Performance Management Guide

Performance Management Guide
AIX 5L Version 5.2
Performance Management Guide
򔻐򗗠򙳰
AIX 5L Version 5.2
Performance Management Guide
򔻐򗗠򙳰
Note
Before using this information and the product it supports, read the information in Appendix H, “Notices” on page 417.
Fifth Edition (October 2002)
This edition applies to AIX 5L Version 5.2 and to all subsequent releases of this product until otherwise indicated in
new editions.
A reader’s comment form is provided at the back of this publication. If the form has been removed, address
comments to Information Development, Department H6DS-905-6C006, 11501 Burnet Road, Austin, Texas
78758-3493. To send comments electronically, use this commercial Internet address: [email protected] Any
information that you supply may be used without incurring any obligation to you.
© Copyright International Business Machines Corporation 1997, 2002. All rights reserved.
US Government Users Restricted Rights – Use, duplication or disclosure restricted by GSA ADP Schedule Contract
with IBM Corp.
Contents
About This Book . . . .
Who Should Use This Book .
Highlighting . . . . . . .
Case-Sensitivity in AIX. . .
ISO 9000 . . . . . . .
Related Publications . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
vii
vii
vii
vii
vii
vii
Chapter 1. Tuning Enhancements for AIX 5.2 . . .
AIX Kernel Tuning Parameter Modifications . . . .
Modifications to vmtune and schedtune . . . . . .
Enhancements to no and nfso . . . . . . . . .
AIX 5.2 Migration Installation and Compatibility Mode.
System Recovery Procedures . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Chapter 2. Performance Concepts . . . .
How Fast is That Computer?. . . . . . .
Understanding the Workload . . . . . . .
Program Execution Dynamics . . . . . .
System Dynamics . . . . . . . . . .
Introducing the Performance-Tuning Process
Performance Benchmarking. . . . . . .
Related Information. . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. 5
. 5
. 5
. 6
. 10
. 11
. 15
. 16
Chapter 3. Resource Management Overview . . . . . .
Performance Overview of the processor Scheduler . . . . .
Performance Overview of the Virtual Memory Manager (VMM) .
Performance Overview of Fixed-Disk Storage Management . .
Support for Pinned Memory. . . . . . . . . . . . . .
Large Page Support . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
19
19
25
33
36
37
Chapter 4. Introduction to Multiprocessing . . . . . .
Symmetrical Multiprocessor (SMP) Concepts and Architecture
SMP Performance Issues . . . . . . . . . . . . .
SMP Workloads . . . . . . . . . . . . . . . . .
SMP Thread Scheduling . . . . . . . . . . . . . .
Thread Tuning . . . . . . . . . . . . . . . . .
SMP Tools . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
39
39
45
46
49
51
55
Chapter 5. Planning and Implementing for Performance
Identifying the Components of the Workload . . . . .
Documenting Performance Requirements . . . . . .
Estimating the Resource Requirements of the Workload .
Designing and Implementing Efficient Programs . . . .
Using Performance-Related Installation Guidelines . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
63
63
64
64
70
78
Chapter 6. System Monitoring and Initial Performance Diagnosis
The Case for Continuous Performance Monitoring . . . . . . .
Using the vmstat, iostat, netstat, and sar Commands . . . . . .
Using the topas Monitor . . . . . . . . . . . . . . . . .
Using the Performance Diagnostic Tool . . . . . . . . . . .
Using the Performance Toolbox . . . . . . . . . . . . . .
Determining the Kind of Performance Problem Reported . . . . .
Identifying the Performance-Limiting Resource . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
83
83
83
85
88
88
90
92
© Copyright IBM Corp. 1997, 2002
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1
1
2
2
3
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
iii
Managing Workload
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Chapter 7. Monitoring and Tuning CPU Use . . . . . . . . . .
Monitoring CPU Use . . . . . . . . . . . . . . . . . . .
Using the time Command to Measure CPU Use . . . . . . . . .
Identifying CPU-Intensive Programs . . . . . . . . . . . . .
Using the tprof Program to Analyze Programs for CPU Use . . . .
Using the pprof Command to Measure CPU usage of Kernel Threads.
Detecting Instruction Emulation with the emstat Tool . . . . . . .
Detecting Alignment Exceptions with the alstat Tool . . . . . . .
Restructuring Executable Programs with the fdpr Program . . . . .
Controlling Contention for the CPU . . . . . . . . . . . . .
CPU-Efficient User ID Administration (The mkpasswd Command) . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . . 99
. . . 99
. . . 106
. . . 108
. . . 110
. . . 117
. . . 119
. . . 121
. . . 121
. . . 123
. . . 129
Chapter 8. Monitoring and Tuning Memory Use . . . . . .
Determining How Much Memory Is Being Used . . . . . . .
Finding Memory-Leaking Programs . . . . . . . . . . .
Assessing Memory Requirements Through the rmss Command .
Tuning VMM Memory Load Control with the schedtune Command
Tuning VMM Page Replacement with the vmtune Command . .
Tuning Paging-Space Thresholds . . . . . . . . . . . .
Choosing a Page Space Allocation Method . . . . . . . .
Using Shared Memory . . . . . . . . . . . . . . . .
Using AIX Memory Affinity Support. . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
131
131
144
145
151
155
160
161
162
163
Chapter 9. File System, Logical Volume, and Disk I/O Performance
Monitoring Disk I/O . . . . . . . . . . . . . . . . . . .
Guidelines for Tuning File Systems . . . . . . . . . . . . .
Changing File System Attributes that Affect Performance . . . . .
Changing Logical Volume Attributes That Affect Performance . . . .
Physical Volume Considerations . . . . . . . . . . . . . .
Volume Group Recommendations . . . . . . . . . . . . . .
Reorganizing Logical Volumes . . . . . . . . . . . . . . .
Reorganizing File Systems . . . . . . . . . . . . . . . .
Reorganizing File System Log and Log Logical Volumes . . . . .
Tuning with vmtune . . . . . . . . . . . . . . . . . . .
Using Disk-I/O Pacing . . . . . . . . . . . . . . . . . .
Tuning Logical Volume Striping . . . . . . . . . . . . . . .
Tuning Asynchronous Disk I/O . . . . . . . . . . . . . . .
Tuning Direct I/O . . . . . . . . . . . . . . . . . . . .
Using Raw Disk I/O . . . . . . . . . . . . . . . . . . .
Using sync/fsync Calls . . . . . . . . . . . . . . . . . .
Setting SCSI-Adapter and Disk-Device Queue Limits . . . . . . .
Expanding the Configuration . . . . . . . . . . . . . . . .
Using RAID . . . . . . . . . . . . . . . . . . . . . .
Using SSA . . . . . . . . . . . . . . . . . . . . . .
Using Fast Write Cache . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
165
165
183
190
192
195
195
196
197
199
200
204
206
209
210
211
212
212
213
213
216
217
Chapter 10. Monitoring and Tuning Communications I/O Use
UDP and TCP/IP Performance Overview . . . . . . . . .
Analyzing Network Performance . . . . . . . . . . . .
Tuning TCP and UDP Performance . . . . . . . . . . .
Tuning mbuf Pool Performance . . . . . . . . . . . . .
Tuning Asynchronous Connections for High-Speed Transfers . .
Tuning Name Resolution . . . . . . . . . . . . . . .
Improving telnetd/rlogind Performance . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
219
219
230
256
274
276
277
278
iv
Performance Management Guide
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Tuning the SP Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
Chapter 11. Monitoring and Tuning NFS Use
NFS Overview . . . . . . . . . . . .
Analyzing NFS Performance . . . . . . .
Tuning for NFS Performance . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
283
283
287
294
Chapter 12. Monitoring and Tuning Java
What is Java? . . . . . . . . . . .
Why Java? . . . . . . . . . . . .
Java Performance Guidelines . . . . .
Monitoring Java . . . . . . . . . .
Tuning Java . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
309
309
309
309
310
310
Chapter 13. Analyzing Performance with the Trace Facility
Understanding the Trace Facility . . . . . . . . . . .
Example of Trace Facility Use . . . . . . . . . . . .
Starting and Controlling Trace from the Command Line . . .
Starting and Controlling Trace from a Program . . . . . .
Using the trcrpt Command to Format a Report . . . . . .
Adding New Trace Events . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
313
313
315
317
318
318
320
Chapter 14. Using Performance Diagnostic Tool (PDT)
Structure of PDT . . . . . . . . . . . . . . .
Scope of PDT Analysis . . . . . . . . . . . . .
Analyzing the PDT Report . . . . . . . . . . . .
Installing and Enabling PDT . . . . . . . . . . .
Customizing PDT . . . . . . . . . . . . . . .
Responding to PDT Report Messages . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
325
325
326
327
330
330
336
Chapter 15. Reporting Performance Problems
Measuring the Baseline . . . . . . . . . .
What is a Performance Problem . . . . . .
Performance Problem Description . . . . . .
Reporting a Performance Problem . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
343
343
344
344
344
Chapter 16. Application Tuning . . . . . .
Profiling . . . . . . . . . . . . . . .
Compiler Optimization Techniques . . . . . .
Optimizing Preprocessors for FORTRAN and C .
Code-Optimization Techniques . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
347
347
352
359
360
Chapter 17. Using POWER4-based Systems . . . .
POWER4 Performance Enhancements . . . . . . .
Scalability Enhancements for POWER4-based Systems .
64-bit Kernel . . . . . . . . . . . . . . . . .
Enhanced Journaled File System (JFS2) . . . . . .
Related Information . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
363
363
364
365
365
366
Chapter 18. Monitoring and Tuning Partitions . .
Performance Considerations with Logical Partitioning .
Workload Management . . . . . . . . . . . .
LPAR Performance Impacts . . . . . . . . . .
CPUs in a partition . . . . . . . . . . . . .
Application Considerations. . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
367
367
368
369
369
370
.
.
.
.
.
.
.
.
.
.
.
.
Contents
v
Appendix A. Monitoring and Tuning Commands and Subroutines.
Performance Reporting and Analysis Commands . . . . . . . .
Performance Tuning Commands . . . . . . . . . . . . . .
Performance-Related Subroutines . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
373
373
376
377
Appendix B. Efficient Use of the ld Command
Rebindable Executable Programs . . . . . .
Prebound Subroutine Libraries . . . . . . .
Examples . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
379
379
379
379
Appendix C. Accessing the Processor Timer .
POWER-based-Architecture-Unique Timer Access
Accessing Timer Registers in PowerPC Systems
Example of the second Subroutine. . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
381
382
383
383
Appendix D. Determining CPU Speed . . . . . . . . . . . . . . . . . . . . . . . . 385
Appendix E. National Language
Programming Considerations . .
Some Simplifying Rules. . . .
Setting the Locale . . . . . .
Support:
. . . .
. . . .
. . . .
Appendix F. Summary of Tunable
Environment Variables . . . . .
Kernel Tunable Parameters . . .
Network Tunable Parameters. . .
Locale
. . .
. . .
. . .
Parameters
. . . . .
. . . . .
. . . . .
.
.
.
.
versus
. . .
. . .
. . .
.
.
.
.
.
.
.
.
Speed
. . .
. . .
. . .
.
.
.
.
Appendix G. Test Case Scenarios . . . . . . . . .
Improve NFS Client Large File Writing Performance . . .
Improve Tivoli Storage Manager (TSM) Backup Performance
Streamline Security Subroutines with Password Indexing .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
389
389
390
390
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
391
391
401
407
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
413
413
414
415
Appendix H. Notices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
Trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
vi
Performance Management Guide
About This Book
This book provides information on concepts, tools, and techniques for assessing and tuning the
performance of systems. Topics covered include efficient system and application design and
implementation, as well as post-implementation tuning of CPU use, memory use, disk I/O, and
communications I/O. Most of the tuning recommendations were developed or validated on AIX Version 4.
Who Should Use This Book
This book is intended for application programmers, customer engineers, experienced end users, enterprise
system administrators, experienced system administrators, system engineers, and system programmers
concerned with performance tuning of operating systems. You should be familiar with the operating system
environment. Introductory sections are included to assist those who are less experienced and to acquaint
experienced users with performance-tuning terminology.
Highlighting
The following highlighting conventions are used in this book:
Bold
Identifies commands, subroutines, keywords, files, structures, directories, and other items
whose names are predefined by the system. Also identifies graphical objects such as buttons,
labels, and icons that the user selects.
Identifies parameters whose actual names or values are to be supplied by the user.
Identifies examples of specific data values, examples of text similar to what you might see
displayed, examples of portions of program code similar to what you might write as a
programmer, messages from the system, or information you should actually type.
Italics
Monospace
Case-Sensitivity in AIX
Everything in the AIX operating system is case-sensitive, which means that it distinguishes between
uppercase and lowercase letters. For example, you can use the ls command to list files. If you type LS, the
system responds that the command is ″not found.″ Likewise, FILEA, FiLea, and filea are three distinct file
names, even if they reside in the same directory. To avoid causing undesirable actions to be performed,
always ensure that you use the correct case.
ISO 9000
ISO 9000 registered quality systems were used in the development and manufacturing of this product.
Related Publications
The following books contain information about or related to performance monitoring:
v AIX 5L Version 5.2 Commands Reference
v
v
v
v
v
v
v
AIX
AIX
AIX
AIX
AIX
AIX
AIX
5L
5L
5L
5L
5L
5L
5L
Version
Version
Version
Version
Version
Version
Version
5.2
5.2
5.2
5.2
5.2
5.2
5.2
Technical Reference
Files Reference
System User’s Guide: Operating System and Devices
System User’s Guide: Communications and Networks
System Management Guide: Operating System and Devices
System Management Guide: Communications and Networks
General Programming Concepts: Writing and Debugging Programs
v Performance Toolbox Version 2 and 3 for AIX: Guide and Reference
© Copyright IBM Corp. 1997, 2002
vii
v PCI Adapter Placement Reference, order number SA38-0538
viii
Performance Management Guide
Chapter 1. Tuning Enhancements for AIX 5.2
There are a few performance tuning changes being introduced in AIX 5.2 that are discussed in this
section:
v AIX kernel tuning parameters
v Modifications to vmtune and schedtune
v Enhancements to no and nfso
v AIX 5.2 Migration Installation and Compatibility Mode
v System Recovery Procedures
AIX Kernel Tuning Parameter Modifications
AIX 5.2 introduces a new method that is more flexible and centralized for setting most of the AIX kernel
tuning parameters. It is now possible to make permanent changes without having to edit any rc files. This
is achieved by placing the reboot values for all tunable parameters in a new stanza file,
/etc/tunables/nextboot. When the machine is rebooted, the values in that file are automatically applied.
Another stanza file, /etc/tunables/lastboot is automatically generated with all the values as they were set
just after the reboot. This provides the capability to return to those values at any time. The log file for any
changes made or impossible to make during reboot is stored in /etc/tunables/lastboot.log. There are sets
of SMIT panels and a WebSm plug-in also available to manipulate current and reboot values for all tuning
parameters as well as the files in the /etc/tunables directory.
There are four new commands introduced in AIX 5.2 to modify the tunables files. The tunsave command
is used to save values to a stanza file. The tunrestore command is used to apply a file, for example, to
change all tunables parameter values to those listed in a file. The command tuncheck must be used to
validate a file created manually and the tundefault command is available to reset tunable parameters to
their default values. All four commands work on both current and reboot tunables parameters values. See
the respective man pages for more information.
For more information about any of these kernel tuning parameter modifications, see the kernel tuning
section in AIX 5L Version 5.2 Performance Tools Guide and Reference.
Modifications to vmtune and schedtune
Vmtune and schedtune are being replaced by the newly supported commands called vmo, ioo, and
schedo. Both vmo and ioo together replace vmtune, while schedo replaces schedtune. All existing
parameters are covered by the new commands.
The ioo command will handle all the I/O related tuning parameters, while the vmo command will handle all
the other VMM parameters previously managed by vmtune. All three commands are part of the new fileset
bos.perf.tune which also contains tunsave, tunrestore, tuncheck, and tundefault. The
bos.adt.samples fileset will still include the vmtune and schedtune commands, which will simply be
compatibility shell scripts calling vmo, ioo, and schedo as appropriate. The compatibility scripts only
support changes to parameters which can be changed interactively. That is, parameters that need
bosboot and then require a reboot of the machine to be effective are no longer supported by the vmtune
script. To change those parameters, users must now use vmo -r. The options (all from vmtune) and
parameters in question are as follows:
vmtune option
parameter name
new command
-C 0|1
page coloring
vmo -r -o pagecoloring=0|1
© Copyright IBM Corp. 1997, 2002
1
-g n1
-L n2
large page size
number of large pages
to reserve
vmo -r -o lpg_size=n1 -o lpg_regions=n2
-m n
memory pools
vmo -r -o mempools=n
-v n
number of frames per
memory pool
vmo -r -o framesets=n
-i n
interval for special data
segment identifiers
vmo -r -o spec_dataseg_int=n
-V n
number of special data
segment identifiers to
reserve
vmo -r -o num_spec_dataseg
-y 0|1
p690 memory affinity
vmo -r -o memory_affinity=0|1
Enhancements to no and nfso
The no and nfso commands have been enhanced to support making permanent changes to tunable
parameters. They now interact with the /etc/tunables/nextboot file to achieve this new functionality. They
both also have a new -h flag which can be used to display help about any parameter. The content of the
help includes the purpose of the parameter, the possible values (default, range and type), and diagnostic
and tuning information to decide when to change the parameter value. This information is also listed
entirely in the respective man pages. Note that all five tuning commands (ioo, nfso, no, vmo, and
schedo) use the same common syntax. See the respective man pages for more details and also the
complete list of tuning parameters supported.
AIX 5.2 Migration Installation and Compatibility Mode
When a machine is migrated to AIX 5.2 from a previous version of AIX, it is automatically set to run in
compatibility mode where the current behavior of the tuning commands, with the exception of the vmtune
parameters mentioned previously, is completely preserved. Contrary to the normal AIX 5.2 tuning mode,
where permanent tunable parameter settings are set by applying values from the /etc/tunables/nextboot
file, in compatibility mode, it is still possible to make permanent changes to tunable parameters by
embedding calls to tuning commands in scripts called during the boot process. The only perceivable
difference is that the /etc/tunables/lastboot and /etc/tunables/lastboot.log files are created during
reboot. The lastboot.log file contains only a warning that AIX is currently running in compatibility mode
and that the nextboot file has not been applied. Furthermore, except for parameters of type Bosboot (see
vmtune and schedtune changes section), none of the new reboot and permanent options ( the -r and -p
flags respectively) of the tuning commands are really meaningful because the content of the file is not
applied at reboot time. The tuning commands are not controlling the reboot values of parameters like they
would in non-compatibility mode. Parameters of type Bosboot are preserved during migration, stored in the
/etc/tunables/nextboot file, and can be modified using the -r option, even when running in compatibility
mode. The /etc/tunables/nextboot file should therefore not be deleted.
The compatibility mode is controlled by a new sys0 attribute called pre520tune, which is automatically set
to enable during a migration installation. In the case of a fresh installation of AIX 5.2, the attribute is set to
disable. In that mode, embedded calls to tuning commands in scripts called during reboot are overwritten
by the content of the nextboot file. The current setting of the pre520tune attribute can be viewed by
running the following command:
# lsattr -E -l sys0
and changed either using the following command:
# chdev -l sys0 -a pre520tune=disable
or using SMIT or Websm.
2
Performance Management Guide
When the compatibility mode is disabled, the other visible change is that the following no parameters,
which are all of type Reboot (they can only be changed during reboot), cannot be changed any more
without using the new -r flag.:
v
v
v
v
v
v
v
v
arptab_bsiz
arptab_nb
extendednetstats
ifsize
inet_stack_size
ipqmaxlen
nstrpush
pseintrstack
Switching to the non-compatibility mode while preserving the current reboot settings can be done by first
changing pre520tune, and then by running the following command:
# tunrestore -r -f lastboot
which will copy the content of the lastboot file to the nextboot file. See AIX 5L Version 5.2 Kernel tuning
in the AIX 5L Version 5.2 Performance Tools Guide and Reference for details about the new AIX 5.2
tuning mode.
System Recovery Procedures
If a machine is unstable after rebooting and pre520tune is set to enable, users should delete the offending
calls to tuning commands from scripts called during reboot. To detect which parameters are set during
reboot, simply look at the /etc/tunables/lastboot file and search for parameters not marked with #
DEFAULT VALUE. For more information on the content of tunable files, see the tunables File Format
section in AIX 5L Version 5.2 Files Reference.
Alternatively, to reset all the tunable parameters to their default values, delete the /etc/tunables/nextboot
file, set pre520tune to disable, run the bosboot command, and reboot the machine.
Chapter 1. Tuning Enhancements for AIX 5.2
3
4
Performance Management Guide
Chapter 2. Performance Concepts
Everyone who uses a computer has an opinion about its performance. Unfortunately, those opinions are
often based on oversimplified ideas about the dynamics of program execution. Uninformed intuition can
lead to expensive and inaccurate guesses about the capacity of a system and the solutions to the
perceived performance problems.
This chapter describes the dynamics of program execution and provides a conceptual framework for
evaluating system performance. It contains the following major sections:
v How Fast is That Computer?
v Understanding the Workload
v
v
v
v
v
Program Execution Dynamics
System Dynamics
Introducing the Performance-Tuning Process
Performance Benchmarking
Related Information
How Fast is That Computer?
Using words like speed and fast to describe contemporary computers, while accepted by precedent, is
extreme oversimplification. There was a time when a programmer could read a program, calculate the sum
of the instruction times, and confidently predict how long it would take the computer to run that program.
Thousands of programmers and engineers have spent the last 30 years making such straightforward
calculations impossible, or at least meaningless.
Today’s computers are more powerful than their predecessors, not just because they use integrated
circuits instead of vacuum tubes and have far shorter cycle times, but because of innumerable hardware
and software architectural inventions. Each advance in integrated-circuit density brings an advance in
computer performance, not just because it allows the same logic to work in a smaller space with a faster
system clock, but because it gives engineers more space in which to implement ideas. In short, computers
have gained capacity by becoming more complex as well as quicker.
The complexity of modern computers and their operating systems is matched by the complexity of the
environment in which they operate. In addition to running individual programs, today’s computer deals with
varying numbers of unpredictably timed interrupts from I/O and communications devices. To the extent that
the engineers’ ideas were based on an assumption of a single program running in a standalone machine,
they may be partly defeated by the randomness of the real world. To the extent that those ideas were
intended to deal with randomness, they may win back some of the loss. The wins and losses change from
program to program and from moment to moment.
The result of all these hardware and software wins and losses is the performance of the system. The
speed of the system is the rate at which it can handle a specific sequence of demands. If the demands
mesh well with the system’s hardware and software architectures, we can say, ″The system runs this
workload fast.″ We cannot say, ″The system is fast,″ or at least we should not.
Understanding the Workload
An accurate and complete definition of the system’s workload is critical to predicting or understanding its
performance. A difference in workload can cause far more variation in the measured performance of a
system than differences in CPU clock speed or random access memory (RAM) size. The workload
definition must include not only the type and rate of requests to the system, but also the exact software
packages and in-house application programs to be executed.
© Copyright IBM Corp. 1997, 2002
5
Whenever possible, observe the current users of existing applications to get authentic, real-world
measurements of the rates at which users interact with their workstations or terminals.
Make sure that you include the work that your system is doing ″under the covers.″ For example, if your
system contains file systems that are NFS-mounted and frequently accessed by other systems, handling
those accesses is probably a significant fraction of the overall workload, even though your system is not
officially a server.
Industry-Standard Benchmarks: A Risky Shortcut
A benchmark is a workload that has been standardized to allow comparisons among dissimilar systems.
Any benchmark that has been in existence long enough to become industry-standard has been studied
exhaustively by systems developers. Operating systems, compilers, and in some cases hardware, have
been tuned to run the benchmark with lightning speed.
Unfortunately, few real workloads duplicate the exact algorithms and environment of a benchmark. Even
those industry-standard benchmarks that were originally derived from real applications might have been
simplified and homogenized to make them portable to a wide variety of hardware platforms. The
environment in which they run has been constrained in the interests of reproducible measurements.
Any reasoning similar to ″System A is rated at 50 percent more MegaThings than System B, so System A
should run my program 50 percent faster than System B″ might be a tempting shortcut, but may be
inaccurate. There is no benchmark with such universal applicability. The only valid use for
industry-standard benchmarks is to narrow the field of candidate systems that will be subjected to a
serious evaluation. There is no substitute for developing a clear understanding of your workload and its
performance in systems under consideration.
Performance Objectives
After defining the workload that the system will have to process, you can choose performance criteria and
set performance objectives based on those criteria. The main overall performance criteria of computer
systems are response time and throughput.
Response time is the elapsed time between when a request is submitted and when the response from that
request is returned. Examples include how long a database query takes, or how long it takes to echo
characters to the terminal, or how long it takes to access a Web page.
Throughput is a measure of the amount of work over a period of time. In other words, it is the number of
workload operations that can be accomplished per unit of time. Examples include database transactions
per minute, kilobytes of a file transferred per second, kilobytes of a file read or written per second, or Web
server hits per minute.
The relationship between these metrics is complex. In some cases, you might have to trade off one
against the other. In other situations, a single change can improve both. Sometimes you can have higher
throughput at the cost of response time or better response time at the cost of throughput. Acceptable
performance is based on reasonable throughput combined with reasonable response time.
In planning for or tuning any system, make sure that you have clear objectives for both response time and
throughput when processing the specified workload. Otherwise, you risk spending analysis time and
resource dollars improving an aspect of system performance that is of secondary importance.
Program Execution Dynamics
Normally, an application programmer thinks of the running program as an uninterrupted sequence of
instructions that perform a specified function. Great amounts of inventiveness and effort have been
expended on the operating system and hardware to ensure that programmers are not distracted from this
idealized view by irrelevant space, speed, and multiprogramming or multiprocessing considerations. If the
6
Performance Management Guide
programmer is seduced by this comfortable illusion, the resulting program might be unnecessarily
expensive to run and might not meet its performance objectives.
To examine clearly the performance characteristics of a workload, a dynamic, rather than a static, model of
program execution is needed, as shown in the following figure.
Figure 1. Program Execution Hierarchy. The figure is a triangle on its base. The left side represents hardware entities
that are matched to the appropriate operating system entity on the right side. A program must go from the lowest level
of being stored on disk, to the highest level being the processor running program instructions. For instance, from
bottom to top, the disk hardware entity holds executable programs; real memory holds waiting operating system
threads and interrupt handlers; the translation lookaside buffer holds detachable threads; cache contains the currently
dispatched thread and the processor pipeline and registers contain the current instruction.
To run, a program must make its way up both the hardware and operating-system hierarchies, more or
less in parallel. Each element in the hardware hierarchy is scarcer and more expensive than the element
below it. Not only does the program have to contend with other programs for each resource, the transition
from one level to the next takes time. To understand the dynamics of program execution, you need a basic
understanding of each of the levels in the hierarchy.
Hardware Hierarchy
Usually, the time required to move from one hardware level to another consists primarily of the latency of
the lower level (the time from the issuing of a request to the receipt of the first data).
Fixed Disks
The slowest operation for a running program (other than waiting for a human keystroke) on a standalone
system is obtaining code or data from a disk, for the following reasons:
v The disk controller must be directed to access the specified blocks (queuing delay).
v The disk arm must seek to the correct cylinder (seek latency).
v The read/write heads must wait until the correct block rotates under them (rotational latency).
v The data must be transmitted to the controller (transmission time) and then conveyed to the application
program (interrupt-handling time).
Chapter 2. Performance Concepts
7
Disk operations can have many causes besides explicit read or write requests in the program.
System-tuning activities frequently prove to be hunts for unnecessary disk I/O.
Real Memory
Real Memory, often referred to as RAM, is fast compared to disk, but much more expensive per byte.
Operating systems try to keep in RAM the code and data that are currently in use, spilling any excess onto
disk (or never bringing them into RAM in the first place).
RAM is not necessarily fast compared to the processor. Typically a RAM latency of dozens of processor
cycles occurs between the time the hardware recognizes the need for a RAM access and the time the
data or instruction is available to the processor.
If the access is to a page of virtual memory that has been spilled to disk (or has not been brought in yet),
a page fault occurs, and the execution of the program is suspended until the page has been read in from
disk.
Translation Lookaside Buffer (TLB)
One of the ways programmers are insulated from the physical limitations of the system is the
implementation of virtual memory. The programmer designs and codes the program as though the memory
were very large, and the system takes responsibility for translating the program’s virtual addresses for
instructions and data into the real addresses that are needed to get the instructions and data from RAM.
Because this address-translation process can be time-consuming, the system keeps the real addresses of
recently accessed virtual-memory pages in a cache called the translation lookaside buffer (TLB).
As long as the running program continues to access a small set of program and data pages, the full
virtual-to-real page-address translation does not need to be redone for each RAM access. When the
program tries to access a virtual-memory page that does not have a TLB entry (a TLB miss), dozens of
processor cycles (the TLB-miss latency) are usually required to perform the address translation.
Caches
To minimize the number of times the program has to experience the RAM latency, systems incorporate
caches for instructions and data. If the required instruction or data is already in the cache (a cache hit), it
is available to the processor on the next cycle (that is, no delay occurs). Otherwise (a cache miss), the
RAM latency occurs.
In some systems, there are two or three levels of cache, usually called L1, L2, and L3. If a particular
storage reference results in an L1 miss, then L2 is checked. If L2 generates a miss, then the reference
goes to the next level, either L3 if present or RAM.
Cache sizes and structures vary by model, but the principles of using them efficiently are identical.
Pipeline and Registers
A pipelined, superscalar architecture makes possible, under certain circumstances, the simultaneous
processing of multiple instructions. Large sets of general-purpose registers and floating-point registers
make it possible to keep considerable amounts of the program’s data in registers, rather than continually
storing and reloading.
The optimizing compilers are designed to take maximum advantage of these capabilities. The compilers’
optimization functions should always be used when generating production programs, however small the
programs are. The Optimization and Tuning Guide for XL Fortran, XL C and XL C++ describes how
programs can be tuned for maximum performance.
Software Hierarchy
To run, a program must also progress through a series of steps in the software hierarchy.
8
Performance Management Guide
Executable Programs
When a user requests a program to run, the operating system performs a number of operations to
transform the executable program on disk to a running program. First, the directories in the user’s current
PATH environment variable must be scanned to find the correct copy of the program. Then, the system
loader (not to be confused with the ld command, which is the binder) must resolve any external references
from the program to shared libraries.
To represent the user’s request, the operating system creates a process, which is a set of resources, such
as a private virtual address segment, required by any running program.
The operating system also automatically creates a single thread within that process. A thread is the current
execution state of a single instance of a program. In AIX Version 4 and later, access to the processor and
other resources is allocated on a thread basis, rather than a process basis. Multiple threads can be
created within a process by the application program. Those threads share the resources owned by the
process within which they are running.
Finally, the system branches to the entry point of the program. If the program page that contains the entry
point is not already in memory (as it might be if the program had been recently compiled, executed, or
copied), the resulting page-fault interrupt causes the page to be read from its backing storage.
Interrupt Handlers
The mechanism for notifying the operating system that an external event has taken place is to interrupt the
currently running thread and transfer control to an interrupt handler. Before the interrupt handler can run,
enough of the hardware state must be saved to ensure that the system can restore the context of the
thread after interrupt handling is complete. Newly invoked interrupt handlers experience all of the delays of
moving up the hardware hierarchy (except page faults). Unless the interrupt handler was run very recently
(or the intervening programs were very economical), it is unlikely that any of its code or data remains in
the TLBs or the caches.
When the interrupted thread is dispatched again, its execution context (such as register contents) is
logically restored, so that it functions correctly. However, the contents of the TLBs and caches must be
reconstructed on the basis of the program’s subsequent demands. Thus, both the interrupt handler and the
interrupted thread can experience significant cache-miss and TLB-miss delays as a result of the interrupt.
Waiting Threads
Whenever an executing program makes a request that cannot be satisfied immediately, such as a
synchronous I/O operation (either explicit or as the result of a page fault), that thread is put in a wait state
until the request is complete. Normally, this results in another set of TLB and cache latencies, in addition
to the time required for the request itself.
Dispatchable Threads
When a thread is dispatchable, but not actually running, it is accomplishing nothing useful. Worse, other
threads that are running may cause the thread’s cache lines to be reused and real memory pages to be
reclaimed, resulting in even more delays when the thread is finally dispatched.
Currently Dispatched Threads
The scheduler chooses the thread that has the strongest claim to the use of the processor. The
considerations that affect that choice are discussed in Performance Overview of the CPU Scheduler. When
the thread is dispatched, the logical state of the processor is restored to the state that was in effect when
the thread was interrupted.
Current Instructions
Most of the machine instructions are capable of executing in a single processor cycle, if no TLB or cache
miss occurs. In contrast, if a program branches rapidly to different areas of the program and accesses
data from a large number of different areas, causing high TLB and cache-miss rates, the average number
of processor cycles per instruction (CPI) executed might be much greater than one. The program is said to
exhibit poor locality of reference. It might be using the minimum number of instructions necessary to do its
Chapter 2. Performance Concepts
9
job, but consuming an unnecessarily large number of cycles. In part because of this poor correlation
between number of instructions and number of cycles, reviewing a program listing to calculate path length
no longer yields a time value directly. While a shorter path is usually faster than a longer path, the speed
ratio can be very different from the path-length ratio.
The compilers rearrange code in sophisticated ways to minimize the number of cycles required for the
execution of the program. The programmer seeking maximum performance must be primarily concerned
with ensuring that the compiler has all the information necessary to optimize effectively, rather than trying
to second-guess the compiler’s optimization techniques (see Effective Use of Preprocessors and the
Compilers). The real measure of optimization effectiveness is the performance of an authentic workload.
System Dynamics
It is not enough to create the most efficient possible individual programs. In many cases, the actual
programs being run were created outside of the control of the person who is responsible for meeting the
organization’s performance objectives. Further, most of the levels of the hierarchy described in Program
Execution Dynamics are managed by one or more parts of the operating system. In any case, after the
application programs have been acquired, or implemented as efficiently as possible, further improvement
in the overall performance of the system becomes a matter of system tuning. The main components that
are subject to system-level tuning are:
Communications I/O
Depending on the type of workload and the type of communications link, it might be necessary to
tune one or more of the communications device drivers, TCP/IP, or NFS.
Fixed Disk
The Logical Volume Manager (LVM) controls the placement of file systems and paging spaces on
the disk, which can significantly affect the amount of seek latency the system experiences. The
disk device drivers control the order in which I/O requests are acted on.
Real Memory
The Virtual Memory Manager (VMM) controls the pool of free real-memory frames and determines
when and from whom to steal frames to replenish the pool.
Running Thread
The scheduler determines which dispatchable entity should next receive control. In AIX, the
dispatchable entity is a thread. See Thread Support.
Classes of Workload
Workloads tend to fall naturally into a small number of classes. The types listed below are sometimes
used to categorize systems. However, because a single system is often called upon to process multiple
classes, workload seems more apt in the context of performance.
Multiuser
A workload that consists of a number of users submitting work through individual terminals.
Typically, the performance objectives of such a workload are either to maximize system throughput
while preserving a specified worst-case response time or to obtain the best possible response time
for a fairly constant workload.
Server
A workload that consists of requests from other systems. For example, a file-server workload is
mostly disk read/write requests. In essence, it is the disk-I/O component of a multiuser workload
(plus NFS or other I/O activity), so the same objective of maximum throughput within a given
response-time limit applies. Other server workloads consist of compute-intensive programs,
database transactions, printer jobs, and so on.
Workstation
A workload that consists of a single user submitting work through the native keyboard and
10
Performance Management Guide
receiving results on the native display of the system. Typically, the highest-priority performance
objective of such a workload is minimum response time to the user’s requests.
When a single system is processing workloads of more than one type, a clear understanding must exist
between the users and the performance analyst as to the relative priorities of the possibly conflicting
performance objectives of the different workloads.
Introducing the Performance-Tuning Process
Performance tuning is primarily a matter of resource management and correct system-parameter setting.
Tuning the workload and the system for efficient resource use consists of the following steps:
1. Identifying the workloads on the system
2. Setting objectives:
a. Determining how the results will be measured
b. Quantifying and prioritizing the objectives
3. Identifying the critical resources that limit the system’s performance
4. Minimizing the workload’s critical-resource requirements:
a. Using the most appropriate resource, if there is a choice
b. Reducing the critical-resource requirements of individual programs or system functions
c. Structuring for parallel resource use
5. Modifying the allocation of resources to reflect priorities
a. Changing the priority or resource limits of individual programs
b. Changing the settings of system resource-management parameters
6. Repeating steps 3 through 5 until objectives are met (or resources are saturated)
7. Applying additional resources, if necessary
There are appropriate tools for each phase of system performance management (see Appendix A.
Monitoring and Tuning Commands and Subroutines). Some of the tools are available from IBM; others are
the products of third parties. The following figure illustrates the phases of performance management in a
simple LAN environment.
Figure 2. Performance Phases. The figure uses five weighted circles to illustrate the steps of performance tuning a
system; plan, install, monitor, tune, and expand. Each circle represents the system in various states of performance;
idle, unbalanced, balanced, and overloaded. Essentially, you expand a system that is overloaded, tune a system until
it is balanced, monitor an unbalanced system and install for more resources when an expansion is necessary.
Chapter 2. Performance Concepts
11
Identifying the Workloads
It is essential that all of the work performed by the system be identified. Especially in LAN-connected
systems, a complex set of cross-mounted file systems can easily develop with only informal agreement
among the users of the systems. These file systems must be identified and taken into account as part of
any tuning activity.
With multiuser workloads, the analyst must quantify both the typical and peak request rates. It is also
important to be realistic about the proportion of the time that a user is actually interacting with the terminal.
An important element of this identification stage is determining whether the measurement and tuning
activity has to be done on the production system or can be accomplished on another system (or off-shift)
with a simulated version of the actual workload. The analyst must weigh the greater authenticity of results
from a production environment against the flexibility of the nonproduction environment, where the analyst
can perform experiments that risk performance degradation or worse.
Setting Objectives
Although you can set objectives in terms of measurable quantities, the actual desired result is often
subjective, such as satisfactory response time. Further, the analyst must resist the temptation to tune what
is measurable rather than what is important. If no system-provided measurement corresponds to the
desired improvement, that measurement must be devised.
The most valuable aspect of quantifying the objectives is not selecting numbers to be achieved, but
making a public decision about the relative importance of (usually) multiple objectives. Unless these
priorities are set in advance, and understood by everyone concerned, the analyst cannot make trade-off
decisions without incessant consultation. The analyst is also apt to be surprised by the reaction of users or
management to aspects of performance that have been ignored. If the support and use of the system
crosses organizational boundaries, you might need a written service-level agreement between the
providers and the users to ensure that there is a clear common understanding of the performance
objectives and priorities.
Identifying Critical Resources
In general, the performance of a given workload is determined by the availability and speed of one or two
critical system resources. The analyst must identify those resources correctly or risk falling into an endless
trial-and-error operation.
Systems have both real and logical resources. Critical real resources are generally easier to identify,
because more system performance tools are available to assess the utilization of real resources. The real
resources that most often affect performance are as follows:
v CPU cycles
v Memory
v I/O bus
v Various adapters
v Disk arms
v Disk space
v Network access
Logical resources are less readily identified. Logical resources are generally programming abstractions that
partition real resources. The partitioning is done to share and manage the real resource.
Some examples of real resources and the logical resources built on them are as follows:
CPU
v Processor time slice
12
Performance Management Guide
Memory
v
v
v
v
Page frames
Stacks
Buffers
Queues
v Tables
v Locks and semaphores
Disk Space
v
v
v
v
Logical volumes
File systems
Files
Partitions
Network Access
v Sessions
v Packets
v Channels
It is important to be aware of logical resources as well as real resources. Threads can be blocked by a
lack of logical resources just as for a lack of real resources, and expanding the underlying real resource
does not necessarily ensure that additional logical resources will be created. For example, consider the
NFS block I/O daemon (biod, see Tuning for NFS Performance). A biod daemon on the client is required
to handle each pending NFS remote I/O request. The number of biod daemons therefore limits the
number of NFS I/O operations that can be in progress simultaneously. When a shortage of biod daemons
exists, system instrumentation may indicate that the CPU and communications links are used only slightly.
You may have the false impression that your system is underused (and slow), when in fact you have a
shortage of biod daemons that is constraining the rest of the resources. A biod daemon uses processor
cycles and memory, but you cannot fix this problem simply by adding real memory or converting to a faster
CPU. The solution is to create more of the logical resource (biod daemons).
Logical resources and bottlenecks can be created inadvertently during application development. A method
of passing data or controlling a device may, in effect, create a logical resource. When such resources are
created by accident, there are generally no tools to monitor their use and no interface to control their
allocation. Their existence may not be appreciated until a specific performance problem highlights their
importance.
Minimizing Critical-Resource Requirements
Consider minimizing the workload’s critical-resource requirements at three levels, as discussed below.
Using the Appropriate Resource
The decision to use one resource over another should be done consciously and with specific goals in
mind. An example of a resource choice during application development would be a trade-off of increased
memory consumption for reduced CPU consumption. A common system configuration decision that
demonstrates resource choice is whether to place files locally on an individual workstation or remotely on
a server.
Reducing the Requirement for the Critical Resource
For locally developed applications, the programs can be reviewed for ways to perform the same function
more efficiently or to remove unnecessary function. At a system-management level, low-priority workloads
that are contending for the critical resource can be moved to other systems, run at other times, or
controlled with the Workload Manager.
Chapter 2. Performance Concepts
13
Structuring for Parallel Use of Resources
Because workloads require multiple system resources to run, take advantage of the fact that the resources
are separate and can be consumed in parallel. For example, the operating system read-ahead algorithm
detects the fact that a program is accessing a file sequentially and schedules additional sequential reads
to be done in parallel with the application’s processing of the previous data. Parallelism applies to system
management as well. For example, if an application accesses two or more files at the same time, adding
an additional disk drive might improve the disk-I/O rate if the files that are accessed at the same time are
placed on different drives.
Reflecting Priorities in Resource Allocation
The operating system provides a number of ways to prioritize activities. Some, such as disk pacing, are
set at the system level. Others, such as process priority, can be set by individual users to reflect the
importance they attach to a specific task.
Repeating the Tuning Steps
A truism of performance analysis is that there is always a next bottleneck. Reducing the use of one
resource means that another resource limits throughput or response time. Suppose, for example, we have
a system in which the utilization levels are as follows:
CPU: 90% Disk: 70% Memory 60%
This workload is CPU-bound. If we successfully tune the workload so that the CPU load is reduced from
90 to 45 percent, we might expect a two-fold improvement in performance. Unfortunately, the workload is
now I/O-limited, with utilizations of approximately the following:
CPU: 45% Disk: 90% Memory 60%
The improved CPU utilization allows the programs to submit disk requests sooner, but then we hit the limit
imposed by the disk drive’s capacity. The performance improvement is perhaps 30 percent instead of the
100 percent we had envisioned.
There is always a new critical resource. The important question is whether we have met the performance
objectives with the resources at hand.
Attention: Improper system tuning with vmtune, schedtune, and other tuning commands can result in
unexpected system behavior like degraded system or application performance, or a system hang.
Changes should only be applied when a bottleneck has been identified by performance analysis.
Note: There is no such thing as a general recommendation for performance dependent tuning settings.
Applying Additional Resources
If, after all of the preceding approaches have been exhausted, the performance of the system still does not
meet its objectives, the critical resource must be enhanced or expanded. If the critical resource is logical
and the underlying real resource is adequate, the logical resource can be expanded at no additional cost.
If the critical resource is real, the analyst must investigate some additional questions:
v How much must the critical resource be enhanced or expanded so that it ceases to be a bottleneck?
v Will the performance of the system then meet its objectives, or will another resource become saturated
first?
v If there will be a succession of critical resources, is it more cost-effective to enhance or expand all of
them, or to divide the current workload with another system?
14
Performance Management Guide
Performance Benchmarking
When we attempt to compare the performance of a given piece of software in different environments, we
are subject to a number of possible errors, some technical, some conceptual. This section contains mostly
cautionary information. Other sections of this book discuss the various ways in which elapsed and
process-specific times can be measured.
When we measure the elapsed (wall-clock) time required to process a system call, we get a number that
consists of the following:
v The actual time during which the instructions to perform the service were executing
v Varying amounts of time during which the processor was stalled while waiting for instructions or data
from memory (that is, the cost of cache and TLB misses)
v The time required to access the clock at the beginning and end of the call
v Time consumed by periodic events, such as system timer interrupts
v Time consumed by more or less random events, such as I/O interrupts
To avoid reporting an inaccurate number, we normally measure the workload a number of times. Because
all of the extraneous factors add to the actual processing time, the typical set of measurements has a
curve of the form shown in the following illustration.
Figure 3. Curve for Typical Set of Measurement.
The extreme low end may represent a low-probability optimum caching situation or may be a rounding
effect.
A regularly recurring extraneous event might give the curve a bimodal form (two maxima), as shown in the
following illustration.
Chapter 2. Performance Concepts
15
"Actual" value
Mean
Figure 4. Bimodal Curve
One or two time-consuming interrupts might skew the curve even further, as shown in the following
illustration:
Figure 5. Skewed Curve
The distribution of the measurements about the actual value is not random, and the classic tests of
inferential statistics can be applied only with great caution. Also, depending on the purpose of the
measurement, it may be that neither the mean nor the actual value is an appropriate characterization of
performance.
Related Information
See the following sections for further information:
v CPU
– The operating system’s management of the CPU resource is described in Performance Overview of
the CPU Scheduler.
– Tools and techniques for managing CPU use are documented in Monitoring and Tuning CPU Use.
v Memory
– The architecture of the operating system’s memory management is described in Performance
Overview of the Virtual Memory Manager (VMM).
16
Performance Management Guide
– Tools and techniques for managing memory use are documented in Monitoring and Tuning Memory
Use.
v Disks
– A description of the structure of the operating system’s fixed-disk support appears in Performance
Overview of Fixed-Disk Storage Management.
– Planning information about the relative performance of fixed disks appears in Disk Preinstallation
Guidelines.
– An extensive discussion of monitoring, reorganizing, and expanding disk storage appears in
Monitoring and Tuning Disk I/O Use.
v Network
– Tools and techniques for managing various forms of communication I/O are documented in
Monitoring and Tuning Communications I/O Use and Monitoring and Tuning NFS Use.
Chapter 2. Performance Concepts
17
18
Performance Management Guide
Chapter 3. Resource Management Overview
This chapter describes the components of the operating system that manage the resources that have the
most effect on system performance, and the ways in which these components can be tuned. This chapter
contains the following major sections:
v Performance Overview of the CPU Scheduler.
v Performance Overview of the Virtual Memory Manager (VMM).
v Performance Overview of Fixed-Disk Storage Management.
Specific tuning recommendations appear in the following:
v Chapter 6. Monitoring and Tuning CPU Use.
v Chapter 7. Monitoring and Tuning Memory Use.
v Chapter 8. Monitoring and Tuning Disk I/O Use.
v Chapter 9. Monitoring and Tuning Communications I/O Use.
v Chapter 10. Monitoring and Tuning NFS Use.
Performance Overview of the processor Scheduler
This section discusses performance related topics for the processor Scheduler.
Thread Support
A thread can be thought of as a low-overhead process. It is a dispatchable entity that requires fewer
resources to create than a process. The fundamental dispatchable entity of the AIX Version 4 scheduler is
the thread.
Processes are composed of one or more threads. In fact, workloads migrated directly from earlier releases
of the operating system continue to create and manage processes. Each new process is created with a
single thread that has its parent process priority and contends for the processor with the threads of other
processes. The process owns the resources used in execution; the thread owns only its current state.
When new or modified applications take advantage of the operating system’s thread support to create
additional threads, those threads are created within the context of the process. They share the process’s
private segment and other resources.
A user thread within a process has a specified contention scope. If the contention scope is global, the
thread contends for processor time with all other threads in the system. The thread that is created when a
process is created has global contention scope. If the contention scope is local, the thread contends with
the other threads within the process to be the recipient of the process’s share of processor time.
The algorithm for determining which thread should be run next is called a scheduling policy.
Processes and Threads
A process is an activity within the system that is started by a command, a shell program, or another
process.
Process properties are as follows:
v pid
v pgid
v uid
v gid
© Copyright IBM Corp. 1997, 2002
19
v
v
v
v
v
environment
cwd
file descriptors
signal actions
process statistics
v nice
These properties are defined in /usr/include/sys/proc.h.
Thread properties are as follows:
v
v
v
v
v
v
stack
scheduling policy
scheduling priority
pending signals
blocked signals
thread-specific data
These thread properties are defined in /usr/include/sys/thread.h.
Each process is made up of one or more threads. A thread is a single sequential flow of control. Multiple
threads of control allow an application to overlap operations, such as reading from a terminal and writing
to a file.
Multiple threads of control also allow an application to service requests from multiple users at the same
time. Threads provide these capabilities without the added overhead of multiple processes such as those
created through the fork() system call.
AIX 4.3.1 introduced a fast fork routine called f_fork(). This routine is very useful for multithreaded
applications that will call the exec() subroutine immediately after they would have called the fork()
subroutine. The fork() subroutine is slower because it has to call fork handlers to acquire all the library
locks before actually forking and letting the child run all child handlers to initialize all the locks. The
f_fork() subroutine bypasses these handlers and calls the kfork() system call directly. Web servers are a
good example of an application that can use the f_fork() subroutine.
Process and Thread Priority
The priority management tools manipulate process priority. In AIX Version 4, process priority is simply a
precursor to thread priority. When the fork() subroutine is called, a process and a thread to run in it are
created. The thread has the priority that would have been attributed to the process.
The kernel maintains a priority value (sometimes termed the scheduling priority) for each thread. The
priority value is a positive integer and varies inversely with the importance of the associated thread. That
is, a smaller priority value indicates a more important thread. When the scheduler is looking for a thread to
dispatch, it chooses the dispatchable thread with the smallest priority value.
A thread can be fixed-priority or nonfixed priority. The priority value of a fixed-priority thread is constant,
while the priority value of a nonfixed-priority thread varies based on the minimum priority level for user
threads (a constant 40), the thread’s nice value (20 by default, optionally set by the nice or renice
command), and its processor-usage penalty.
The priority of a thread can be fixed at a certain value, which can have a priority value less than 40, if their
priority is set (fixed) through the setpri() subroutine. These threads are immune to the scheduler
20
Performance Management Guide
recalculation algorithms. If their priority values are fixed to be less than 40, these threads will run and
complete before any user threads can run. For example, a thread with a fixed value of 10 will run before a
thread with a fixed value of 15.
Users can apply the nice command to make a thread’s nonfixed priority less favorable. The system
manager can apply a negative nice value to a thread, thus giving it a better priority.
The following illustration shows some of the ways in which the priority value can change.
Figure 6. How the Priority Value is Determined. The illustration shows how the scheduling priority value of a thread
can change during execution or after applying the nice command. The smaller the priority value, the higher the thread
priority. At initiation, the nice value defaults to 20 and the base priority defaults to 40. After some execution and a
processor penality, the nice value remains 20 and the base priority remains 40. After running the renice —5 command
and with the same processor usage as before, the nice value is now 15 and the base priority remains 40. After issuing
the setpri() subroutine with a value of 50, fixed priority is now 50 and the nice value and processor usage is irrelevant.
The nice value of a thread is set when the thread is created and is constant over the life of the thread,
unless explicitly changed by the user through the renice command or the setpri(), setpriority(),
thread_setsched(), or nice() system calls.
The processor penalty is an integer that is calculated from the recent processor usage of a thread. The
recent processor usage increases by approximately 1 each time the thread is in control of the processor at
the end of a 10 ms clock tick, up to a maximum value of 120. The actual priority penalty per tick increases
with the nice value. Once per second, the recent processor usage values for all threads are recalculated.
The result is the following:
v The priority of a nonfixed-priority thread becomes less favorable as its recent processor usage
increases and vice versa. This implies that, on average, the more time slices a thread has been
allocated recently, the less likely it is that the thread will be allocated the next time slice.
v The priority of a nonfixed-priority thread becomes less favorable as its nice value increases, and vice
versa.
Note: With the use of multiple processor run queues and their load balancing mechanism, nice or renice
values might not have the expected effect on thread priorities because less favored priorities might
have equal or greater run time than favored priorities. Threads requiring the expected effects of
nice or renice should be placed on the global run queue.
You can use the ps command to display the priority value, nice value, and short-term processor-usage
values for a process.
Chapter 3. Resource Management Overview
21
See Controlling Contention for the processor for a more detailed discussion on using the nice and renice
commands.
See Tuning the Thread-Priority-Value Calculation, for the details of the calculation of the processor penalty
and the decay of the recent processor usage values.
The priority mechanism is also used by AIX Workload Manager to enforce processor resource
management. Because threads classified under the Workload Manager have their priorities managed by
the Workload Manager, they might have different priority behavior over threads not classified under the
Workload Manager.
Scheduling Policy for Threads
The following are the possible values for thread scheduling policy:
SCHED_FIFO
After a thread with this policy is scheduled, it runs to completion unless it is blocked, it voluntarily
yields control of the processor, or a higher-priority thread becomes dispatchable. Only fixed-priority
threads can have a SCHED_FIFO scheduling policy.
SCHED_RR
When a SCHED_RR thread has control at the end of the time slice, it moves to the tail of the
queue of dispatchable threads of its priority. Only fixed-priority threads can have a SCHED_RR
scheduling policy.
SCHED_OTHER
This policy is defined by POSIX Standard 1003.4a as implementation-defined. The recalculation of
the running thread’s priority value at each clock interrupt means that a thread may lose control
because its priority value has risen above that of another dispatchable thread.
SCHED_FIFO2
The policy is the same as for SCHED_FIFO, except that it allows a thread which has slept for only
a short amount of time to be put at the head of its run queue when it is awakened. This time
period is the affinity limit (tunable with schedtune -a). This policy is only available beginning with
AIX 4.3.3.
SCHED_FIFO3
A thread whose scheduling policy is set to SCHED_FIFO3 is always put at the head of a run
queue. To prevent a thread belonging to SCHED_FIFO2 scheduling policy from being put ahead of
SCHED_FIFO3, the run queue parameters are changed when a SCHED_FIFO3 thread is
enqueued, so that no thread belonging to SCHED_FIFO2 will satisfy the criterion that enables it to
join the head of the run queue. This policy is only available beginning with AIX 4.3.3.
SCHED_FIFO4
A higher priority SCHED_FIFO4 scheduling class thread does not preempt the currently running
low priority thread as long as their priorities differ by a value of 1. The default behavior is the
preemption of the currently running low priority thread on a given CPU by a high priority thread
that becomes eligible to run on the same CPU. This policy is only available beginning with AIX 5L
Version 5100-01 + APAR IY22854.
The scheduling policies are set with the thread_setsched() system call and are only effective for the
calling thread. However, a thread can be set to the SCHED_RR scheduling policy by issuing a setpri() call
specifying the process ID; the caller of setpri() and the target of setpri() do not have to match.
Only processes that have root authority can issue the setpri() system call. Only threads that have root
authority can change the scheduling policy to any of the SCHED_FIFO options or SCHED_RR. If the
scheduling policy is SCHED_OTHER, the priority parameter is ignored by the thread_setsched()
subroutine.
22
Performance Management Guide
Threads are primarily of interest for applications that currently consist of several asynchronous processes.
These applications might impose a lighter load on the system if converted to a multithreaded structure.
Scheduler Run Queue
The scheduler maintains a run queue of all of the threads that are ready to be dispatched. The following
illustration depicts the run queue symbolically.
Figure 7. Run Queue. This illustration simply shows how threads with a lower priority value are passed through the run
queue before threads with a higher priority value. The range of possible priority values is 0 to 127 which directly relate
to a total of 128 total run queues.
All the dispatchable threads of a given priority occupy positions in the run queue.
The fundamental dispatchable entity of the scheduler is the thread. AIX 5.1 maintains 256 run queues (128
in AIX 4.3 and prior releases). In AIX 5.1, run queues relate directly to the range of possible values (0
through 255) for the priority field for each thread.. This method makes it easier for the scheduler to
determine which thread is most favored to run. Without having to search a single large run queue, the
scheduler consults a mask where a bit is on to indicate the presence of a ready-to-run thread in the
corresponding run queue.
The priority value of a thread changes rapidly and frequently. The constant movement is due to the way
that the scheduler recalculates priorities. This is not true, however, for fixed-priority threads.
Starting with AIX 4.3.3, each processor has its own run queue. The run queue values reported in the
performance tools will be the sum of all the threads in each run queue. Having a per-processor run queue
saves overhead on dispatching locks and improves overall processor affinity. Threads will tend to stay on
the same processor more often. If a thread becomes runnable because of an event on another processor
than the one in which the newly runnable thread had been running on, then this thread would only get
dispatched immediately if there was an idle processor. No preemption occurs until the processor’s state
can be examined (such as an interrupt on this thread’s processor).
On multiprocessor systems with multiple run queues, transient priority inversions can occur. It is possible
at any point in time that one run queue could have several threads having more favorable priority than
another run queue. AIX has mechanisms for priority balancing over time, but if strict priority is required (for
Chapter 3. Resource Management Overview
23
example, for real-time applications) an environment variable called RT_GRQ exists, that, if set to ON, will
cause this thread to be on a global run queue. In that case, the global run queue is searched to see which
thread has the best priority. This can improve performance for threads that are interrupt driven. Threads
that are running at fixed priority are placed on the global run queue if schedtune -F is set to 1.
The average number of threads in the run queue can be seen in the first column of the vmstat command
output. If you divide this number by the number of processors, the result is the average number of threads
that can be run on each processor. If this value is greater than one, these threads must wait their turn for
the processor (the greater the number, the more likely it is that performance delays are noticed).
When a thread is moved to the end of the run queue (for example, when the thread has control at the end
of a time slice), it is moved to a position after the last thread in the queue that has the same priority value.
Scheduler processor Time Slice
The processor time slice is the amount of time a SCHED_RR thread can absorb before the scheduler
switches to another thread at the same priority. You can use the -t option of the schedtune command to
increase the number of clock ticks in the time slice by 10 millisecond increments (see Modifying the
Scheduler Time Slice with the schedtune Command).
Note: The time slice is not a guaranteed amount of processor time. It is the longest time that a thread can
be in control before it faces the possibility of being replaced by another thread. There are many
ways in which a thread can lose control of the processor before it has had control for a full time
slice.
Mode Switching
A user process undergoes a mode switch when it needs access to system resources. This is implemented
through the system call interface or by interrupts such as page faults. There are two modes:
v User mode
v Kernel mode
Processor time spent in user mode (application and shared libraries) is reflected as user time in the output
of commands such as the vmstat, iostat, and sar commands. Processor time spent in kernel mode is
reflected as system time in the output of these commands.
User Mode
Programs that execute in the user protection domain are user processes. Code that executes in this
protection domain executes in user execution mode, and has the following access:
v Read/write access to user data in the process private region
v Read access to the user text and shared text regions
v Access to shared data regions using the shared memory functions
Programs executing in the user protection domain do not have access to the kernel or kernel data
segments, except indirectly through the use of system calls. A program in this protection domain can only
affect its own execution environment and executes in the process or unprivileged state.
Kernel Mode
Programs that execute in the kernel protection domain include interrupt handlers, kernel processes, the
base kernel, and kernel extensions (device driver, system calls and file systems). This protection domain
implies that code executes in kernel execution mode, and has the following access:
v Read/write access to the global kernel address space
v Read/write access to the kernel data in the process region when executing within a process
Kernel services must be used to access user data within the process address space.
24
Performance Management Guide
Programs executing in this protection domain can affect the execution environments of all programs,
because they have the following characteristics:
v They can access global system data
v They can use kernel services
v They are exempt from all security restraints
v They execute in the processor privileged state.
Mode Switches
The use of a system call by a user-mode process allows a kernel function to be called from user mode.
Access to functions that directly or indirectly invoke system calls is typically provided by programming
libraries, which provide access to operating system functions.
Mode switches should be differentiated from the context switches seen in the output of the vmstat (cs
column) and sar (cswch/s) commands. A context switch occurs when the currently running thread is
different from the previously running thread on that processor.
The scheduler performs a context switch when any of the following occurs:
v A thread must wait for a resource (voluntarily), such as disk I/O, network I/O, sleep, or locks
v A higher priority thread wakes up (involuntarily)
v The thread has used up its time slice (usually 10 ms).
Context switch time, system calls, device interrupts, NFS I/O, and any other activity in the kernel is
considered as system time.
Performance Overview of the Virtual Memory Manager (VMM)
The virtual address space is partitioned into segments. A segment is a 256 MB, contiguous portion of the
virtual-memory address space into which a data object can be mapped.
Process addressability to data is managed at the segment (or object) level so that a segment can be
shared between processes or maintained as private. For example, processes can share code segments
yet have separate and private data segments.
Real-Memory Management
Virtual-memory segments are partitioned into fixed-size units called pages. The default page size is 4096
bytes. Some systems also support a larger page size, typically accessed only through the shmat system
call.Each page in a segment can be in real memory (RAM), or stored on disk until it is needed. Similarly,
real memory is divided into 4096-byte page frames. The role of the VMM is to manage the allocation of
real-memory page frames and to resolve references by the program to virtual-memory pages that are not
currently in real memory or do not yet exist (for example, when a process makes the first reference to a
page of its data segment).
Because the amount of virtual memory that is in use at any given instant can be larger than real memory,
the VMM must store the surplus on disk. From the performance standpoint, the VMM has two, somewhat
opposed, objectives:
v Minimize the overall processor-time and disk-bandwidth cost of the use of virtual memory
v Minimize the response-time cost of page faults
In pursuit of these objectives, the VMM maintains a free list of page frames that are available to satisfy a
page fault. The VMM uses a page-replacement algorithm to determine which virtual-memory pages
currently in memory will have their page frames reassigned to the free list. The page-replacement
algorithm uses several mechanisms:
v Virtual-memory segments are classified into either persistent segments or working segments.
Chapter 3. Resource Management Overview
25
v
v
v
v
v
Virtual-memory segments are classified as containing either computational or file memory.
Virtual-memory pages whose access causes a page fault are tracked.
Page faults are classified as new-page faults or as repage faults.
Statistics are maintained on the rate of repage faults in each virtual-memory segment.
User-tunable thresholds influence the page-replacement algorithm’s decisions.
The following sections describe the free list and the page-replacement mechanisms in more detail.
Free List
The VMM maintains a logical list of free page frames that it uses to accommodate page faults. In most
environments, the VMM must occasionally add to the free list by reassigning some page frames owned by
running processes. The virtual-memory pages whose page frames are to be reassigned are selected by
the VMM’s page-replacement algorithm. The VMM thresholds determine the number of frames reassigned.
Persistent versus Working Segments
The pages of a persistent segment have permanent storage locations on disk. Files containing data or
executable programs are mapped to persistent segments. Because each page of a persistent segment has
a permanent disk storage location, the VMM writes the page back to that location when the page has been
changed and can no longer be kept in real memory. If the page has not changed when selected for
placement on a free list, no I/O is required. If the page is referenced again later, a new copy is read in
from its permanent disk-storage location.
Working segments are transitory, exist only during their use by a process, and have no permanent
disk-storage location. Process stack and data regions are mapped to working segments, as are the kernel
text segment, the kernel-extension text segments, as well as the shared-library text and data segments.
Pages of working segments must also have disk-storage locations to occupy when they cannot be kept in
real memory. The disk-paging space is used for this purpose.
The following illustration shows the relationship between some of the types of segments and the locations
of their pages on disk. It also shows the actual (arbitrary) locations of the pages when they are in real
memory.
26
Performance Management Guide
Figure 8. Persistent and Working Storage Segments. This illustration shows the relationship between some of the
types of segments and the locations of their pages on disk. It also shows the actual (arbitrary) locations of the pages
when they are in real memory. Working segments are transitory, meaning they exist only during their use by a process
and have no permanent disk-storage location. Process stack and data regions are mapped to working segments, as
are the kernel text segment, the kernel-extension text segments, and the shared-library text and data segments.
Pages of working segments must also have disk-storage locations to occupy when they cannot be kept in real
memory. The disk-paging space is used for this purpose.
Persistent-segment types are further classified. Client segments are used to map remote files (for
example, files that are being accessed through NFS), including remote executable programs. Pages from
client segments are saved and restored over the network to their permanent file location, not on the
local-disk paging space. Journaled and deferred segments are persistent segments that must be
atomically updated. If a page from a journaled or deferred segment is selected to be removed from real
memory (paged out), it must be written to disk paging space unless it is in a state that allows it to be
committed (written to its permanent file location).
Computational versus File Memory
Computational memory, also known as computational pages, consists of the pages that belong to
working-storage segments or program text (executable files) segments.
File memory (or file pages) consists of the remaining pages. These are usually pages from permanent
data files in persistent storage.
Page Replacement
When the number of available real memory frames on the free list becomes low, a page stealer is invoked.
A page stealer moves through the Page Frame Table (PFT), looking for pages to steal.
The PFT includes flags to signal which pages have been referenced and which have been modified. If the
page stealer encounters a page that has been referenced, it does not steal that page, but instead, resets
the reference flag for that page. The next time the clock hand (page stealer) passes that page and the
reference bit is still off, that page is stolen. A page that was not referenced in the first pass is immediately
stolen.
The modify flag indicates that the data on that page has been changed since it was brought into memory.
When a page is to be stolen, if the modify flag is set, a pageout call is made before stealing the page.
Pages that are part of working segments are written to paging space; persistent segments are written to
disk.
Chapter 3. Resource Management Overview
27
Figure 9. Page Replacement Example. The illustration consists of excerpts from three tables. The first table is the
page frame table with four columns that contain the real address, the segment type, a reference flag, and a modify
flag. A second table is called the free list table and contains addresses of all free pages. The last table represents the
resulting page frame table after all of the free addresses have been removed.
In addition to the page-replacement, the algorithm keeps track of both new page faults (referenced for the
first time) and repage faults (referencing pages that have been paged out), by using a history buffer that
contains the IDs of the most recent page faults. It then tries to balance file (persistent data) page outs with
computational (working storage or program text) page outs.
When a process exits, its working storage is released immediately and its associated memory frames are
put back on the free list. However, any files that the process may have opened can stay in memory.
Page replacement is done directly within the scope of the thread if running on a uniprocessor. On a
multiprocessor system, page replacement is done through the lrud kernel process, which is dispatched to
a CPU when the minfree threshold has been reached. Starting with AIX 4.3.3, the lrud kernel process is
multithreaded with one thread per memory pool. Real memory is split into evenly sized memory pools
based on the number of CPUs and the amount of RAM. The number of memory pools on a system can be
determined by running the vmtune -A command.
In AIX 4.3.3 and later use the vmtune -m <number of memory pools> command to change the number of
memory pools that will be configured at system boot. The values for minfree and maxfree in the vmtune
command output will be the sum of the minfree and maxfree for each memory pool.
Repaging
A page fault is considered to be either a new page fault or a repage fault. A new page fault occurs when
there is no record of the page having been referenced recently. A repage fault occurs when a page that is
28
Performance Management Guide
known to have been referenced recently is referenced again, and is not found in memory because the
page has been replaced (and perhaps written to disk) since it was last accessed.
A perfect page-replacement policy would eliminate repage faults entirely (assuming adequate real memory)
by always stealing frames from pages that are not going to be referenced again. Thus, the number of
repage faults is an inverse measure of the effectiveness of the page-replacement algorithm in keeping
frequently reused pages in memory, thereby reducing overall I/O demand and potentially improving system
performance.
To classify a page fault as new or repage, the VMM maintains a repage history buffer that contains the
page IDs of the N most recent page faults, where N is the number of frames that the memory can hold.
For example, 512 MB memory requires a 128 KB repage history buffer. At page-in, if the page’s ID is
found in the repage history buffer, it is counted as a repage. Also, the VMM estimates the
computational-memory repaging rate and the file-memory repaging rate separately by maintaining counts
of repage faults for each type of memory. The repaging rates are multiplied by 0.9 each time the
page-replacement algorithm runs, so that they reflect recent repaging activity more strongly than historical
repaging activity.
VMM Thresholds
Several numerical thresholds define the objectives of the VMM. When one of these thresholds is
breached, the VMM takes appropriate action to bring the state of memory back within bounds. This section
discusses the thresholds that the system administrator can alter through the vmtune command.
The number of page frames on the free list is controlled by the following parameters:
minfree
Minimum acceptable number of real-memory page frames in the free list. When the size of the
free list falls below this number, the VMM begins stealing pages. It continues stealing pages until
the size of the free list reaches maxfree.
maxfree
Maximum size to which the free list will grow by VMM page-stealing. The size of the free list may
exceed this number as a result of processes terminating and freeing their working-segment pages
or the deletion of files that have pages in memory.
The VMM attempts to keep the size of the free list greater than or equal to minfree. When page faults or
system demands cause the free list size to fall below minfree, the page-replacement algorithm runs. The
size of the free list must be kept above a certain level (the default value of minfree) for several reasons.
For example, the operating system’s sequential-prefetch algorithm requires several frames at a time for
each process that is doing sequential reads. Also, the VMM must avoid deadlocks within the operating
system itself, which could occur if there were not enough space to read in a page that was required to free
a page frame.
The following thresholds are expressed as percentages. They represent the fraction of the total real
memory of the machine that is occupied by file pages (pages of noncomputational segments).
minperm
If the percentage of real memory occupied by file pages falls below this level, the
page-replacement algorithm steals both file and computational pages, regardless of repage rates.
maxperm
If the percentage of real memory occupied by file pages rises above this level, the
page-replacement algorithm steals only file pages.
maxclient
If the percentage of real memory occupied by file pages is above this level, the page-replacement
algorithm steals only client pages.
Chapter 3. Resource Management Overview
29
When the percentage of real memory occupied by file pages is between minperm and maxperm, the
VMM normally steals only file pages, but if the repaging rate for file pages is higher than the repaging rate
for computational pages, computational pages are stolen as well.
The main intent of the page-replacement algorithm is to ensure that computational pages are given fair
treatment. For example, the sequential reading of a long data file into memory should not cause the loss
of program text pages that are likely to be used again soon. The page-replacement algorithm’s use of the
thresholds and repaging rates ensures that both types of pages get treated fairly, with a slight bias in favor
of computational pages.
VMM Memory Load Control Facility
A process requires real-memory pages to execute. When a process references a virtual-memory page that
is on disk, because it either has been paged-out or has never been read, the referenced page must be
paged-in and, on average, one or more pages must be paged out (if replaced pages had been modified),
creating I/O traffic and delaying the progress of the process.
The operating system attempts to steal real memory from pages that are unlikely to be referenced in the
near future, through the page-replacement algorithm. A successful page-replacement algorithm allows the
operating system to keep enough processes active in memory to keep the CPU busy. But at some level of
competition for memory, no pages are good candidates for paging out to disk because they will all be
reused in the near future by the active set of processes. This situation depends on the following:
v Total amount of memory in the system
v The number of processes
v The time-varying memory requirements of each process
v The page-replacement algorithm
When this happens, continuous paging-in and paging-out occurs. This condition is called thrashing.
Thrashing results in incessant I/O to the paging disk and causes each process to encounter a page fault
almost as soon as it is dispatched, with the result that none of the processes make any significant
progress.
The most destructive aspect of thrashing is that, although thrashing may have been triggered by a brief,
random peak in workload (such as all of the users of a system happening to press Enter in the same
second), the system might continue thrashing for an indefinitely long time.
The operating system has a memory load-control algorithm that detects when the system is starting to
thrash and then suspends active processes and delays the initiation of new processes for a period of time.
Five parameters set rates and bounds for the algorithm. The default values of these parameters have
been chosen to be ″fail safe″ across a wide range of workloads. In AIX Version 4, memory load control is
disabled by default on systems that have available memory frames that add up to greater than or equal to
128 MB.
Memory Load Control Algorithm
The memory load control mechanism assesses, once per second, whether sufficient memory is available
for the set of active processes. When a memory-overcommitment condition is detected, some processes
are suspended, decreasing the number of active processes and thereby decreasing the level of memory
overcommitment.
When a process is suspended, all of its threads are suspended when they reach a suspendable state. The
pages of the suspended processes quickly become stale and are paged out by the page-replacement
algorithm, releasing enough page frames to allow the remaining active processes to progress. During the
interval in which existing processes are suspended, newly created processes are also suspended,
preventing new work from entering the system. Suspended processes are not reactivated until a
subsequent interval passes during which no potential thrashing condition exists. Once this safe interval
has passed, the threads of the suspended processes are gradually reactivated.
30
Performance Management Guide
Memory load-control parameters specify the following:
v The system memory overcommitment threshold (schedtune -h)
v The number of seconds required to make a safe interval (schedtune -w)
v The individual process memory overcommitment threshold by which an individual process is qualified as
a suspension candidate (schedtune -p)
v The minimum number of active processes when processes are being suspended (schedtune -m)
v The minimum number of elapsed seconds of activity for a process after reactivation (schedtune -e)
For information on setting and tuning these parameters, see Tuning VMM Memory Load Control with the
schedtune Command.
Once per second, the scheduler (process 0) examines the values of all the above measures that have
been collected over the preceding one-second interval, and determines if processes are to be suspended
or activated. If processes are to be suspended, every process eligible for suspension by the -p and -e
parameter test is marked for suspension. When that process next receives the CPU in user mode, it is
suspended (unless doing so would reduce the number of active processes below the -m value). The
user-mode criterion is applied so that a process is ineligible for suspension during critical system activities
performed on its behalf. If, during subsequent one-second intervals, the thrashing criterion is still being
met, additional process candidates meeting the criteria set by -p and -e are marked for suspension. When
the scheduler subsequently determines that the safe-interval criterion has been met and processes are to
be reactivated, some number of suspended processes are put on the run queue (made active) every
second.
Suspended processes are reactivated by:
1. Priority
2. The order in which they were suspended
The suspended processes are not all reactivated at once. A value for the number of processes reactivated
is selected by a formula that recognizes the number of then-active processes and reactivates either
one-fifth of the number of then-active processes or a monotonically increasing lower bound, whichever is
greater. This cautious strategy results in increasing the degree of multiprogramming roughly 20 percent per
second. The intent of this strategy is to make the rate of reactivation relatively slow during the first second
after the safe interval has expired, while steadily increasing the reintroduction rate in subsequent seconds.
If the memory-overcommitment condition recurs during the course of reactivating processes, the following
occur:
v Reactivation is halted
v The marked-to-be reactivated processes are again marked suspended
v Additional processes are suspended in accordance with the above rules
Allocation and Reclamation of Paging Space Slots
The operating system supports three allocation methods for working storage, also referred to as
paging-space slots, as follows:
v Late allocation
v Early allocation
v Deferred allocation
Note: Paging-space slots are only released by process (not thread) termination or by the disclaim()
system call. The slots are not released by the free() system call.
Chapter 3. Resource Management Overview
31
Late Allocation Algorithm
Prior to AIX 4.3.2 with the late allocation algorithm, a paging slot is allocated to a page of virtual memory
only when that page is first touched. That is the first time that the page’s content is of interest to the
executing program.
Many programs exploit late allocation by allocating virtual-memory address ranges for maximum-sized
structures and then only using as much of the structure as the situation requires. The pages of the
virtual-memory address range that are never accessed never require real-memory frames or paging-space
slots.
This technique does involve some degree of risk. If all of the programs running in a machine happened to
encounter maximum-size situations simultaneously, paging space might be exhausted. Some programs
might not be able to continue to completion.
Early Allocation Algorithm
The second operating system’s paging-space-slot-allocation method is intended for use in installations
where this situation is likely, or where the cost of failure to complete is intolerably high. Aptly called early
allocation, this algorithm causes the appropriate number of paging-space slots to be allocated at the time
the virtual-memory address range is allocated, for example, with the malloc() subroutine. If there are not
enough paging-space slots to support the malloc() subroutine, an error code is set. The early-allocation
algorithm is invoked as follows:
# export PSALLOC=early
This example causes all future programs to be executed in the environment to use early allocation. The
currently executing shell is not affected.
Early allocation is of interest to the performance analyst mainly because of its paging-space size
implications. If early allocation is turned on for those programs, paging-space requirements can increase
many times. Whereas the normal recommendation for paging-space size is at least twice the size of the
system’s real memory, the recommendation for systems that use PSALLOC=early is at least four times
the real memory size. Actually, this is just a starting point. Analyze the virtual storage requirements of your
workload and allocate paging spaces to accommodate them. As an example, at one time, the AIXwindows
server required 250 MB of paging space when run with early allocation.
When using PSALLOC=early, the user should set a handler for the following SIGSEGV signal by
pre-allocating and setting the memory as a stack using the sigaltstack function. Even though
PSALLOC=early is specified, when there is not enough paging space and a program attempts to expand
the stack, the program may receive the SIGSEGV signal.
Deferred Allocation Algorithm
The third operating system’s paging-space-slot-allocation method is the default beginning with AIX 4.3.2
Deferred Page Space Allocation (DPSA) policy delays allocation of paging space until it is necessary to
page out the page, which results in no wasted paging space allocation. This method can save huge
amounts of paging space, which means disk space.
On some systems, paging space might not ever be needed even if all the pages accessed have been
touched. This situation is most common on systems with very large amount of RAM. However, this may
result in overcommitment of paging space in cases where more virtual memory than available RAM is
accessed.
To disable DPSA and preserve the Late Page Space Allocation policy, run the following command:
# /usr/samples/kernel/vmtune -d 0
To activate DPSA, run the following command:
# /usr/samples/kernel/vmtune -d 1
32
Performance Management Guide
In general, system performance can be improved by DPSA, because the overhead of allocating page
space after page faults is avoided the. Paging space devices need less disk space if DPSA is used.
For further information, see Choosing a Page Space Allocation Method and Placement and Sizes of
Paging Spaces.
Performance Overview of Fixed-Disk Storage Management
The following illustration shows the hierarchy of structures used by the operating system to manage
fixed-disk storage. Each individual disk drive, called a physical volume (PV), has a name, such as
/dev/hdisk0. If the physical volume is in use, it belongs to a volume group (VG). All of the physical
volumes in a volume group are divided into physical partitions (PPs) of the same size (by default, 4 MB in
volume groups that include physical volumes smaller than 4 GB; 8 MB or more with bigger disks).
For space-allocation purposes, each physical volume is divided into five regions. See Position on Physical
Volume for more information. The number of physical partitions in each region varies, depending on the
total capacity of the disk drive.
Chapter 3. Resource Management Overview
33
Figure 10. Organization of Fixed-Disk Data (Unmirrored). The illustration shows the hierarchy of a physical volume that
is partitioned into one or more logical volumes. These partitions or logical volumes contain file systems with directory
structures which contain individual files. Files are written to blocks contained in tracks on the storage media and these
blocks are usually not contiguous. Disk fragmenting occurs when data gets erased and new data files are written to
the empty blocks that are randomly scattered around multiple tracks on the media.
Within each volume group, one or more logical volumes (LVs) are defined. Each logical volume consists of
one or more logical partitions. Each logical partition corresponds to at least one physical partition. If
mirroring is specified for the logical volume, additional physical partitions are allocated to store the
additional copies of each logical partition. Although the logical partitions are numbered consecutively, the
underlying physical partitions are not necessarily consecutive or contiguous.
Logical volumes can serve a number of system purposes, such as paging, but each logical volume that
holds ordinary system data or user data or programs contains a single journaled file system (JFS or
Enhanced JFS). Each JFS consists of a pool of page-size (4096-byte) blocks. When data is to be written
to a file, one or more additional blocks are allocated to that file. These blocks may or may not be
contiguous with one another and with other blocks previously allocated to the file.
34
Performance Management Guide
For purposes of illustration, the previous figure shows a bad (but not the worst possible) situation that
might arise in a file system that had been in use for a long period without reorganization. The
/op/filename file is physically recorded on a large number of blocks that are physically distant from one
another. Reading the file sequentially would result in many time-consuming seek operations.
While an operating system’s file is conceptually a sequential and contiguous string of bytes, the physical
reality might be very different. Fragmentation may arise from multiple extensions to logical volumes as well
as allocation/release/reallocation activity within a file system. A file system is fragmented when its available
space consists of large numbers of small chunks of space, making it impossible to write out a new file in
contiguous blocks.
Access to files in a highly fragmented file system may result in a large number of seeks and longer I/O
response times (seek latency dominates I/O response time). For example, if the file is accessed
sequentially, a file placement that consists of many, widely separated chunks requires more seeks than a
placement that consists of one or a few large contiguous chunks. If the file is accessed randomly, a
placement that is widely dispersed requires longer seeks than a placement in which the file’s blocks are
close together.
The effect of a file’s placement on I/O performance diminishes when the file is buffered in memory. When
a file is opened in the operating system, it is mapped to a persistent data segment in virtual memory. The
segment represents a virtual buffer for the file; the file’s blocks map directly to segment pages. The VMM
manages the segment pages, reading file blocks into segment pages upon demand (as they are
accessed). There are several circumstances that cause the VMM to write a page back to its corresponding
block in the file on disk; but, in general, the VMM keeps a page in memory if it has been accessed
recently. Thus, frequently accessed pages tend to stay in memory longer, and logical file accesses to the
corresponding blocks can be satisfied without physical disk accesses.
At some point, the user or system administrator can choose to reorganize the placement of files within
logical volumes and the placement of logical volumes within physical volumes to reduce fragmentation and
to more evenly distribute the total I/O load. Chapter 8. Monitoring and Tuning Disk I/O Use contains further
details about detecting and correcting disk placement and fragmentation problems.
Sequential-Access Read Ahead
The VMM tries to anticipate the future need for pages of a sequential file by observing the pattern in which
a program is accessing the file. When the program accesses two successive pages of the file, the VMM
assumes that the program will continue to access the file sequentially, and the VMM schedules additional
sequential reads of the file. These reads are overlapped with the program processing, and will make the
data available to the program sooner than if the VMM had waited for the program to access the next page
before initiating the I/O. The number of pages to be read ahead is determined by two VMM thresholds:
minpgahead
Number of pages read ahead when the VMM first detects the sequential access pattern. If the
program continues to access the file sequentially, the next read ahead will be for 2 times
minpgahead, the next for 4 times minpgahead, and so on until the number of pages reaches
maxpgahead.
maxpgahead
Maximum number of pages the VMM will read ahead in a sequential file.
The number of pages to read ahead on Enhanced JFS is determined by the two thresholds:
j2_minPageReadAhead
and
j2_maxPageReadAhead.
Chapter 3. Resource Management Overview
35
Write Behind
To increase write performance, limit the number of dirty file pages in memory, reduce system overhead,
and minimize disk fragmentation, the file system divides each file into 16 KB partitions. The pages of a
given partition are not written to disk until the program writes the first byte of the next 16 KB partition. At
that point, the file system forces the four dirty pages of the first partition to be written to disk. The pages of
data remain in memory until their frames are reused, at which point no additional I/O is required. If a
program accesses any of the pages before their frames are reused, no I/O is required.
If a large number of dirty file pages remain in memory and do not get reused, the sync daemon writes
them to disk, which might result in abnormal disk utilization. To distribute the I/O activity more efficiently
across the workload, write-behind can be turned on to tell the system how many pages to keep in memory
before writing them to disk. The write-behind threshold is on a per-file basis, which causes pages to be
written to disk before the sync daemon runs. The I/O is spread more evenly throughout the workload.
There are two types of write-behind: sequential and random. The size of the write-behind partitions and
the write-behind threshold can be changed with the vmtune command (see VMM Write-Behind).
Memory Mapped Files and Write Behind
Normal files are automatically mapped to segments to provide mapped files. This means that normal file
access bypasses traditional kernel buffers and block I/O routines, allowing files to use more memory when
the extra memory is available (file caching is not limited to the declared kernel buffer area).
Files can be mapped explicitly with the shmat() or mmap() subroutines, but this provides no additional
memory space for their caching. Applications that use the shmat() or mmap() subroutines to map a file
explicitly and access it by address rather than by the read() and write() subroutines may avoid some path
length of the system-call overhead, but they lose the benefit of the system write-behind feature.
When applications do not use the write() subroutine, modified pages tend to accumulate in memory and
be written randomly when purged by the VMM page-replacement algorithm or the sync daemon. This
results in many small writes to the disk that cause inefficiencies in CPU and disk utilization, as well as
fragmentation that might slow future reads of the file.
Disk-I/O Pacing
Because most writes are asynchronous, FIFO I/O queues of several megabytes can build up, which can
take several seconds to complete. The performance of an interactive process is severely impacted if every
disk read spends several seconds working its way through the queue. In response to this problem, the
VMM has an option called I/O pacing to control writes.
I/O pacing does not change the interface or processing logic of I/O. It simply limits the number of I/Os that
can be outstanding against a file. When a process tries to exceed that limit, it is suspended until enough
outstanding requests have been processed to reach a lower threshold. Using Disk-I/O Pacing describes
I/O pacing in more detail.
Support for Pinned Memory
AIX 4.3.3 and AIX 5.1 enable memory pages to be maintained in real memory all the time. This
mechanism is called pinning memory. Pinning a memory region prohibits the pager from stealing pages
from the pages backing the pinned memory region. Memory regions defined in either system space or
user space may be pinned. After a memory region is pinned, accessing that region does not result in a
page fault until the region is subsequently unpinned. While a portion of the kernel remains pinned, many
regions are pageable and are only pinned while being accessed.
36
Performance Management Guide
The advantage of having portions of memory pinned is that, when accessing a page that is pinned, you
can retrieve the page without going through the page replacement algorithm. An adverse side effect of
having too many pinned memory pages is that it can increase paging activity for unpinned pages, which
would degrade performance.
To tune pinned memory, use the vmtune command to dedicate a number of pages at boot time for pinned
memory. The following flags affect how AIX manages pinned memory:
maxpin
Specifies the maximum percentage of real memory that can be pinned.
v_pinshm
Setting the v_pinshm parameter to 1 (-S 1) causes pages in shared memory segments to be
pinned by VMM, if the application, which does the shmget(), specifies SHM_PIN as part of the
flags. The default value is 0. This option is available only in AIX 4.3.3 and later.
Large Page Support
In addition to regular page sizes of 4 kilobytes, beginning with AIX 5.1, the operating system supports
large, 16–MB pages. Applications can use large pages with the shmget and shmat system calls. For the
system to be able to use large pages, the pages must be enabled by specifying the SHM_LGPAGE flag
with the shmget system call. Use this flag in conjunction with the SHM_PIN flag, and enable with the
vmtune command.
To enable support for large pages, use the following flags with the vmtune command:
Table 1.
-gLargePageSize
Specifies the size in bytes of the hardware-supported large pages used for the
implementation for the shmget system call with the SHM_LGPAGE flag. Large
pages must be enabled with a non-zero value for the -L flag and the bosboot
command must be run and the system restarted for this change to take effect.
-LLargePages
Specifies the number of large pages to reserve for implementing the shmget system
call with the SHM_LGPAGE flag. For this change to take effect, you must specify the
-g flag, run the bosboot command, and restart the system.
Use the following flags with the shmget system call:
SHM_LGPAGE
Creates the region so it can be mapped through hardware-supported, large-page mechanisms, if
enabled. This flag must be used in conjunction with the SHM_PIN flag and enabled with the
vmtune -L command, to reserve memory for the region (which requires a restart) and vmtune -S
to enable SHM_PIN. This has no effect on shared memory regions created with the EXTSHM=ON
environment variable.
SHM_PIN
Pins the shared memory region if enabled. This flag must be enabled with the vmtune command.
This has no effect on shared memory regions created with EXTSHM=ON environment variable.
Chapter 3. Resource Management Overview
37
38
Performance Management Guide
Chapter 4. Introduction to Multiprocessing
At any given time, a technological limit exists on the speed with which a single processor chip can
operate. If a system’s workload cannot be handled satisfactorily by a single processor, one response is to
apply multiple processors to the problem.
The success of this response depends not only on the skill of the system designers, but also on whether
the workload is amenable to multiprocessing. In terms of human tasks, adding people might be a good
idea if the task is answering calls to a toll-free number, but is dubious if the task is driving a car.
If improved performance is the objective of a proposed migration from a uniprocessor to a multiprocessor
system, the following conditions must be true:
v The workload is processor-limited and has saturated its uniprocessor system.
v The workload contains multiple processor-intensive elements, such as transactions or complex
calculations, that can be performed simultaneously and independently.
v The existing uniprocessor cannot be upgraded or replaced with another uniprocessor of adequate
power.
Although unchanged single-thread applications normally function correctly in a multiprocessor environment,
their performance often changes in unexpected ways. Migration to a multiprocessor can improve the
throughput of a system, and can improve the execution time of complex, multithreaded applications, but
seldom improves the response time of individual, single-thread commands.
Getting the best possible performance from a multiprocessor system requires an understanding of the
operating-system and hardware-execution dynamics that are unique to the multiprocessor environment.
This chapter includes the following major sections:
v Symmetrical Multiprocessor (SMP) Concepts and Architecture
v SMP Performance Issues
v SMP Workloads
v SMP Thread Scheduling
v Thread Tuning
v SMP Tools
Symmetrical Multiprocessor (SMP) Concepts and Architecture
As with any change that increases the complexity of the system, the use of multiple processors generates
design considerations that must be addressed for satisfactory operation and performance. The additional
complexity gives more scope for hardware/software tradeoffs and requires closer hardware/software
design coordination than in uniprocessor systems. The different combinations of design responses and
tradeoffs give rise to a wide variety of multiprocessor system architectures.
This section describes the main design considerations of multiprocessor systems and the hardware
responses to those considerations.
Types of Multiprocessing
Several categories of multiprocessing (MP) systems exist, as described below:
Shared Nothing MP (pure cluster)
Each processor is a complete stand-alone machine and runs a copy of the operating system. The
processors share nothing (each has its own memory, caches, and disks), but they are interconnected.
© Copyright IBM Corp. 1997, 2002
39
When LAN-connected, processors are loosely coupled. When connected by a switch, the processors are
tightly coupled. Communication between processors is done through message-passing.
The advantages of such a system are very good scalability and high availability. The disadvantages of
such a system are an unfamiliar programming model (message passing).
Shared Disks MP
Processors have their own memory and cache. The processors run in parallel and share disks. Each
processor runs a copy of the operating system and the processors are loosely coupled (connected through
LAN). Communication between processors is done through message-passing.
The advantages of shared disks are that part of a familiar programming model is retained (disk data is
addressable and coherent, memory is not), and high availability is much easier than with shared-memory
systems. The disadvantages are limited scalability due to bottlenecks in physical and logical access to
shared data.
Shared Memory Cluster (SMC)
All of the processors in a shared memory cluster have their own resources (main memory, disks, I/O) and
each processor runs a copy of the operating system. Processors are tightly coupled (connected through a
switch). Communication between the processors is done through shared memory.
Shared Memory MP
All of the processors are tightly coupled inside the same box with a high-speed bus or a switch. The
processors share the same global memory, disks, and I/O devices. Only one copy of the operating system
runs across all of the processors, and the operating system must be designed to exploit this architecture
(multithreaded operating system).
SMPs have several advantages:
v They are a cost-effective way to increase throughput.
v They offer a single system image since the Operating System is shared between all the processors
(administration is easy).
v They apply multiple processors to a single problem (parallel programming).
v Load balancing is done by the operating system.
v The uniprocessor (UP) programming model can be used in an SMP.
v They are scalable for shared data.
v All data is addressable by all the processors and kept coherent by the hardware snooping logic.
v There is no need to use message-passing libraries to communicate between processors because
communication is done through the global shared memory.
v More power requirements can be solved by adding more processors to the system. However, you must
set realistic expectations about the increase in performance when adding more processors to an SMP
system.
v More and more applications and tools are available today. Most UP applications can run on or are
ported to SMP architecture.
There are some limitations of SMP systems, as follows:
v There are limits on scalability due to cache coherency, locking mechanism, shared objects, and others.
v There is a need for new skills to exploit multiprocessors, such as threads programming and device
drivers programming.
Parallelizing an Application
An application can be parallelized on an SMP in two ways, as follows:
v The traditional way is to break the application into multiple processes. These processes communicate
using inter-process communication (IPC) such as pipes, semaphores or shared memory. The processes
40
Performance Management Guide
must be able to block waiting for events such as messages from other processes, and they must
coordinate access to shared objects with something like locks.
v Another way is to use the portable operating system interface for UNIX (POSIX) threads. Threads have
similar coordination problems as processes and similar mechanisms to deal with them. Thus a single
process can have any number of its threads running simultaneously on different processors.
Coordinating them and serializing access to shared data are the developer’s responsibility.
Consider the advantages of both threads and processes when you are determining which method to use
for parallelizing an application. Threads may be faster than processes and memory sharing is easier. On
another hand, a process implementation will distribute more easily to multiple machines or clusters. If an
application needs to create or delete new instances, then threads are faster (more overhead in forking
processes). For other functions, the overhead of threads is about the same as that of processes.
Data Serialization
Any storage element that can be read or written by more than one thread may change while the program
is running. This is generally true of multiprogramming environments as well as multiprocessing
environments, but the advent of multiprocessors adds to the scope and importance of this consideration in
two ways:
v Multiprocessors and thread support make it attractive and easier to write applications that share data
among threads.
v The kernel can no longer solve the serialization problem simply by disabling interrupts.
Note: To avoid serious problems, programs that share data must arrange to access that data serially,
rather than in parallel. Before a program updates a shared data item, it must ensure that no other
program (including another copy of itself running on another thread) will change the item. Reads
can usually be done in parallel.
The primary mechanism that is used to keep programs from interfering with one another is the lock. A lock
is an abstraction that represents permission to access one or more data items. Lock and unlock requests
are atomic; that is, they are implemented in such a way that neither interrupts nor multiprocessor access
affect the outcome. All programs that access a shared data item must obtain the lock that corresponds to
that data item before manipulating it. If the lock is already held by another program (or another thread
running the same program), the requesting program must defer its access until the lock becomes
available.
Besides the time spent waiting for the lock, serialization adds to the number of times a thread becomes
nondispatchable. While the thread is nondispatchable, other threads are probably causing the
nondispatchable thread’s cache lines to be replaced, which results in increased memory-latency costs
when the thread finally gets the lock and is dispatched.
The operating system’s kernel contains many shared data items, so it must perform serialization internally.
Serialization delays can therefore occur even in an application program that does not share data with other
programs, because the kernel services used by the program have to serialize shared kernel data.
Types of Locks
The Open Software Foundation/1 (OSF/1) 1.1 locking methodology was used as a model for the AIX
multiprocessor lock functions. However, because the system is preemptable and pageable, some
characteristics have been added to the OSF/1 1.1 Locking Model. Simple locks and complex locks are
preemptable. Also, a thread may sleep when trying to acquire a busy simple lock if the owner of the lock is
not currently running. In addition, a simple lock becomes a sleep lock when a processor has been spinning
on a simple lock for a certain amount of time (this amount of time is a systemwide variable).
Chapter 4. Introduction to Multiprocessing
41
AIX Version 4 Simple Locks
A simple lock in operating system version 4 is a spin lock that will sleep under certain conditions
preventing a thread from spinning indefinitely. Simple locks are preemptable, meaning that a kernel thread
can be preempted by another higher priority kernel thread while it holds a simple lock. On a multiprocessor
system, simple locks, which protect thread-interrupt critical sections, must be used in conjunction with
interrupt control in order to serialize execution both within the executing processor and between different
processors.
On a uniprocessor system, interrupt control is sufficient; there is no need to use locks. Simple locks are
intended to protect thread-thread and thread-interrupt critical sections. Simple locks will spin until the lock
becomes available if in an interrupt handler. They have two states: locked or unlocked.
AIX Version 4 Complex Locks
The complex locks in AIX are read-write locks which protect thread-thread critical sections. These locks
are preemptable. Complex locks are spin locks that will sleep under certain conditions. By default, they are
not recursive, but can become recursive through the lock_set_recursive() kernel service. They have three
states: exclusive-write, shared-read, or unlocked.
Lock Granularity
A programmer working in a multiprocessor environment must decide how many separate locks must be
created for shared data. If there is a single lock to serialize the entire set of shared data items, lock
contention is comparatively likely. The existence of widely used locks places an upper limit on the
throughput of the system.
If each distinct data item has its own lock, the probability of two threads contending for that lock is
comparatively low. Each additional lock and unlock call costs processor time, however, and the existence
of multiple locks makes a deadlock possible. At its simplest, deadlock is the situation shown in the
following illustration, in which Thread 1 owns Lock A and is waiting for Lock B. Meanwhile, Thread 2 owns
Lock B and is waiting for Lock A. Neither program will ever reach the unlock() call that would break the
deadlock. The usual preventive for deadlock is to establish a protocol by which all of the programs that
use a given set of locks must always acquire them in exactly the same sequence.
Figure 11. Deadlock. Shown in the following illustration is a deadlock in which a column named Thread 1 owns Lock A
and is waiting for Lock B. Meanwhile, the column named Thread 2 owns Lock B and is waiting for Lock A. Neither
program thread will ever reach the unlock call that would break the deadlock.
According to queuing theory, the less idle a resource, the longer the average wait to get it. The
relationship is nonlinear; if the lock is doubled, the average wait time for that lock more than doubles.
42
Performance Management Guide
The most effective way to reduce wait time for a lock is to reduce the size of what the lock is protecting.
Here are some guidelines:
v Reduce the frequency with which any lock is requested.
v Lock just the code that accesses shared data, not all the code in a component (this will reduce lock
holding time).
v Lock only specific data items or structures and not entire routines.
v Always associate locks with specific data items or structures, not with routines.
v For large data structures, choose one lock for each element of the structure rather than one lock for the
whole structure.
v Never perform synchronous I/O or any other blocking activity while holding a lock.
v If you have more than one access to the same data in your component, try to move them together so
they can be covered by one lock-unlock action.
v Avoid double wake-up. If you modify some data under a lock and have to notify someone that you have
done it, release the lock before you post the wake-up.
v If you must hold two locks simultaneously, request the busiest one last.
On the other hand, a too-fine granularity will increase the frequency of locks requests and locks releases,
which therefore will add additional instructions. You must locate a balance between a too-fine and
too-coarse granularity. The optimum granularity will have to be found by trial and error, and is one of the
big challenges in an MP system. The following graph shows the relation between the throughput and the
granularity of locks.
Figure 12. Relationship Between Throughput and Granularity. This illustration is a simple two axis chart. The vertical,
or y axis, represents throughput. The horizontal, or x axis, represents granularity going from fine to coarse as it moves
out on the scale. An elongated bell curve shows the relationship of granularity on throughput. As granularity goes from
fine to coarse, throughput gradually increases to a maximum level and then slowly starts to decline. It shows that a
compromise in granularity is necessary to reach maximum throughput.
Locking Overhead
Requesting locks, waiting for locks, and releasing locks add processing overhead in several ways:
v A program that supports multiprocessing always does the same lock and unlock processing, even
though it is running in a uniprocessor or is the only user in a multiprocessor system of the locks in
question.
Chapter 4. Introduction to Multiprocessing
43
v When one thread requests a lock held by another thread, the requesting thread may spin for a while or
be put to sleep and, if possible, another thread dispatched. This consumes processor time.
v The existence of widely used locks places an upper bound on the throughput of the system. For
example, if a given program spends 20 percent of its execution time holding a mutual-exclusion lock, at
most five instances of that program can run simultaneously, regardless of the number of processors in
the system. In fact, even five instances would probably never be so nicely synchronized as to avoid
waiting for one another (see Multiprocessor Throughput Scalability).
Waiting for Locks
When a thread wants a lock already owned by another thread, the thread is blocked and must wait until
the lock becomes free. There are two different ways of waiting:
v Spin locks are suitable for locks that are held only for very short times. It allows the waiting thread to
keep its processor, repeatedly checking the lock bit in a tight loop (spin) until the lock becomes free.
Spinning results in increased CPU time (system time for kernel or kernel extension locks).
v Sleeping locks are suitable for locks that may be held for longer periods. The thread sleeps until the
lock is free and is put back in the run queue when the lock becomes free. Sleeping results in more idle
time.
Waiting always decreases system performance. If a spin lock is used, the processor is busy, but it is not
doing useful work (not contributing to throughput). If a sleeping lock is used, the overhead of context
switching and dispatching as well as the consequent increase in cache misses is incurred.
Operating system developers can choose between two types of locks: mutually exclusive simple locks that
allow the process to spin and sleep while waiting for the lock to become available, and complex read-write
locks that can spin and block the process while waiting for the lock to become available.
Conventions govern the rules about using locks. Neither hardware nor software has an enforcement or
checking mechanism. Although using locks has made the AIX Version 4 ″MP Safe,″ developers are
responsible to define and implement an appropriate locking strategy to protect their own global data.
Cache Coherency
In designing a multiprocessor, engineers give considerable attention to ensuring cache coherency. They
succeed; but cache coherency has a performance cost. We need to understand the problem being
attacked:
If each processor has a cache that reflects the state of various parts of memory, it is possible that two or
more caches may have copies of the same line. It is also possible that a given line may contain more than
one lockable data item. If two threads make appropriately serialized changes to those data items, the
result could be that both caches end up with different, incorrect versions of the line of memory. In other
words, the system’s state is no longer coherent because the system contains two different versions of
what is supposed to be the content of a specific area of memory.
The solutions to the cache coherency problem usually include invalidating all but one of the duplicate lines
when the line is modified. Although the hardware uses snooping logic to invalidate, without any software
intervention, any processor whose cache line has been invalidated will have a cache miss, with its
attendant delay, the next time that line is addressed.
Snooping is the logic used to resolve the problem of cache consistency. Snooping logic in the processor
broadcasts a message over the bus each time a word in its cache has been modified. The snooping logic
also snoops on the bus looking for such messages from other processors.
When a processor detects that another processor has changed a value at an address existing in its own
cache, the snooping logic invalidates that entry in its cache. This is called cross invalidate. Cross
invalidate reminds the processor that the value in the cache is not valid, and it must look for the correct
44
Performance Management Guide
value somewhere else (memory or other cache). Since cross invalidates increase cache misses and the
snooping protocol adds to the bus traffic, solving the cache consistency problem reduces the performance
and scalability of all SMPs.
Processor Affinity and Binding
If a thread is interrupted and later redispatched to the same processor, the processor’s cache might still
contain lines that belong to the thread. If the thread is dispatched to a different processor, it will probably
experience a series of cache misses until its cache working set has been retrieved from RAM or the other
processor’s cache. On the other hand, if a dispatchable thread has to wait until the processor that it was
previously running on is available, the thread may experience an even longer delay.
Processor affinity is the probability of dispatching of a thread to the processor that was previously
executing it. The degree of emphasis on processor affinity should vary directly with the size of the thread’s
cache working set and inversely with the length of time since it was last dispatched. The AIX Version 4
dispatcher enforces affinity with the processors, so affinity is done implicitly by the operating system.
The highest possible degree of processor affinity is to bind a thread to a specific processor. Binding means
that the thread will be dispatched to that processor only, regardless of the availability of other processors.
The bindprocessor command and the bindprocessor() subroutine bind the thread (or threads) of a
specified process to a particular processor (see The bindprocessor Command). Explicit binding is inherited
through fork() and exec() system calls.
The binding can be useful for CPU-intensive programs that experience few interrupts. It can sometimes be
counterproductive for ordinary programs, because it may delay the redispatch of a thread after an I/O until
the processor to which the thread is bound becomes available. If the thread has been blocked for the
duration of an I/O operation, it is unlikely that much of its processing context remains in the caches of the
processor to which it is bound. The thread would probably be better served if it were dispatched to the
next available processor.
Memory and Bus Contention
In a uniprocessor, contention for some internal resources, such as banks of memory and I/O or memory
buses, is usually a minor component using time. In a multiprocessor, these effects can become more
significant, particularly if cache-coherency algorithms add to the number of accesses to RAM.
SMP Performance Issues
To effectively use an SMP, take the following into account when you are attempting to enhance
performance:
Workload Concurrency
The primary performance issue that is unique to SMP systems is workload concurrency, which can be
expressed as, ″Now that we have n processors, how do we keep them all usefully employed″? If only one
processor in a four-way multiprocessor system is doing useful work at any given time, it is no better than a
uniprocessor. It could possibly be worse, because of the extra code to avoid interprocessor interference.
Workload concurrency is the complement of serialization. To the extent that the system software or the
application workload (or the interaction of the two) require serialization, workload concurrency suffers.
Workload concurrency may also be decreased, more desirably, by increased processor affinity. The
improved cache efficiency gained from processor affinity may result in quicker completion of the program.
Workload concurrency is reduced (unless there are more dispatchable threads available), but response
time is improved.
Chapter 4. Introduction to Multiprocessing
45
A component of workload concurrency, process concurrency, is the degree to which a multithreaded
process has multiple dispatchable threads at all times.
Throughput
The throughput of an SMP system is mainly dependent on:
v A consistently high level of workload concurrency. More dispatchable threads than processors at certain
times cannot compensate for idle processors at other times.
v The amount of lock contention.
v The degree of processor affinity.
Response Time
The response time of a particular program in an SMP system is dependent on:
v The process-concurrency level of the program. If the program consistently has two or more dispatchable
threads, its response time will probably improve in an SMP environment. If the program consists of a
single thread, its response time will be, at best, comparable to that in a uniprocessor of the same
speed.
v The amount of lock contention of other instances of the program or with other programs that use the
same locks.
v The degree of processor affinity of the program. If each dispatch of the program is to a different
processor that has none of the program’s cache lines, the program may run more slowly than in a
comparable uniprocessor.
SMP Workloads
The effect of additional processors on performance is dominated by certain characteristics of the specific
workload being handled. This section discusses those critical characteristics and their effects.
The following terms are used to describe the extent to which an existing program has been modified, or a
new program designed, to operate in an SMP environment:
SMP safe
Avoidance in a program of any action, such as unserialized access to shared data, that would
cause functional problems in an SMP environment. This term, when used alone, usually refers to a
program that has undergone only the minimum changes necessary for correct functioning in an
SMP environment.
SMP efficient
Avoidance in a program of any action that would cause functional or performance problems in an
SMP environment. A program that is described as SMP-efficient is SMP-safe as well. An
SMP-efficient program has usually undergone additional changes to minimize incipient bottlenecks.
SMP exploiting
Adding features to a program that are specifically intended to make effective use of an SMP
environment, such as multithreading. A program that is described as SMP-exploiting is generally
assumed to be SMP-safe and SMP-efficient as well.
Workload Multiprocessing
Multiprogramming operating systems running heavy workloads on fast computers give our human senses
the impression that several things are happening simultaneously. In fact, many demanding workloads do
not have large numbers of dispatchable threads at any given instant, even when running on a
single-processor system where serialization is less of a problem. Unless there are always at least as many
dispatchable threads as there are processors, one or more processors will be idle part of the time.
The number of dispatchable threads is the total number of threads in the system
46
Performance Management Guide
v
v
v
v
Minus
Minus
Minus
Minus
the
the
the
the
number
number
number
number
of
of
of
of
threads
threads
threads
threads
that
that
that
that
are
are
are
are
waiting for I/O,
waiting for a shared resource,
waiting for the results of another thread,
sleeping at their own request.
A workload can be said to be multiprocessable to the extent that it presents at all times as many
dispatchable threads as there are processors in the system. Note that this does not mean simply an
average number of dispatchable threads equal to the processor count. If the number of dispatchable
threads is zero half the time and twice the processor count the rest of the time, the average number of
dispatchable threads will equal the processor count, but any given processor in the system will be working
only half the time.
Increasing the multiprocessability of a workload involves one or both of the following:
v Identifying and resolving any bottlenecks that cause threads to wait
v Increasing the total number of threads in the system
These solutions are not independent. If there is a single, major system bottleneck, increasing the number
of threads of the existing workload that pass through the bottleneck will simply increase the proportion of
threads waiting. If there is not currently a bottleneck, increasing the number of threads may create one.
Multiprocessor Throughput Scalability
Real workloads do not scale perfectly on an SMP system. Some factors that inhibit perfect scaling are as
follows:
v
v
v
v
v
Bus/switch contention increases while the number of processors increases
Memory contention increases (all the memory is shared by all the processors)
Increased cost of cache misses as memory gets farther away
Cache cross-invalidates and reads from another cache to maintain cache coherency
Increased cache misses because of higher dispatching rates (more processes/threads need to be
dispatched on the system)
v Increased cost of synchronization instructions
v Increased cache misses because of larger operating system and application data structures
v Increased operating system and application path lengths for lock-unlock
v Increased operating system and application path lengths waiting for locks
All of these factors contribute to what is called the scalability of a workload. Scalability is the degree to
which workload throughput benefits from the availability of additional processors. It is usually expressed as
the quotient of the throughput of the workload on a multiprocessor divided by the throughput on a
comparable uniprocessor. For example, if a uniprocessor achieved 20 requests per second on a given
workload and a four-processor system achieved 58 requests per second, the scaling factor would be 2.9.
That workload is highly scalable. A workload that consisted exclusively of long-running, compute-intensive
programs with negligible I/O or other kernel activity and no shared data might approach a scaling factor of
3.2 to 3.9 on a 4-way system. However, most real-world workloads would not. Because scalability is very
difficult to estimate, scalability assumptions should be based on measurements of authentic workloads.
The following figure illustrates the problems of scaling. The workload consists of a series of hypothetical
commands. Each command is about one-third normal processing, one-third I/O wait, and one-third
processing with a lock held. On the uniprocessor, only one command can actually be processing at a time,
regardless of whether the lock is held. In the time interval shown (five times the standalone execution time
of the command), the uniprocessor handles 7.67 of the commands.
Chapter 4. Introduction to Multiprocessing
47
Figure 13. Multiprocessor Scaling. This figure illustrates the problems of scaling. The workload consists of a series of
hypothetical commands. Each command is about one-third normal processing, one-third I/O wait, and one-third
processing with a lock held. On the uniprocessor, only one command can actually be processing at a time, regardless
of whether the lock is held. In the same time interval, the uniprocessor handles 7.67 of the commands. In that same
period, the multiprocessor handles 14 commands for a scaling factor of 1.83..
On the multiprocessor, two processors handle program execution, but there is still only one lock. For
simplicity, all of the lock contention is shown affecting processor B. In the period shown, the multiprocessor
handles 14 commands. The scaling factor is thus 1.83. We stop at two processors because more would
not change the situation. The lock is now in use 100 percent of the time. In a four-way multiprocessor, the
scaling factor would be 1.83 or less.
Real programs are seldom as symmetrical as the commands in the illustration. In addition we have only
taken into account one dimension of contention: locking. If we had included cache-coherency and
processor-affinity effects, the scaling factor would almost certainly be lower.
This example illustrates that workloads often cannot be made to run faster simply by adding processors. It
is also necessary to identify and minimize the sources of contention among the threads.
Scaling is workload-dependent. Some published benchmark results imply that high levels of scalability are
easy to achieve. Most such benchmarks are constructed by running combinations of small, CPU-intensive
programs that use almost no kernel services. These benchmark results represent an upper bound on
scalability, not a realistic expectation.
Another interesting point to note for benchmarks is that in general, a one-way SMP will run slower (about
5-15 percent) than the equivalent uniprocessor running the UP version of the operating system.
Multiprocessor Response Time
A multiprocessor can only improve the execution time of an individual program to the extent that the
program can run in multithreaded mode. There are several ways to achieve parallel execution of parts of a
single program:
v Making explicit calls to libpthreads.a subroutines (or, in older programs, to the fork() subroutine) to
create multiple threads that run simultaneously.
v Processing the program with a parallelizing compiler or preprocessor that detects sequences of code
that can be executed simultaneously and generates multiple threads to run them in parallel.
v Using a software package that is itself multithreaded.
48
Performance Management Guide
Unless one or more of these techniques is used, the program will run no faster in a multiprocessor system
than in a comparable uniprocessor. In fact, because it may experience more locking overhead and delays
due to being dispatched to different processors at different times, it may be slower.
Even if all of the applicable techniques are exploited, the maximum improvement is limited by a rule that
has been called Amdahl’s Law:
v If a fraction x of a program’s uniprocessor execution time, t, can only be processed sequentially, the
improvement in execution time in an n-way multiprocessor over execution time in a comparable
uniprocessor (the speed-up) is given by the equation:
Figure 14. Amdahl’s Law. Amdahl’s Law says speed-up equals uniprocessor time divided by sequence time plus
multiprocessor time or 1 divided by x plus (x over n). Lim speed-up equals 1 divided by x and n equals infinity.
As an example, if 50 percent of a program’s processing must be done sequentially, and 50 percent can be
done in parallel, the maximum response-time improvement is less than a factor of 2 (in an otherwise-idle
4-way multiprocessor, it is at most 1.6).
SMP Thread Scheduling
Thread support divides program-execution control into two elements:
v A process is a collection of physical resources required to run the program, such as memory and
access to files.
v A thread is the execution state of an instance of the program, such as the current contents of the
instruction-address register and the general-purpose registers. Each thread runs within the context of a
given process and uses that process’s resources. Multiple threads can run within a single process,
sharing its resources.
In the SMP environment, the availability of thread support makes it easier and less expensive to
implement SMP-exploiting applications. Forking multiple processes to create multiple flows of control is
cumbersome and expensive, because each process has its own set of memory resources and requires
considerable system processing to set up. Creating multiple threads within a single process requires less
processing and uses less memory.
Thread support exists at two levels:
v libpthreads.a support in the application program environment
v Kernel thread support
Although threads are normally a convenient and efficient mechanism to exploit multiprocessing, there are
scalability limits associated with threads. Because threads share process resources and state, locking and
serialization of these resources can sometimes limit scalability.
Chapter 4. Introduction to Multiprocessing
49
Default Scheduler Processing of Migrated Workloads
The division between processes and threads is invisible to existing programs. In fact, workloads migrated
directly from earlier releases of the operating system create processes as they have always done. Each
new process is created with a single thread (the initial thread) that contends for the CPU with the threads
of other processes.
The default attributes of the initial thread, in conjunction with the new scheduler algorithms, minimize
changes in system dynamics for unchanged workloads.
Priorities can be manipulated with the nice and renice commands and the setpri() and setpriority()
system calls, as before. The scheduler allows a given thread to run for at most one time slice (normally 10
ms) before forcing it to yield to the next dispatchable thread of the same or higher priority. See Controlling
Contention for the CPU for more detail.
Scheduling Algorithm Variables
Several variables affect the scheduling of threads. Some are unique to thread support; others are
elaborations of process-scheduling considerations:
Priority
A thread’s priority value is the basic indicator of its precedence in the contention for processor
time.
Scheduler run queue position
A thread’s position in the scheduler’s queue of dispatchable threads reflects a number of
preceding conditions.
Scheduling policy
This thread attribute determines what happens to a running thread at the end of the time slice.
Contention scope
A thread’s contention scope determines whether it competes only with the other threads within its
process or with all threads in the system. A pthread created with process contention scope is
scheduled by the library, while those created with system scope are scheduled by the kernel. The
library scheduler utilizes a pool of kernels threads to schedule pthreads with process scope.
Generally, create pthreads with system scope, if they are performing I/O. Process scope is useful,
when there is a lot of intra-process synchronizations. Contention scope is a libpthreads.a
concept.
Processor affinity
The degree to which affinity is enforced affects performance.
The combinations of these considerations can seem complex, but you can choose from three distinct
approaches when you are managing a given process:
Default
The process has one thread, whose priority varies with CPU consumption and whose scheduling
policy is SCHED_OTHER.
Process-level control
The process can have one or more threads, but the scheduling policy of those threads is left as
the default SCHED_OTHER, which permits the use of the existing methods of controlling nice
values and fixed priorities. All of these methods affect all of the threads in the process identically. If
the setpri() subroutine is used, the scheduling policy of all of the threads in the process is set to
SCHED_RR.
Thread-level control
The process can have one or more threads. The scheduling policy of these threads is set to
SCHED_RR or SCHED_FIFOn, as appropriate. The priority of each thread is fixed and is
manipulated with thread-level subroutines.
50
Performance Management Guide
The scheduling policies are described in Scheduling Policy for Threads.
Thread Tuning
User threads provide independent flow of control within a process. If the user threads need to access
kernel services (such as system calls), the user threads will be serviced by associated kernel threads.
User threads are provided in various software packages with the most notable being the pthreads shared
library (libpthreads.a). With the libpthreads implementation, user threads sit on top of virtual processors
(VP) which are themselves on top of kernel threads. A multithreaded user process can use one of two
models, as follows:
1:1 Thread Model
The 1:1 model indicates that each user thread will have exactly one kernel thread mapped to it.
This is the default model on AIX 4.1, AIX 4.2, and AIX 4.3. In this model, each user thread is
bound to a VP and linked to exactly one kernel thread. The VP is not necessarily bound to a real
CPU (unless binding to a processor was done). A thread which is bound to a VP is said to have
system scope because it is directly scheduled with all the other user threads by the kernel
scheduler.
M:N Thread Model
The M:N model was implemented in AIX 4.3.1 and is also now the default model. In this model,
several user threads can share the same virtual processor or the same pool of VPs. Each VP can
be thought of as a virtual CPU available for executing user code and system calls. A thread which
is not bound to a VP is said to be a local or process scope because it is not directly scheduled
with all the other threads by the kernel scheduler. The pthreads library will handle the scheduling
of user threads to the VP and then the kernel will schedule the associated kernel thread. As of AIX
4.3.2, the default is to have one kernel thread mapped to eight user threads. This is tunable from
within the application or through an environment variable.
Thread Environment Variables
Within the libpthreads.a framework, a series of tuning knobs have been provided that might impact the
performance of the application. If possible, the application developer should provide a front-end shell script
to invoke the binary executable programs, in which the developer may specify new values to override the
system defaults. These environment variables are as follows:
AIXTHREAD_COND_DEBUG (AIX 4.3.3 and subsequent versions)
The AIXTHREAD_COND_DEBUG varible maintains a list of condition variables for use by the debugger. If
the program contains a large number of active condition variables and frequently creates and destroys
condition variables, this may create higher overhead for maintaining the list of condition variables. Setting
the variable to OFF will disable the list. Leaving this variable turned on makes debugging threaded
applications easier, but may impose some overhead.
AIXTHREAD_ENRUSG
This variable enables or disables the pthread resource collection. Turning it on allows for resource
collection of all pthreads in a process, but will impose some overhead.
AIXTHREAD_GUARDPAGES=n
For AIX 4.3 and later:
*
*
*
*
*
*
*
*
*
*
*
+-----------------------+
| pthread attr
|
+-----------------------+
| pthread struct
|
+-----------------------+
| pthread stack
|
|
|
|
|
V
|
+-----------------------+
| RED ZONE
|
+-----------------------+
<--- pthread->pt_attr
<--- pthread->pt_stk.st_limit
<--- pthread->pt_stk.st_base
<--- pthread->pt_guardaddr
Chapter 4. Introduction to Multiprocessing
51
*
*
| pthread private data |
+-----------------------+ <--- pthread->pt_data
The RED ZONE on this illustration is called the Guardpage.
Starting with AIX 5.2, the pthread attr, pthread, and ctx represent the PTH_FIXED part of the memory
allocated for a pthread.
The approximate byte sizes in the diagram below are in [] for 32-bit. For 64-bit, expect the pieces
comprising PTH_FIXED to be slightly larger and the key data to be 8 Kb, but otherwise the same.
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
+-----------------------+
| page alignment 2
|
| [8K-4K+PTH_FIXED-a1] |
+-----------------------+
| pthread ctx [368]
|
+-----------------------+<--- pthread->pt_attr
| pthread attr [112]
|
+-----------------------+ <--- pthread->pt_attr
| pthread struct [960] |
+-----------------------+ <--- pthread
| pthread stack
|
pthread->pt_stk.st_limit
|
|[96K+4K-PTH_FIXED] |
|
V
|
+-----------------------+ <--- pthread->pt_stk.st_base
| RED ZONE [4K]
|
+-----------------------+ <--- pthread->pt_guardaddr
| pthread key data [4K] |
+-----------------------+ <--- pthread->pt_data
| page alignment 1 (a1) |
| [<4K]
|
+-----------------------+
The RED ZONE on this illustration is called the Guardpage.
The decimal number of guard pages to add to the end of the pthread stack is n. It overrides the attribute
values that are specified at pthread creation time. If the application specifies its own stack, no guard pages
are created. The default is 0 and n must be a positive value.
The guardpage size in bytes is determined by multiplying n by the PAGESIZE. Pagesize is a system
determined size.
AIXTHREAD_MNRATIO (AIX 4.3 and later)
AIXTHREAD_MNRATIO controls the scaling factor of the library. This ratio is used when creating and
terminating pthreads. It ay be useful for applications with a very large number of threads. However, always
test a ratio of 1:1 because it may provide for better performance.
AIXTHREAD_MUTEX_DEBUG (AIX 4.3.3 and later)
This variable maintains a list of active mutexes for use by the debugger. If the program contains a large
number of active mutexes and frequently creates and destroys mutexes, this may create higher overhead
for maintaining the list of mutexes. Setting the variable to ON makes debugging threaded applications
easier, but may impose the additional overhead. Leaving the variable off will disable the list.
AIXTHREAD_RWLOCK_DEBUG (AIX 4.3.3 and later)
Maintains a list of read-write locks for use by the debugger. If the program contains a large number of
active read-write locks and frequently creates and destroys read-write locks, this may create higher
overhead for maintaining the list of read-write locks. Setting the variable to OFF will disable the list.
AIXTHREAD_SCOPE={P|S} (AIX 4.3.1 and later)
P signifies process-wide contention scope (M:N) and S signifies system-wide contention scope (1:1). Either
P or S should be specified and the current default is process-wide scope.
52
Performance Management Guide
The use of this environment variable impacts only those threads created with the default attribute. The
default attribute is employed, when the attr parameter to the pthread_create() subroutine is NULL.
If a user thread is created with system-wide scope, it is bound to a kernel thread and it is scheduled by
the kernel. The underlying kernel thread is not shared with any other user thread.
If a user thread is created with process-wide scope, it is subject to the user scheduler. It does not have a
dedicated kernel thread; it sleeps in user mode; it is placed on the user run queue when it is waiting for a
processor; and it is subjected to time slicing by the user scheduler.
Tests on AIX 4.3.2 have shown that certain applications can perform much better with the 1:1 model.
AIXTHREAD_SLPRATIO (AIX 4.3 and later)
This thread tuning variable controls the number of kernel threads that should be held in reserve for
sleeping threads. In general, fewer kernel threads are required to support sleeping pthreads because they
are generally woken one at a time. This conserves kernel resources.
AIXTHREAD_STK=n (AIX 4.3.3 ML 09 and AIX 5.1.1)
The decimal number number of bytes that should be allocated for each pthread. This value may be
overridden by pthread_attr_setstacksize.
MALLOCBUCKETS
Malloc buckets provides an optional buckets-based extension of the default allocator. It is intended to
improve malloc performance for applications that issue large numbers of small allocation requests. When
malloc buckets is enabled, allocation requests that fall within a predefined range of block sizes are
processed by malloc buckets. All other requests are processed in the usual manner by the default
allocator.
Malloc buckets is not enabled by default. It is enabled and configured prior to process startup by setting
the MALLOCTYPE and MALLOCBUCKETS environment variables.
For more information on mallos buckets, see General Programming Concepts: Writing and Debugging
Programs.
MALLOCMULTIHEAP={considersize,heaps:n} (AIX 4.3.1 and later)
Multiple heaps are required so that a threaded application can have more than one thread issuing
malloc(), free(), and realloc() subroutine calls. With a single heap, all threads trying to do a malloc(),
free(), or realloc() call would be serialized (that is only one thread can do malloc/free/realloc at a time).
The result is a serious impact on multi-processor machines. With multiple heaps, each thread gets its own
heap. If all heaps are being used then any new threads trying to malloc/free/realloc will have to wait (that
is serialize) until one or more of the heaps becomes available again. We still have serialization, but its
likelihood and impact are greatly reduced.
The thread-safe locking has been changed to handle this approach. Each heap has its own lock, and the
locking routine ″intelligently″ selects a heap to try to prevent serialization. If considersize is set in the
MALLOCMULTIHEAP environment variable, then the selection will also try to select any available heap
that has enough free space to handle the request instead of just selecting the next unlocked heap.
More than one option can be specified (and in any order) as long as they are comma-separated, for
example, MALLOCMULTIHEAP=considersize,heaps:3. The options are:
heaps:n
The number of heaps used can be changed with this option. If n is not valid (that is, n<=0 or
n>32), 32 is used.
considersize
Uses a different heap-selection algorithm that tries to minimize the working set size of the process.
The default is not to consider it and use the faster algorithm.
Chapter 4. Introduction to Multiprocessing
53
The default for MALLOCMULTIHEAP is NOT SET (only the first heap is used). If the environment variable
MALLOCMULTIHEAP is set (for example, MALLOCMULTIHEAP=1) then the threaded application will be
able to use all of the 32 heaps. Setting MALLOCMULTIHEAP=heaps:n will limit the number of heaps to n
instead of the 32 heaps.
For more information, see the Malloc Multiheap section in AIX 5L Version 5.2 General Programming
Concepts: Writing and Debugging Programs.
SPINLOOPTIME=n
Controls the number of times that the system will try to get a busy mutex or spin lock without taking a
secondary action such as calling the kernel to yield the process. This control is intended for MP systems,
where it is hoped that the lock being held by another actively running pthread will be released. The
parameter works only within libpthreads (user threads). The kernel parameter MAXSPIN affects spinning in
the kernel lock routines (see The schedtune -s Command). If locks are usually available within a short
amount of time, you may want to increase the spin time by setting this environment variable. The number
of times to retry a busy lock before yielding to another pthread is n. The default is 40 and n must be a
positive value.
YIELDLOOPTIME=n
Controls the number of times that the system yields the processor when trying to acquire a busy mutex or
spin lock before actually going to sleep on the lock. The processor is yielded to another kernel thread,
assuming there is another runnable one with sufficient priority. This variable has been shown to be
effective in complex applications, where multiple locks are in use. The number of times to yield the
processor before blocking on a busy lock is n. The default is 0 and n must be a positive value.
Variables for Process-Wide Contention Scope
The following environment variables impact the scheduling of pthreads created with process-wide
contention scope.
AIXTHREAD_MNRATIO=p:k
where k is the number of kernel threads that should be employed to handle p runnable pthreads.
This environment variable controls the scaling factor of the library. This ratio is used when creating
and terminating pthreads. The variable is only valid with process-wide scope; with system-wide
scope, this environment variable is ignored. The default setting is 8:1.
AIXTHREAD_SLPRATIO=k:p
where k is the number of kernel threads that should be held in reserve for p sleeping pthreads.
The sleep ratio is the number of kernel threads to keep on the side in support of sleeping
pthreads. In general, fewer kernel threads are required to support sleeping pthreads, since they
are generally woken one at a time. This conserves kernel resources. Any positive integer value
may be specified for p and k. If k>p, then the ratio is treated as 1:1. The default is 1:12.
AIXTHREAD_MINKTHREADS=n
where n is the minimum number of kernel threads that should be used. The library scheduler will
not reclaim kernel threads below this figure. A kernel thread may be reclaimed at virtually any
point. Generally, a kernel thread is targeted for reclaim as a result of a pthread terminating. The
default is 8.
Thread Debug Options
The pthreads library maintains a list of active mutexes, condition variables, and read-write locks for use by
the debugger.
When a lock is initialized, it is added to the list, assuming that it is not already on the list. The list is held
as a linked list, so determining that a new lock is not already on the list has a performance implication
when the list gets large. The problem is compounded by the fact that the list is protected by a lock
(dbx__mutexes), which is held across the search of the list. In this case other calls to the
pthread_mutex_init() subroutine are held while the search is done.
54
Performance Management Guide
If the following environment variables are set to OFF, which is the default, then the appropriate debugging
list will be disabled completely. That means the dbx command (or any debugger using the pthread debug
library) will show no objects in existence.
v AIXTHREAD_MUTEX_DEBUG
v AIXTHREAD_COND_DEBUG
v AIXTHREAD_RWLOCK_DEBUG
To set any of these environment variables to ON, use the following command:
# export variable_name=ON
Thread Tuning Summary
Depending on the type of application, the administrator can choose to use a different thread model. Tests
on AIX 4.3.2 have shown that certain applications can perform much better with the 1:1 model. This is an
important point because the default as of AIX 4.3.1 is M:N. By simply setting the environment variable
AIXTHREAD_SCOPE=S for that process, we can set the thread model to 1:1 and then compare the
performance to its previous performance when the thread model was M:N.
If you see an application creating and deleting threads, it could be the kernel threads are being harvested
because of the 8:1 default ratio of user threads to kernel threads. This harvesting along with the overhead
of the library scheduling can affect the performance. On the other hand, when thousands of user threads
exist, there may be less overhead to schedule them in user space in the library rather than manage
thousands of kernel threads. You should always try changing the scope if you encounter a performance
problem when using pthreads; in many cases, the system scope can provide better performance.
If an application is running on an SMP system, then if a user thread cannot acquire a mutex, it will attempt
to spin for up to 40 times. It could easily be the case that the mutex was available within a short amount of
time, so it may be worthwhile to spin for a longer period of time. As you add more CPUs, if the
performance goes down, this usually indicates a locking problem. You might want to increase the spin time
by setting the environment variable SPINLOOPTIME=n, where n is the number of spins. It is not unusual
to set the value as high as in the thousands depending on the speed of the CPUs and the number of
CPUs. Once the spin count has been exhausted, the thread can go to sleep waiting for the mutex to
become available or it can issue the yield() system call and simply give up the CPU but stay in a runnable
state rather than going to sleep. By default, it will go to sleep, but by setting the YIELDLOOPTIME
environment variable to a number, it will yield up to that many times before going to sleep. Each time it
gets the CPU after it yields, it can try to acquire the mutex.
Certain multithreaded user processes that use the malloc subsystem heavily may obtain better
performance by exporting the environment variable MALLOCMULTIHEAP=1 before starting the
application. The potential performance improvement is particularly likely for multithreaded C++ programs,
because these may make use of the malloc subsystem whenever a constructor or destructor is called. Any
available performance improvement will be most evident when the multithreaded user process is running
on an SMP system, and particularly when system scope threads are used (M:N ratio of 1:1). However, in
some cases, improvement may also be evident under other conditions, and on uniprocessors.
SMP Tools
All performance tools of the operating system support SMP machines. Some performance tools provide
individual processor utilization statistics. Other performance tools average out the utilization statistics for all
processors and display only the averages.
This section describes the tools that are only supported on SMP. For details on all other performance
tools, see the appropriate chapters.
Chapter 4. Introduction to Multiprocessing
55
The bindprocessor Command
Use the bindprocessor command to bind or unbind the kernel threads of a process to a processor. Root
authority is necessary to bind or unbind threads in processes that you do not own.
Note: The bindprocessor command is meant for multiprocessor systems. Although it will also work on
uniprocessor systems, binding has no effect on such systems.
To query the available processors, run the following:
# bindprocessor -q
The available processors are: 0 1 2 3
The output shows the logical processor numbers for the available processors, which are used with the
bindprocessor command as will be seen.
To bind a process whose PID is 14596 to processor 1, run the following:
# bindprocessor 14596 1
No return message is given if the command was successful. To verify if a process is bound or unbound to
a processor, use the ps -mo THREAD command as explained in Using the ps Command:
# ps
USER
root
root
root
root
root
-
-mo THREAD
PID PPID
3292 7130
14596 3292
15606 3292
16634 3292
18048 3292
-
TID
14309
15629
16895
15107
17801
ST
A
S
A
R
A
R
A
R
A
R
CP
1
1
73
73
74
74
73
73
14
14
PRI
60
60
100
100
101
101
100
100
67
67
SC
1
1
1
1
1
1
1
1
1
1
WCHAN
-
F
TT BND COMMAND
240001 pts/0
- -ksh
400
- 200001 pts/0 1 /tmp/cpubound
0
- 1 200001 pts/0
- /tmp/cpubound
0
- 200001 pts/0
- /tmp/cpubound
0
- 200001 pts/0
- ps -mo THREAD
0
- -
The column BND shows the number of the processor that the process is bound to or a dash (-) if the
process is not bound at all.
To unbind a process whose PID is 14596, use the following command:
# bindprocessor -u 14596
# ps -mo THREAD
USER
PID PPID
TID ST
root 3292 7130
- A
- 14309 S
root 14596 3292
- A
- 15629 R
root 15606 3292
- A
- 16895 R
root 16634 3292
- A
- 15107 R
root 18052 3292
- A
- 17805 R
CP
2
2
120
120
120
120
120
120
12
12
PRI
61
61
124
124
124
124
124
124
66
66
SC
1
1
1
1
1
1
0
0
1
1
WCHAN
-
F
TT BND COMMAND
240001 pts/0
- -ksh
400
- 200001 pts/0 - /tmp/cpubound
0
- - 200001 pts/0
- /tmp/cpubound
0
- 200001 pts/0
- /tmp/cpubound
0
- 200001 pts/0
- ps -mo THREAD
0
- -
When the bindprocessor command is used on a process, all of its threads will then be bound to one
processor and unbound from their former processor. Unbinding the process will also unbind all its threads.
You cannot bind or unbind an individual thread using the bindprocessor command.
However, within a program, you can use the bindprocessor() function call to bind individual threads. If the
bindprocessor() function is used within a piece of code to bind threads to processors, the threads remain
with these processors and cannot be unbound. If the bindprocessor command is used on that process,
all of its threads will then be bound to one processor and unbound from their respective former
processors. An unbinding of the whole process will also unbind all the threads.
56
Performance Management Guide
A process cannot be bound until it is started; that is, it must exist in order to be bound. When a process
does not exist, the following error displays:
# bindprocessor 7359 1
1730-002: Process 7359 does not match an existing process
When a processor does not exist, the following error displays:
# bindprocessor 7358 4
1730-001: Processor 4 is not available
Note: Do not use the bindprocessor command on the wait processes kproc.
Considerations
Binding can be useful for CPU-intensive programs that experience few interrupts. It can sometimes be
counterproductive for ordinary programs because it may delay the redispatch of a thread after an I/O until
the processor to which the thread is bound becomes available. If the thread has been blocked for the
duration of an I/O operation, it is unlikely that much of its processing context remains in the caches of the
processor to which it is bound. The thread would probably be better served if it were dispatched to the
next available processor.
Binding does not prevent other processes from being dispatched on the processor on which you bound
your process. Binding is different from partitioning. Without Workload Manager (WLM), introduced in AIX
4.3.3, it is not possible to dedicate a set of processors to a specific workload and another set of
processors to another workload. Therefore, a higher priority process might be dispatched on the processor
where you bound your process. In this case, your process will not be dispatched on other processors, and
therefore, you will not always increase the performance of the bound process. Better results may be
achieved if you increase the priority of the bound process.
If you bind a process on a heavily loaded system, you might decrease its performance because when a
processor becomes idle, the process will not be able to run on the idle processor if it is not the processor
on which the process is bound.
If the process is multithreaded, binding the process will bind all its threads to the same processor.
Therefore, the process does not take advantage of the multiprocessing, and performance will not be
improved.
Note: Use process binding with care, because it disrupts the natural load balancing provided by AIX
Version 4, and the overall performance of the system could degrade. If the workload of the machine
changes from that which is monitored when making the initial binding, system performance can
suffer. If you use the bindprocessor command, take care to monitor the machine regularly
because the environment might change, making the bound process adversely affect system
performance.
The lockstat Command
The lockstat command is only available in AIX Version 4.
As described earlier in this chapter, the use of locks and finding the right granularity is one of the big
challenges in a MP operating system. You need to have a way to determine if locks are posing a problem
on the system (for example, lock contention). The lockstat command displays lock-contention statistics for
operating system locks on SMP systems.
To determine whether the lockstat command is installed and available, run the following command:
# lslpp -lI perfagent.tools
Before you use the lockstat command, create as root a new bosboot image with the -L option to enable
lock instrumentation. Assume that the boot disk is hdisk0. Run the following:
Chapter 4. Introduction to Multiprocessing
57
# bosboot -a -d /dev/hdisk0 -L
After you run the command, reboot the machine to enable lock instrumentation. At this time, you can use
the lockstat command to look at the locking activity.
It is only possible to see which kernel locks are generated by the workload. Application locks cannot be
seen directly with the lockstat command. However, they can be seen indirectly. In that case, check the
application for bottlenecks, such as:
v One message queue with lots of processes writing and one reading (VMM)
v All processes yield (Dispatcher)
The lockstat command can be CPU-intensive because there is overhead involved with lock
instrumentation, which is why it is not turned on by default. The overhead of enabling lock instrumentation
is typically 3-5 percent. Also note that trace buffers fill up much quicker when using this option because
there are a lot of locks being used.
AIX Version 4 defines subsystems comprised of lock classes in /usr/include/sys/lockname.h. Each time
an operating system developer needs to acquire a lock, they pick up or create a lock class which serves to
identify the lock.
The lockstat command generates a report for each kernel lock that meets all specified conditions. When
no conditions are specified, the default values are used. The following are the parameters that can be
used to filter the data collected:
-a
Displays a supplementary list showing the most requested (or active) locks, regardless of the
conditions defined by other flags.
-c LockCount
Specifies how many times a lock must be requested during an interval in order to be displayed. A
lock request is a lock operation which in some cases cannot be satisfied immediately. All lock
requests are counted. The default is 200.
-b BlockRatio
Specifies a block ratio. When a lock request is not satisfied, it is said to be blocked. A lock must
have a block ratio that is higher than BlockRatio to appear in the list. The default of BlockRatio is
5 percent.
-nCheckCount
Specifies the number of locks that are to be checked. The lockstat command sorts locks
according to lock activity. This parameter determines how many of the most active locks will be
subject to further checking. Limiting the number of locks that are checked maximizes system
performance, particularly if the lockstat command is executed in intervals. The default value is 40.
-p LockRate
Specifies a percentage of the activity of the most-requested lock in the kernel. Only locks that are
more active than this percentage will be listed. The default value is 2, which means that the only
locks listed are those requested at least 2 percent as often as the most active lock.
-t MaxLocks
Specifies the maximum number of locks to be displayed. The default is 10.
If the lockstat command is executed with no options, an output similar to the following is displayed:
# lockstat
Subsys Name
Ocn
Ref/s
%Ref
%Block %Sleep
____________________________________________________________________
PFS
PROC
58
IRDWR_LOCK_CLASS
PROC_INT_CLASS
Performance Management Guide
259
1
75356
12842
37.49
6.39
9.44
17.75
0.21
0.00
The first column is the subsystem (Subsys) to which the lock belongs. Some common subsystems are as
follows:
PROC Scheduler, dispatcher or interrupt handlers
VMM
Pages, segment and free list
TCP
Sockets, NFS
PFS
I-nodes, i-cache
Next, the symbolic name of the lock class is shown. Some common classes are as follows:
TOD_LOCK_CLASS
All interrupts that need the Time-of-Day (TOD) timer
PROC_INT_CLASS
Interrupts for processes
U_TIMER_CLASS
Per-process timer lock
VMM_LOCK_VMKER
Free list
VMM_LOCK_PDT
Paging device table
VMM_LOCK_LV
Per paging space
ICACHE_LOCK_CLASS
I-node cache
The other columns are as follows:
v The Ocn column provides the occurrence number of the lock in its class.
v The reference number (Ref/s) is the number of lock requests per second.
v The %Ref column is the reference rate expressed as a percentage of all lock requests.
v The last two columns present respectively the ratio of blocking lock requests to total lock requests
(%Block) and the percentage of lock requests that cause the calling thread to sleep (%Sleep).
As a guideline, be concerned if a lock has a reference number (Ref/s) above 10000. In the example, both
classes shown present a very high rate. In this case, you may want to use the vmstat command to
investigate further. Refer to The vmstat Command for more information. If the output of the vmstat
command shows a significant amount of CPU idle time when the system seems subjectively to be running
slowly, delays might be due to kernel lock contention, because lock requestors go into blocked mode. Lock
contentions cause wasted cycles because a thread may be spinning on a busy lock or sleeping until the
lock is granted. Improper designs may even lead to deadlocks. The wasted cycles would degrade system
performance.
The lockstat command output does not indicate exactly which application is causing a problem to the
system. The lock-contentions problem can only be solved at the source-code level. For example, if your
application has a high number of processes that read and write a unique message queue, you might have
lock contention for the inter-process communication (IPC) subsystem. Adding more message queues may
reduce the level of lock contention.
In this example, many instances of a process that opens the same file for read-only were running
simultaneously on the system. In the operating system, every time a file is accessed, its i-node is updated
with the last access time. That is the reason for the high reference number observed for the lock class
IRDWR_LOCK_CLASS. Many threads were trying to update the i-node of the same file concurrently.
Chapter 4. Introduction to Multiprocessing
59
When the lockstat command is run without options, only the locks with %Block above 5 percent are listed.
You can change this behavior by specifying another BlockRatio with the -b option, as follows:
# lockstat -b 1
Subsys Name
Ocn
Ref/s
%Ref
%Block %Sleep
____________________________________________________________________
PFS
PROC
PROC
IRDWR_LOCK_CLASS
PROC_INT_CLASS
PROC_INT_CLASS
258
1
2
95660
5798
2359
60.22
3.65
1.48
69.15
4.73
1.02
0.16
0.00
0.00
In this case, all the lock requests with %Block above 1 percent will be shown.
If no lock has a BlockRatio within the given range, the output would be as follows:
# lockstat
No Contention
However, this might also indicate that the lock instrumentation has not been activated.
The -a option additionally lists the 10 most-requested (or active) locks, as follows:
# lockstat -a
Subsys Name
Ocn
Ref/s
%Ref
%Block %Sleep
____________________________________________________________________
PFS
PROC
IRDWR_LOCK_CLASS
PROC_INT_CLASS
259
1
75356
12842
37.49
6.39
9.44
17.75
0.21
0.00
First 10 largest reference rate locks :
Subsys Name
Ocn
Ref/s
%Ref
%Block %Sleep
____________________________________________________________________
PFS
PROC
PROC
PROC
XPSE
IOS
XPSE
XPSE
XPSE
XPTY
IRDWR_LOCK_CLASS
PROC_INT_CLASS
TOD_LOCK_CLASS
PROC_INT_CLASS
PSE_OPEN_LOCK
SELPOLL_LOCK_CLASS
PSE_SQH_LOCK
PSE_SQH_LOCK
PSE_SQH_LOCK
PTY_LOCK_CLASS
259
1
-2
--95
105
75
6
75356
12842
5949
5288
4498
4276
4223
4213
3585
3336
37.49
6.39
2.96
2.63
2.24
2.13
2.10
2.10
1.78
1.66
9.44
17.75
1.68
3.97
0.87
3.20
0.62
0.50
0.31
0.00
0.21
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
The meaning of the fields is the same as in the previous example. The first table is a list of locks with
%Block above 5 percent. A list of the top 10 reference-rate locks, sorted in decreasing order, is then
provided. The number of locks in the most-requested list can be changed with the -t option, as follows:
# lockstat -a -t 3
Subsys Name
Ocn
Ref/s
%Ref
%Block %Sleep
____________________________________________________________________
PFS
PROC
IRDWR_LOCK_CLASS
PROC_INT_CLASS
259
1
75356
12842
37.49
6.39
9.44
17.75
0.21
0.00
First 3 largest reference rate locks :
Subsys Name
Ocn
Ref/s
%Ref
%Block %Sleep
____________________________________________________________________
PFS
PROC
PROC
IRDWR_LOCK_CLASS
PROC_INT_CLASS
TOD_LOCK_CLASS
259
1
--
75356
12842
5949
37.49
6.39
2.96
9.44
17.75
1.68
0.21
0.00
0.00
In the previous example, the -t option specifies that only the top three reference-rate locks will be shown.
60
Performance Management Guide
If the output of the lockstat -a command looks similar to the following:
No Contention
First 10 largest reference rate locks :
Subsys Name
Ocn
Ref/s
%Ref
%Block %Sleep
____________________________________________________________________
then an empty most-requested lock list means that the lock instrumentation has not been enabled. It can
be enabled by executing the bosboot command as explained at the beginning of this section.
The lockstat command can also be run in intervals, as follows:
# lockstat 10 100
The first number passed in the command line specifies the amount of time (in seconds) between each
report. Each report contains statistics collected during the interval since the previous report. If no interval is
specified, the system gives information covering an interval of one second and then exits. The second
number determines the number of reports generated. The second number can only be specified if an
interval is given.
Note: Under excessive lock contention on large SMPs, the lockstat command does not scale well and
might not return in the time period specified.
The schedtune -s Command
If a thread wants to acquire a lock when another thread currently owns that lock and is running on another
CPU, the thread that wants the lock will spin on the CPU until the owner thread releases the lock. Prior to
AIX 4.3.1, this thread would spin indefinitely. In AIX 4.3.1, the thread spins up to a certain value as
specified by a tunable parameter called MAXSPIN.
The default value of MAXSPIN was previously 0xFFFFFFFF (the hexadecimal representation of a very
large number in decimal form) on SMP systems and 1 on UP systems. In AIX 4.3.1, the default value of
MAXSPIN is 0x4000 (16384) for SMP systems and remains at 1 on UP systems. If you notice more idle or
I/O wait time on a system that had not shown this previously, it could be that threads are going to sleep
more often. If this is causing a performance problem, then tune MAXSPIN such that it is a higher value or
set to -1 which means to spin up to 0xFFFFFFFF times.
To revise the number of times to spin before going to sleep use the -s option of the schedtune command.
To reduce CPU usage that might be caused by excessive spins, reduce the value of MAXSPIN as follows:
# /usr/samples/kernel/schedtune -s 8192
You may observe an increase in context-switching. If context-switching becomes the bottleneck, increase
MAXSPIN.
To determine whether the schedtune command is installed and available, run the following command:
# lslpp -lI bos.adt.samples
To change the value, you must be the root user.
Chapter 4. Introduction to Multiprocessing
61
62
Performance Management Guide
Chapter 5. Planning and Implementing for Performance
A program that does not perform acceptably is not functional. Every program must satisfy a set of users,
sometimes a large and diverse set. If the performance of the program is truly unacceptable to a significant
number of those users, it will not be used. A program that is not being used is not performing its intended
function.
This situation is true of licensed software packages as well as user-written applications, although most
developers of software packages are aware of the effects of poor performance and take pains to make
their programs run as fast as possible. Unfortunately, they cannot anticipate all of the environments and
uses that their programs will experience. Final responsibility for acceptable performance falls on the people
who select or write, plan for, and install software packages.
This chapter describes the stages by which a programmer or system administrator can ensure that a
newly written or purchased program has acceptable performance. (Wherever the word programmer
appears alone, the term includes system administrators and anyone else who is responsible for the
ultimate success of a program.)
To achieve acceptable performance in a program, identify and quantify acceptability at the start of the
project and never lose sight of the measures and resources needed to achieve it. Although this method
sounds elementary, some programming projects consciously reject it. They adopt a policy that might be
fairly described as design, code, debug, maybe document, and if we have time, fix the performance.
The only way that programs can predictably be made to function in time, not just in logic, is by integrating
performance considerations in the software planning and development process. Advance planning is
perhaps more critical when existing software is being installed, because the installer has less freedom than
the developer.
Although the detail of this process might seem burdensome for a small program, remember that we have a
second ″agenda.″ Not only must the new program have satisfactory performance, we must also ensure
that the addition of that program to an existing system does not degrade the performance of other
programs run on that system.
This chapter includes the following major sections:
v Identifying the Components of the Workload
v Documenting Performance Requirements
v Estimating the Resource Requirements of the Workload
v Designing and Implementing Efficient Programs
v Using Performance-Related Installation Guidelines
Identifying the Components of the Workload
Whether the program is new or purchased, small or large, the developers, the installers, and the
prospective users have assumptions about the use of the program, such as:
v Who will be using the program
v Situations in which the program will be run
v How often those situations will arise and at what times of the hour, day, month, or year
v Whether those situations will also require additional uses of existing programs
v Which systems the program will run on
v How much data will be handled, and from where
v Whether data created by or for the program will be used in other ways
© Copyright IBM Corp. 1997, 2002
63
Unless these ideas are elicited as part of the design process, they will probably be vague, and the
programmers will almost certainly have different assumptions than the prospective users. Even in the
apparently trivial case in which the programmer is also the user, leaving the assumptions unarticulated
makes it impossible to compare design to assumptions in any rigorous way. Worse, it is impossible to
identify performance requirements without a complete understanding of the work being performed.
Documenting Performance Requirements
In identifying and quantifying performance requirements, it is important to identify the reasoning behind a
particular requirement. This is part of the general capacity planning process. Users might be basing their
statements of requirements on assumptions about the logic of the program that do not match the
programmer’s assumptions. At a minimum, a set of performance requirements should document the
following:
v The maximum satisfactory response time to be experienced most of the time for each distinct type of
user-computer interaction, along with a definition of most of the time. Response time is measured from
the time that the user performs the action that says ″Go″ until the user receives enough feedback from
the computer to continue the task. It is the user’s subjective wait time. It is not from entry to a
subroutine until the first write statement.
v
v
v
v
v
v
If the user denies interest in response time and indicates that only the result is of interest, you can ask
whether ″ten times your current estimate of stand-alone execution time″ would be acceptable. If the
answer is ″yes,″ you can proceed to discuss throughput. Otherwise, you can continue the discussion of
response time with the user’s full attention.
The response time that is minimally acceptable the rest of the time. A longer response time can cause
users to think the system is down. You also need to specify rest of the time; for example, the peak
minute of a day, 1 percent of interactions. Response time degradations can be more costly or painful at
a particular time of the day.
The typical throughput required and the times it will be taking place. This is not a casual consideration.
For example, the requirement for one program might be that it runs twice a day: at 10:00 a.m. and 3:15
p.m. If this is a CPU-limited program that runs for 15 minutes and is planned to run on a multiuser
system, some negotiation is in order.
The size and timing of maximum-throughput periods.
The mix of requests expected and how the mix varies with time.
The number of users per machine and total number of users, if this is a multiuser application. This
description should include the times these users log on and off, as well as their assumed rates of
keystrokes, completed requests, and think times. You may want to investigate whether think times vary
systematically with the preceding and following request.
Any assumptions that the user is making about the machines the workload will run on. If the user has a
specific existing machine in mind, make sure you know that early on. Similarly, if the user is assuming a
particular type, size, cost, location, interconnection, or any other variable that will constrain your ability
to satisfy the preceding requirements, that assumption also becomes part of the requirements.
Satisfaction will probably not be assessed on the system where the program is developed, tested, or
first installed.
Estimating the Resource Requirements of the Workload
Unless you are purchasing a software package that comes with detailed resource-requirement
documentation, estimating resources can be the most difficult task in the performance-planning process.
The difficulty has several causes, as follows:
v There are several ways to do any task. You can write a C (or other high-level language) program, a
shell script, a perl script, an awk script, a sed script, an AIXwindows dialog, and so on. Some
techniques that may seem particularly suitable for the algorithm and for programmer productivity are
extraordinarily expensive from the performance perspective.
64
Performance Management Guide
A useful guideline is that, the higher the level of abstraction, the more caution is needed to ensure that
one does not receive a performance surprise. Consider carefully the data volumes and number of
iterations implied by some apparently harmless constructs.
v The precise cost of a single process is difficult to define. This difficulty is not merely technical; it is
philosophical. If multiple instances of a given program run by multiple users are sharing pages of
program text, which process should be charged with those pages of memory? The operating system
leaves recently used file pages in memory to provide a caching effect for programs that reaccess that
data. Should programs that reaccess data be charged for the space that was used to retain the data?
The granularity of some measurements such as the system clock can cause variations in the CPU time
attributed to successive instances of the same program.
Two approaches deal with resource-report ambiguity and variability. The first is to ignore the ambiguity
and to keep eliminating sources of variability until the measurements become acceptably consistent.
The second approach is to try to make the measurements as realistic as possible and describe the
results statistically. Note that the latter yields results that have some correlation with production
situations.
v Systems are rarely dedicated to running a single instance of a single program. There are almost always
daemons running, there is frequently communications activity, and often workload from multiple users.
These activities seldom add up linearly. For example, increasing the number of instances of a given
program may result in few new program text pages being used, because most of the program was
already in memory. However, the additional processes may result in more contention for the processor’s
caches, so that not only do the other processes have to share processor time with the newcomer, but
all processes may experience more cycles per instruction. This is, in effect, a slowdown of the
processor, as a result of more frequent cache misses.
Make your estimate as realistic as the specific situation allows, using the following guidelines:
v If the program exists, measure the existing installation that most closely resembles your own
requirements. The best method is to use a capacity planning tool such as BEST/1.
v If no suitable installation is available, do a trial installation and measure a synthetic workload.
v If it is impractical to generate a synthetic workload that matches the requirements, measure individual
interactions and use the results as input to a simulation.
v If the program does not exist yet, find a comparable program that uses the same language and general
structure, and measure it. Again, the more abstract the language, the more care is needed in
determining comparability.
v If no comparable program exists, develop a prototype of the main algorithms in the planned language,
measure the prototype, and model the workload.
v Only if measurement of any kind is impossible or infeasible should you make an educated guess. If it is
necessary to guess at resource requirements during the planning stage, it is critical that the actual
program be measured at the earliest possible stage of its development.
Keep in mind that independent software vendors (ISV) often have sizing guidelines for their applications.
In estimating resources, we are primarily interested in four dimensions (in no particular order):
CPU time
Processor cost of the workload
Disk accesses
Rate at which the workload generates disk reads or writes
LAN traffic
Number of packets the workload generates and the number of bytes of data exchanged
Real memory
Amount of RAM the workload requires
The following sections discuss how to determine these values in various situations.
Chapter 5. Planning and Implementing for Performance
65
Measuring Workload Resources
If the real program, a comparable program, or a prototype is available for measurement, the choice of
technique depends on the following:
v Whether the system is processing other work in addition to the workload we want to measure.
v Whether we have permission to use tools that may degrade performance (for example, is this system in
production or is it dedicated to our use for the duration of the measurement?).
v The degree to which we can simulate or observe an authentic workload.
Measuring a Complete Workload on a Dedicated System
Using a dedicated system is the ideal situation because we can use measurements that include system
overhead as well as the cost of individual processes.
To measure overall system performance for most of the system activity, use the vmstat command:
# vmstat 5 >vmstat.output
This gives us a picture of the state of the system every 5 seconds during the measurement run. The first
set of vmstat output contains the cumulative data from the last boot to the start of the vmstat command.
The remaining sets are the results for the preceding interval, in this case 5 seconds. A typical set of
vmstat output on a system looks similar to the following:
kthr
memory
page
faults
cpu
----- ----------- ------------------------ ------------ ----------r b
avm
fre re pi po fr
sr cy in
sy cs us sy id wa
0 1 75186
192
0
0
0
0
1
0 344 1998 403 6 2 92 0
To measure CPU and disk activity, use the iostat command:
# iostat 5 >iostat.output
This gives us a picture of the state of the system every 5 seconds during the measurement run. The first
set of iostat output contains the cumulative data from the last boot to the start of the iostat command.
The remaining sets are the results for the preceding interval, in this case 5 seconds. A typical set of iostat
output on a system looks similar to the following:
tty:
tin
0.0
Disks:
hdisk0
hdisk1
cd0
tout
0.0
% tm_act
8.0
0.0
0.0
avg-cpu:
Kbps
34.5
0.0
0.0
% user
19.4
tps
8.2
0.0
0.0
% sys
5.7
Kb_read
12
0
0
% idle
70.8
% iowait
4.1
Kb_wrtn
164
0
0
To measure memory, use the svmon command. The svmon -G command gives a picture of overall
memory use. The statistics are in terms of 4 KB pages (example from AIX 5.2):
# svmon -G
memory
pg space
pin
in use
size
65527
131072
inuse
65406
37218
free
121
work
5972
54177
pers
0
9023
clnt
0
2206
pin
5963
virtual
74711
lpage
0
0
In this example, the machine’s 256 MB memory is fully used. About 83 percent of RAM is in use for
working segments, the read/write memory of running programs (the rest is for caching files). If there are
long-running processes in which we are interested, we can review their memory requirements in detail.
The following example determines the memory used by a process of user hoetzel.
66
Performance Management Guide
# ps -fu hoetzel
UID
PID PPID
hoetzel 24896 33604
hoetzel 32496 25350
C
STIME
0 09:27:35
6 15:16:34
TTY TIME CMD
pts/3 0:00 /usr/bin/ksh
pts/5 0:00 ps -fu hoetzel
# svmon -P 24896
-----------------------------------------------------------------------------Pid Command
Inuse
Pin
Pgsp Virtual
64-bit
Mthrd
LPage
24896 ksh
7547
4045
1186
7486
N
N
N
Vsid
0
6a89aa
72d3cb
401100
3d40f
16925a
Esid
0
d
2
1
f
-
Type
work
work
work
pers
work
pers
Description
kernel seg
shared library text
process private
code,/dev/hd2:6250
shared library data
/dev/hd4:447
LPage
-
Inuse
Pin Pgsp Virtual
6324 4041 1186 6324
1064
0
0 1064
75
4
0
75
59
0
23
0
0
23
2
0
-
The working segment (5176), with 4 pages in use, is the cost of this instance of the ksh program. The
2619-page cost of the shared library and the 58-page cost of the ksh program are spread across all of the
running programs and all instances of the ksh program, respectively.
If we believe that our 256 MB system is larger than necessary, use the rmss command to reduce the
effective size of the machine and remeasure the workload. If paging increases significantly or response
time deteriorates, we have reduced memory too much. This technique can be continued until we find a
size that runs our workload without degradation. See Assessing Memory Requirements Through the rmss
Command for more information on this technique.
The primary command for measuring network usage is the netstat program. The following example shows
the activity of a specific Token-Ring interface:
# netstat -I tr0 5
input
(tr0)
output
packets errs packets errs colls
35552822 213488 30283693
0
0
300
0
426
0
0
272
2
190
0
0
231
0
192
0
0
143
0
113
0
0
408
1
176
0
0
input
(Total)
output
packets errs packets errs colls
35608011 213488 30338882
0
0
300
0
426
0
0
272
2
190
0
0
231
0
192
0
0
143
0
113
0
0
408
1
176
0
0
The first line of the report shows the cumulative network traffic since the last boot. Each subsequent line
shows the activity for the preceding 5-second interval.
Measuring a Complete Workload on a Production System
The techniques of measurement on production systems are similar to those on dedicated systems, but we
must be careful to avoid degrading system performance.
Probably the most cost-effective tool is the vmstat command, which supplies data on memory, I/O, and
CPU usage in a single report. If the vmstat intervals are kept reasonably long, for example, 10 seconds,
the average cost is relatively low. See Identifying the Performance-Limiting Resource for more information
on using the vmstat command.
Measuring a Partial Workload on a Production System
By partial workload, we mean measuring a part of the production system’s workload for possible transfer
to or duplication on a different system. Because this is a production system, we must be as unobtrusive as
possible. At the same time, we must analyze the workload in more detail to distinguish between the parts
we are interested in and those we are not. To do a partial measurement, we must discover what the
workload elements of interest have in common. Are they:
v The same program or a small set of related programs?
Chapter 5. Planning and Implementing for Performance
67
v Work performed by one or more specific users of the system?
v Work that comes from one or more specific terminals?
Depending on the commonality, we could use one of the following
# ps -ef | grep pgmname
# ps -fuusername, . . .
# ps -ftttyname, . . .
to identify the processes of interest and report the cumulative CPU time consumption of those processes.
We can then use the svmon command (judiciously) to assess the memory use of the processes.
Measuring an Individual Program
Many tools are available for measuring the resource consumption of individual programs. Some of these
programs are capable of more comprehensive workload measurements as well, but are too intrusive for
use on production systems. Most of these tools are discussed in depth in the chapters that discuss tuning
for minimum consumption of specific resources. Some of the more prominent are:
svmon
Measures the real memory used by a process. Discussed in Determining How Much Memory Is
Being Used.
time
Measures the elapsed execution time and CPU consumption of an individual program. Discussed
in Using the time Command to Measure CPU Use.
tprof
Measures the relative CPU consumption of programs, subroutine libraries, and the operating
system’s kernel. Discussed in Using the tprof Program to Analyze Programs for CPU Use.
vmstat -s
Measures the I/O load generated by a program. Discussed in Assessing Overall Disk I/O with the
vmstat Command.
Estimating Resources Required by a New Program
It is impossible to make precise estimates of unwritten programs. The invention and redesign that take
place during the coding phase defy prediction, but the following guidelines can help you to get a general
sense of the requirements. As a starting point, a minimal program would need the following:
v About 50 milliseconds of CPU time, mostly system time.
v Real Memory
– One page for program text
– About 15 pages (of which 2 are pinned) for the working (data) segment
– Access to libc.a. Normally this is shared with all other programs and is considered part of the base
cost of the operating system.
v About 12 page-in Disk I/O operations, if the program has not been compiled, copied, or used recently.
Otherwise, none required.
To the above, add the basic cost allowances for demands implied by the design (the units given are for
example purposes only):
v CPU time
– The CPU consumption of an ordinary program that does not contain high levels of iteration or costly
subroutine calls is almost unmeasurably small.
– If the proposed program contains a computationally expensive algorithm, develop a prototype and
measure the algorithm.
– If the proposed program uses computationally expensive library subroutines, such as X or Motif
constructs or the printf() subroutine, measure their CPU consumption with otherwise trivial
programs.
v Real Memory
68
Performance Management Guide
– Allow approximately 350 lines of code per page of program text, which is about 12 bytes per line.
Keep in mind that coding style and compiler options can make a difference of a factor or two in
either direction. This allowance is for pages that are touched in your typical scenario. If your design
places infrequently executed subroutines at the end of the executable program, those pages do not
normally consume real memory.
– References to shared libraries other than libc.a increase the memory requirement only to the extent
that those libraries are not shared with other programs or instances of the program being estimated.
To measure the size of these libraries, write a trivial, long-running program that refers to them and
use the svmon -P command against the process.
– Estimate the amount of storage that will be required by the data structures identified in the design.
Round up to the nearest page.
– In the short run, each disk I/O operation will use one page of memory. Assume that the page has to
be available already. Do not assume that the program will wait for another program’s page to be
freed.
v Disk I/O
– For sequential I/O, each 4096 bytes read or written causes one I/O operation, unless the file has
been accessed recently enough that some of its pages are still in memory.
– For random I/O, each access, however small, to a different 4096-byte page causes one I/O
operation, unless the file has been accessed recently enough that some of its pages are still in
memory.
– Each sequential read or write of a 4 KB page in a large file takes about 100 units. Each random read
or write of a 4 KB page takes about 300 units. Remember that real files are not necessarily stored
sequentially on disk, even though they are written and read sequentially by the program.
Consequently, the typical CPU cost of an actual disk access will be closer to the random-access cost
than to the sequential-access cost.
v Communications I/O
– If disk I/O is actually to Network File System (NFS) remote-mounted file systems, the disk I/O is
performed on the server, but the client experiences higher CPU and memory demands.
– RPCs of any kind contribute substantially to the CPU load. The proposed RPCs in the design should
be minimized, batched, prototyped, and measured in advance.
– Each sequential NFS read or write of an 4 KB page takes about 600 units on the client. Each
random NFS read or write of a 4 KB page takes about 1000 units on the client.
– Web browsing and Web serving implies considerable network I/O, with TCP connections opening
and closing quite frequently.
Transforming Program-Level Estimates to Workload Estimates
The best method for estimating peak and typical resource requirements is to use a queuing model such as
BEST/1. Static models can be used, but you run the risk of overestimating or underestimating the peak
resource. In either case, you need to understand how multiple programs in a workload interact from the
standpoint of resource requirements.
If you are building a static model, use a time interval that is the specified worst-acceptable response time
for the most frequent or demanding program (usually they are the same). Determine which programs will
typically be running during each interval, based on your projected number of users, their think time, their
key entry rate, and the anticipated mix of operations.
Use the following guidelines:
v CPU time
– Add together the CPU requirements for the all of the programs that are running during the interval.
Include the CPU requirements of the disk and communications I/O the programs will be doing.
– If this number is greater than 75 percent of the available CPU time during the interval, consider
fewer users or more CPUs.
Chapter 5. Planning and Implementing for Performance
69
v Real Memory
– The operating system memory requirement scales with the amount of physical memory. Start with 6
to 8 MB for the operating system itself. The lower figure is for a standalone system. The latter figure
is for a system that is LAN-connected and uses TCP/IP and NFS.
– Add together the working segment requirements of all of the instances of the programs that will be
running during the interval, including the space estimated for the program’s data structures.
– Add to that total the memory requirement of the text segment of each distinct program that will be
running (one copy of the program text serves all instances of that program). Remember that any
(and only) subroutines that are from unshared libraries will be part of the executable program, but
the libraries themselves will not be in memory.
– Add to the total the amount of space consumed by each of the shared libraries that will be used by
any program in the workload. Again, one copy serves all.
– To allow adequate space for some file caching and the free list, your total memory projection should
not exceed 80 percent of the size of the machine to be used.
v Disk I/O
– Add the number of I/Os implied by each instance of each program. Keep separate totals for I/Os to
small files (or randomly to large files) versus purely sequential reading or writing of large files (more
than 32 KB).
– Subtract those I/Os that you believe will be satisfied from memory. Any record that was read or
written in the previous interval is probably still available in the current interval. Beyond that, examine
the size of the proposed machine versus the total RAM requirements of the machine’s workload. Any
space remaining after the operating system’s requirement and the workload’s requirements probably
contains the most recently read or written file pages. If your application’s design is such that there is
a high probability that you will reuse recently accessed data, you can calculate an allowance for the
caching effect. Remember that the reuse is at the page level, not at the record level. If the probability
of reuse of a given record is low, but there are a lot of records per page, it is likely that some of the
records needed in any given interval will fall in the same page as other, recently used, records.
– Compare the net I/O requirements (disk I/Os per second per disk) to the approximate capabilities of
current disk drives. If the random or sequential requirement is greater than 75 percent of the total
corresponding capability of the disks that will hold application data, tuning (and possibly expansion)
will be needed when the application is in production.
v Communications I/O
– Calculate the bandwidth consumption of the workload. If the total bandwidth consumption of all of the
nodes on the LAN is greater than 70 percent of nominal bandwidth (50 percent for Ethernet), you
might want to use a network with higher bandwidth.
– Perform a similar analysis of CPU, memory, and I/O requirements of the added load that will be
placed on the server.
Note: Remember that these guidelines are intended for use only when no extensive measurement is
possible. Any application-specific measurement that can be used in place of a guideline will
considerably improve the accuracy of the estimate.
Designing and Implementing Efficient Programs
If you have determined which resource will limit the speed of your program, you can go directly to the
section that discusses appropriate techniques for minimizing the use of that resource. Otherwise, assume
that the program will be balanced and that all of the recommendations in this chapter apply. Once the
program is implemented, proceed to Identifying the Performance-Limiting Resource.
CPU-Limited Programs
The maximum speed of a truly processor-limited program is determined by:
v The algorithm used
70
Performance Management Guide
v
v
v
v
The
The
The
The
source code and data structures created by the programmer
sequence of machine-language instructions generated by the compiler
sizes and structures of the processor’s caches
architecture and clock rate of the processor itself (see Appendix D. Determining CPU Speed)
If the program is CPU-limited because it consists almost entirely of numerical computation, the chosen
algorithm will have a major effect on the performance of the program. A discussion of alternative
algorithms is beyond the scope of this book. It is assumed that computational efficiency has been
considered in choosing the algorithm.
Given an algorithm, the only items in the preceding list that the programmer can affect are the source
code, the compiler options used, and possibly the data structures. The following sections deal with
techniques that can be used to improve the efficiency of an individual program for which the user has the
source code. If the source code is not available, attempt to use tuning or workload-management
techniques.
Design and Coding for Effective Use of Caches
In Performance Concepts, we indicated that processors have a multilevel hierarchy of memory:
1. Instruction pipeline and the CPU registers
2. Instruction and data cache(s) and the corresponding translation lookaside buffers
3. RAM
4. Disk
As instructions and data move up the hierarchy, they move into storage that is faster than the level below
it, but also smaller and more expensive. To obtain the maximum possible performance from a given
machine, therefore, the programmer must make the most effective use of the available storage at each
level.
Effective use of storage means keeping it full of instructions and data that are likely to be used. An
obstacle to achieving this objective is the fact that storage is allocated in fixed-length blocks such as cache
lines and real memory pages that usually do not correspond to boundaries within programs or data
structures. Programs and data structures that are designed without regard to the storage hierarchy often
make inefficient use of the storage allocated to them, with adverse performance effects in small or heavily
loaded systems.
Taking the storage hierarchy into account means understanding and adapting to the general principles of
efficient programming in a cached or virtual-memory environment. Repackaging techniques can yield
significant improvements without recoding, and any new code should be designed with efficient storage
use in mind.
Two terms are essential to any discussion of the efficient use of hierarchical storage: locality of reference
and working set.
v The locality of reference of a program is the degree to which its instruction-execution addresses and
data references are clustered in a small area of storage during a given time interval.
v The working set of a program during that same interval is the set of storage blocks that are in use, or,
more precisely, the code or data that occupy those blocks.
A program with good locality of reference has a minimal working set, because the blocks that are in use
are tightly packed with executing code or data. A functionally equivalent program with poor locality of
reference has a larger working set, because more blocks are needed to accommodate the wider range of
addresses being accessed.
Chapter 5. Planning and Implementing for Performance
71
Because each block takes a significant amount of time to load into a given level of the hierarchy, the
objective of efficient programming for a hierarchical-storage system is to design and package code in such
a way that the working set remains as small as practical.
The following figure illustrates good and bad practice at a subroutine level. The first version of the program
is packaged in the sequence in which it was probably written. The first subroutine PriSub1 contains the
entry point of the program. It always uses primary subroutines PriSub2 and PriSub3. Some infrequently
used functions of the program require secondary subroutines SecSub1 and SecSub2. On rare occasions,
the error subroutines ErrSub1 and ErrSub2 are needed.
Figure 15. Locality of Reference. The top half of the figure describes how a binary program is packaged which shows
low locality of reference. The instructions for PriSub1 is in the binary executable first, followed by the instructions for
SecSub1, ErrSub1, PriSub2, SecSub2, ErrSub2, and PriSub3. In this executable, the instructions for PriSub1,
SecSub1, and ErrSub1 occupy into the first page of memory. The instructions for PriSub2, SecSub2, and ErrSub2
occupy the second page of memory, and the instructions for PriSub3 occupy the third page of memory. SecSub1 and
SecSub2 are infrequently used; also ErrSub1 and ErrSub2 are rarely used, if ever. Therefore, the packaging of this
program exhibits poor locality of reference and may use more memory than required. In the second half of the figure,
PriSub1, PriSub2, and PriSub3 are located next to each other and occupy the first page of memory. Following PriSub3
is SecSub1, SecSub2, and ErrSub1 which all occupy the second page of memory. Finally, ErrSub2 is at the end and
occupies the third page of memory. Because ErrSub2 may never be needed, it would reduce the memory
requirements by one page in this case.
The initial version of the program has poor locality of reference because it takes three pages of memory to
run in the normal case. The secondary and error subroutines separate the main path of the program into
three, physically distant sections.
The improved version of the program places the primary subroutines adjacent to one another and puts the
low-frequency function after that. The necessary error subroutines (which are rarely-used) are left at the
end of the executable program. The most common functions of the program can now be handled with only
one disk read and one page of memory instead of the three previously required.
Remember that locality of reference and working set are defined with respect to time. If a program works
in stages, each of which takes a significant time and uses a different set of subroutines, try to minimize the
working set of each stage.
Registers and Pipeline
In general, allocating and optimizing of register space and keeping the pipeline full are the responsibilities
of the compilers. The programmer’s main obligation is to avoid structures that defeat compiler-optimization
techniques. For example, if you use one of your subroutines in one of the critical loops of your program, it
72
Performance Management Guide
may be appropriate for the compiler to inline that subroutine to minimize execution time. If the subroutine
has been packaged in a different .c module, however, it cannot be inlined by the compiler.
Cache and TLBs
Depending on the processor architecture and model, processors have from one to several caches to hold
the following:
v Parts of executing programs
v Data used by executing programs
v Translation lookaside buffers (TLBs), which contain the mapping from virtual address to real address of
recently used pages of instruction text or data
If a cache miss occurs, loading a complete cache line can take dozens of processor cycles. If a TLB miss
occurs, calculating the virtual-to-real mapping of a page can take several dozen cycles. The exact cost is
implementation-dependent.
Even if a program and its data fit in the caches, the more lines or TLB entries used (that is, the lower the
locality of reference), the more CPU cycles it takes to get everything loaded in. Unless the instructions and
data are reused many times, the overhead of loading them is a significant fraction of total program
execution time, resulting in degraded system performance.
Good programming techniques keep the main-line, typical-case flow of the program as compact as
possible. The main procedure and all of the subroutines it calls frequently should be contiguous.
Low-probability conditions, such as obscure errors, should be tested for only in the main line. If the
condition actually occurs, its processing should take place in a separate subroutine. All such subroutines
should be grouped together at the end of the module. This arrangement reduces the probability that
low-usage code will take up space in a high-usage cache line. In large modules, some or all of the
low-usage subroutines might occupy a page that almost never has to be read into memory.
The same principle applies to data structures, although it is sometimes necessary to change the code to
compensate for the compiler’s rules about data layout.
For example, some matrix operations, such as matrix multiplication, involve algorithms that, if coded
simplistically, have poor locality of reference. Matrix operations generally involve accessing the matrix data
sequentially, such as row elements acting on column elements. Each compiler has specific rules about the
storage layout of matrixes. The FORTRAN compiler lays out matrixes in column-major format (that is, all of
the elements of column 1, followed by all the elements of column 2, and so forth). The C compiler lays out
matrixes in row-major format. If the matrixes are small, the row and column elements can be contained in
the data cache, and the processor and floating-point unit can run at full speed. However, as the size of the
matrixes increases, the locality of reference of such row/column operations deteriorates to a point where
the data can no longer be maintained in the cache. In fact, the natural access pattern of the row/column
operations generates a thrashing pattern for the cache where a string of elements accessed is larger than
the cache, forcing the initially accessed elements out and then repeating the access pattern again for the
same data.
The general solution to such matrix access patterns is to partition the operation into blocks, so that
multiple operations on the same elements can be performed while they remain in the cache. This general
technique is given the name strip mining.
Experts in numerical analysis were asked to code versions of the matrix-manipulation algorithms that
made use of strip mining and other optimization techniques. The result was a 30-fold improvement in
matrix-multiplication performance. The tuned routines are in the Basic Linear Algebra Subroutines (BLAS)
library, /usr/lib/libblas.a. A larger set of performance-tuned subroutines is the Engineering and Scientific
Subroutine Library (ESSL) licensed program.
Chapter 5. Planning and Implementing for Performance
73
The functions and interfaces of the Basic Linear Algebra Subroutines are documented in AIX 5L Version
5.2 Technical Reference. The FORTRAN run-time environment must be installed to use the library. Users
should generally use this library for their matrix and vector operations because its subroutines are tuned to
a degree that users are unlikely to achieve by themselves.
If the data structures are controlled by the programmer, other efficiencies are possible. The general
principle is to pack frequently used data together whenever possible. If a structure contains frequently
accessed control information and occasionally accessed detailed data, make sure that the control
information is allocated in consecutive bytes. This will increase the probability that all of the control
information will be loaded into the cache with a single (or at least with the minimum number of) cache
misses.
Effective Use of Preprocessors and the Compilers
The programmer who wants to obtain the highest possible performance from a given program running on
a given machine must deal with several considerations:
v There are preprocessors that can rearrange some source code structures to form a functionally
equivalent source module that can be compiled into more efficient executable code.
v Just as there are several variants of the architecture, there are several compiler options to allow optimal
compilation for a specific variant or set of variants.
v The programmer can use the #pragma feature to inform the C compiler of certain aspects of the
program that will allow the compiler to generate more efficient code by relaxing some of its worst-case
assumptions.
v There are several levels of optimization that give the compiler different degrees of freedom in instruction
rearrangement.
Programmers who are unable to experiment, should always optimize. The difference in performance
between optimized and unoptimized code is almost always so large that basic optimization (the -O option
of the compiler commands) should always be used. The only exceptions are testing situations in which
there is a specific need for straightforward code generation, such as statement-level performance analysis
using the tprof tool.
These techniques yield additional performance improvement for some programs, but the determination of
which combination yields the best performance for a specific program might require considerable
recompilation and measurement.
For an extensive discussion of the techniques for efficient use of compilers, see Optimization and Tuning
Guide for XL Fortran, XL C and XL C++.
Levels of Optimization
The levels of optimization in the compilers are as follows:
No Optimization
In the absence of any version of the -O flag, the compiler generates straightforward code with no
instruction reordering or other attempt at performance improvement.
-O or -O2
These equivalent flags cause the compiler to optimize on the basis of conservative assumptions about
code reordering. Only explicit relaxations such as the #pragma directives are used. This level performs no
software pipelining, loop unrolling, or simple predictive commoning. It also constrains the amount of
memory the compiler can use.
-O3
This flag directs the compiler to be aggressive about the optimization techniques used and to use as much
memory as necessary for maximum optimization.
74
Performance Management Guide
This level of optimization may result in functional changes to the program if the program is sensitive to the
following:
v Floating-point exceptions
v The sign of zero
v Precision effects of reordering calculations
These side effects can be avoided, at some performance cost, by using the -qstrict option in combination
with -O3.
The -qhot option, in combination with -O3, enables predictive commoning and some unrolling.
The result of these changes is that large or complex routines should have the same or better performance
with the -O3 option (possibly in conjunction with -qstrict or -qhot) that they had with the -O option in
earlier versions of the compiler.
-O4
This flag is equivalent to -O3 -qipa with automatic generation of architecture and tuning option ideal for
that platform.
-O5
This flag is similiar to -O4, except in this case,-qipa = level = 2.
Compiling for Specific Hardware Platforms (-qarch, -qtune)
Systems can use several type of processors. By using the -qarch and -qtune options, you can optimize
programs for the special instructions and particular strengths of these processors.
Follow these guidelines:
v If your program will be run only on a single system, or on a group of systems with the same processor
type, use the -qarch option to specify the processor type.
v If your program will be run on systems with different processor types, and you can identify one
processor type as the most important, use the appropriate -qarch and -qtune settings. FORTRAN and
HPF users can use the xxlf and xxlhpf commands to select these settings interactively.
v If your program is intended to run on the full range of processor implementations, and is not intended
primarily for one processor type, do not use either -qarch or -qtune.
C Options for string.h Subroutine Performance
The operating system provides the ability to embed the string subroutines in the application program rather
than using them from libc.a, saving call and return linkage time. To embed the string subroutines, the
source code of the application must have the following statement prior to the use of the subroutine(s):
#include <string.h>
C and C++ Coding Style for Best Performance
In many cases, the performance cost of a C construct is not obvious, and sometimes is even
counter-intuitive. Some of these situations are as follows:
v Whenever possible, use int instead of char or short.
In most cases, char and short data items take more instructions to manipulate. The extra instructions
cost time, and, except in large arrays, any space that is saved by using the smaller data types is more
than offset by the increased size of the executable program.
v If you have to use a char, make it unsigned, if possible.
A signed char takes another two instructions more than an unsigned char each time the variable is
loaded into a register.
v Use local (automatic) variables rather than global variables whenever possible.
Chapter 5. Planning and Implementing for Performance
75
Global variables require more instructions to access than local variables. Also, in the absence of
information to the contrary, the compiler assumes that any global variable may have been changed by a
subroutine call. This change has an adverse effect on optimization because the value of any global
variable used after a subroutine call will have to be reloaded.
v When it is necessary to access a global variable (that is not shared with other threads), copy the value
into a local variable and use the copy.
Unless the global variable is accessed only once, it is more efficient to use the local copy.
v Use binary codes rather than strings to record and test for situations. Strings consume both data and
instruction space. For example, the sequence:
#define situation_1 1
#define situation_2 2
#define situation_3 3
int situation_val;
situation_val = situation_2;
. . .
if (situation_val == situation_1)
. . .
is much more efficient than the following sequence:
char situation_val[20];
strcpy(situation_val,"situation_2");
. . .
if ((strcmp(situation_val,"situation_1"))==0)
. . .
v When strings are necessary, use fixed-length strings rather than null-terminated variable-length strings
wherever possible.
The mem*() family of routines, such as memcpy(), is faster than the corresponding str*() routines, such
as strcpy(), because the str*() routines must check each byte for null and the mem*() routines do not.
Compiler Execution Time
In the operating system, the C compiler can be invoked by two different commands: cc and xlc. The cc
command, which has historically been used to invoke the system’s C compiler, causes the C compiler to
run in langlevel=extended mode. This mode allows the compilation of existing C programs that are not
ANSI-compliant. It also consumes processor time.
If the program being compiled is, in fact, ANSI-compliant, it is more efficient to invoke the C compiler by
using the xlc command.
Use of the -O3 flag implicitly includes the -qmaxmem option. This option allows the compiler to use as
much memory as necessary for maximum optimization. This situation can have two effects:
v On a multiuser system, a large -O3 compilation may consume enough memory to have an adverse
effect on the performance experienced by other users.
v On a system with small real memory, a large -O3 compilation may consume enough memory to cause
high paging rates, making compilation slow.
Memory-Limited Programs
To programmers accustomed to struggling with the addressing limitations of, for instance, the DOS
environment, 256 MB virtual memory segments seem effectively infinite. The programmer is tempted to
ignore storage constraints and code for minimum path length and maximum simplicity. Unfortunately, there
is a drawback to this attitude. Virtual memory is large, but it is variable-speed. The more memory used,
the slower it becomes, and the relationship is not linear. As long as the total amount of virtual storage
actually being touched by all programs (that is, the sum of the working sets) is slightly less than the
amount of unpinned real memory in the machine, virtual memory performs at about the speed of real
76
Performance Management Guide
memory. As the sum of the working sets of all executing programs passes the number of available page
frames, memory performance degrades rapidly (if VMM memory load control is turned off) by up to two
orders of magnitude. When the system reaches this point, it is said to be thrashing. It is spending almost
all of its time paging, and no useful work is being done because each process is trying to steal back from
other processes the storage necessary to accommodate its working set. If VMM memory load control is
active, it can avoid this self-perpetuating thrashing, but at the cost of significantly increased response
times.
The degradation caused by inefficient use of memory is much greater than that from inefficient use of the
caches because the difference in speed between memory and disk is so much higher than the difference
between cache and memory. Where a cache miss can take a few dozen CPU cycles, a page fault typically
takes 10 milliseconds or more, which is at least 400,000 CPU cycles.
Although VMM memory load control can ensure that incipient thrashing situations do not become
self-perpetuating, unnecessary page faults still exact a cost in degraded response time and reduced
throughput (see Tuning VMM Memory Load Control with the schedtune Command).
Structuring of Pageable Code
To minimize the code working set of a program, the general objective is to pack code that is frequently
executed into a small area, separating it from infrequently executed code. Specifically:
v Do not put long blocks of error-handling code in line. Place them in separate subroutines, preferably in
separate source-code modules. This applies not only to error paths, but to any functional option that is
infrequently used.
v Do not structure load modules arbitrarily. Try to ensure that frequently called object modules are located
as close to their callers as possible. Object modules consisting (ideally) of infrequently called
subroutines should be concentrated at the end of the load module. The pages they inhabit will seldom
be read in.
Structuring of Pageable Data
To minimize the data working set, try to concentrate the frequently used data and avoid unnecessary
references to virtual-storage pages. Specifically:
v Use the malloc() or calloc() subroutines to request only as much space as you actually need. Never
request and then initialize a maximum-sized array when the actual situation uses only a fraction of it.
When you touch a new page to initialize the array elements, you effectively force the VMM to steal a
page of real memory from someone. Later, this results in a page fault when the process that owned that
page tries to access it again. The difference between the malloc() and calloc() subroutines is not just in
the interface.
v Because the calloc() subroutine zeroes the allocated storage, it touches every page that is allocated,
whereas the malloc() subroutine touches only the first page. If you use the calloc() subroutine to
allocate a large area and then use only a small portion at the beginning, you place an unnecessary load
on the system. Not only do the pages have to be initialized; if their real-memory frames are reclaimed,
the initialized and never-to-be-used pages must be written out to paging space. This situation wastes
both I/O and paging-space slots.
v Linked lists of large structures (such as buffers) can result in similar problems. If your program does a
lot of chain-following looking for a particular key, consider maintaining the links and keys separately
from the data or using a hash-table approach instead.
v Locality of reference means locality in time, not just in address space. Initialize data structures just prior
to when they are used (if at all). In a heavily loaded system, data structures that are resident for a long
time between initialization and use risk having their frames stolen. Your program would then experience
an unnecessary page fault when it began to use the data structure.
v Similarly, if a large structure is used early and then left untouched for the remainder of the program, it
should be released. It is not sufficient to use the free() subroutine to free the space that was allocated
with the malloc() or calloc() subroutines. The free() subroutine releases only the address range that
the structure occupied. To release the real memory and paging space, use the disclaim() subroutine to
disclaim the space as well. The call to disclaim() should be before the call to free().
Chapter 5. Planning and Implementing for Performance
77
Misuse of Pinned Storage
To avoid circularities and time-outs, a small fraction of the system must be pinned in real memory. For this
code and data, the concept of working set is meaningless, because all of the pinned information is in real
storage all the time, whether or not it is used. Any program (such as a user-written device driver) that pins
code or data must be carefully designed (or scrutinized, if ported) to ensure that only minimal amounts of
pinned storage are used. Some cautionary examples are as follows:
v Code is pinned on a load-module (executable file) basis. If a component has some object modules that
must be pinned and others that can be pageable, package the pinned object modules in a separate
load module.
v Pinning a module or a data structure because there might be a problem is irresponsible. The designer
should understand the conditions under which the information could be required and whether a page
fault could be tolerated at that point.
v Pinned structures whose required size is load-dependent, such as buffer pools, should be tunable by
the system administrator.
Using Performance-Related Installation Guidelines
This topic provides recommended actions you should take (or not take) before and during the installation
process.
Operating System Preinstallation Guidelines
Two situations require consideration, as follows:
v Installing the Operating System on a New System
Before you begin the installation process, be sure that you have made decisions about the size and
location of disk file systems and paging spaces, and that you understand how to communicate those
decisions to the operating system.
v Installing a New Level of the Operating System on an Existing System
If you are upgrading to a new level of the operating system, do the following:
– Identify all uses in your present environment of the release-specific schedtune and vmtune
performance tools. Because these tools can only be run by the root user, their use should not be
widespread.
– If these programs are used during system boot, such as from /etc/inittab, they should be temporarily
removed or bypassed until you are convinced by documentation or experiment that your use of these
tools work correctly in the new release of the operating system.
CPU Preinstallation Guidelines
Use the default CPU scheduling parameters, such as the time-slice duration. Unless you have extensive
monitoring and tuning experience with the same workload on a nearly identical configuration, leave these
parameters unchanged at installation time.
See Monitoring and Tuning CPU Use for post-installation recommendations.
Memory Preinstallation Guidelines
Do not make any memory-threshold changes until you have had experience with the response of the
system to the actual workload.
See Monitoring and Tuning Memory Use for post-installation recommendations.
78
Performance Management Guide
Disk Preinstallation Guidelines
The mechanisms for defining and expanding logical volumes attempt to make the best possible default
choices. However, satisfactory disk-I/O performance is much more likely if the installer of the system tailors
the size and placement of the logical volumes to the expected data storage and workload requirements.
Recommendations are as follows:
v If possible, the default volume group, rootvg, should consist only of the physical volume on which the
system is initially installed. Define one or more other volume groups to control the other physical
volumes in the system. This recommendation has system management, as well as performance,
advantages.
v If a volume group consists of more than one physical volume, you may gain performance by:
– Initially defining the volume group with a single physical volume.
– Defining a logical volume within the new volume group. This definition causes the allocation of the
volume group’s journal logical volume on the first physical volume.
– Adding the remaining physical volumes to the volume group.
– Defining the high-activity file systems on the newly added physical volumes.
– Defining only very-low-activity file systems, if any, on the physical volume containing the journal
logical volume. This affects performance only if I/O would cause journaled file system (JFS) log
transactions.
This approach separates journaled I/O activity from the high-activity data I/O, increasing the
probability of overlap. This technique can have an especially significant effect on NFS server
performance, because both data and journal writes must be complete before NFS signals I/O
complete for a write operation.
v At the earliest opportunity, define or expand the logical volumes to their maximum expected sizes. To
maximize the probability that performance-critical logical volumes will be contiguous and in the desired
location, define or expand them first.
v High-usage logical volumes should occupy parts of multiple disk drives. If the RANGE of physical
volumes option on the Add a Logical Volume screen of the SMIT program (fast path: smitty mklv) is
set to maximum, the new logical volume will be divided among the physical volumes of the volume
group (or the set of physical volumes explicitly listed).
v If the system has drives of different types (or you are trying to decide which drives to order), consider
the following guidelines:
– Place large files that are normally accessed sequentially on the fastest available disk drive.
– If you expect frequent sequential accesses to large files on the fastest disk drives, limit the number
of disk drivers per disk adapter.
– When possible, attach drives with critical, high-volume performance requirements to a high speed
adapter. These adapters have features, such as back-to-back write capability, that are not available
on other disk adapters.
– On the smaller disk drives, logical volumes that will hold large, frequently accessed sequential files
should be allocated in the outer_edge of the physical volume. These disks have more blocks per
track in their outer sections, which improves sequential performance.
– On the original SCSI bus, the highest-numbered drives (those with the numerically largest SCSI
addresses, as set on the physical drives) have the highest priority. Subsequent specifications usually
attempt to maintain compatibility with the original specification. Thus, the order from highest to lowest
priority is as follows: 7-6-5-4-3-2-1-0-15-14-13-12-11-10-9-8.
In most situations this effect is not noticeable, but large sequential file operations have been known
to exclude low-numbered drives from access to the bus. You should probably configure the disk
drives holding the most response-time-critical data at the highest addresses on each SCSI bus.
The lsdev -Cs scsi command reports on the current address assignments on each SCSI bus. For
the original SCSI adapter, the SCSI address is the first number in the fourth pair of numbers in the
output. In the following output example, one 400 GB disk is at SCSI address 4, another at address 5,
the 8mm tape drive at address 1, and the CDROM drive is at address 3.
Chapter 5. Planning and Implementing for Performance
79
cd0
hdisk0
hdisk1
rmt0
Available
Available
Available
Available
10-80-00-3,0
10-80-00-4,0
10-80-00-5,0
10-80-00-1,0
SCSI Multimedia CD-ROM Drive
16 Bit SCSI Disk Drive
16 Bit SCSI Disk Drive
2.3 GB 8mm Tape Drive
– Large files that are heavily used and are normally accessed randomly, such as data bases, should
be spread across two or more physical volumes.
See Monitoring and Tuning Disk I/O Use for post-installation recommendations.
Placement and Sizes of Paging Spaces
The general recommendation is that the sum of the sizes of the paging spaces should be equal to at least
twice the size of the real memory of the machine, up to a memory size of 256 MB (512 MB of paging
space).
Note: For memories larger than 256 MB, the following is recommended:
total paging space = 512 MB + (memory size - 256 MB) * 1.25
However, starting with AIX 4.3.2 and Deferred Page Space Allocation, this guideline may tie up
more disk space than actually necessary. See Choosing a Page Space Allocation Method for more
information.
Ideally, there should be several paging spaces of roughly equal size, each on a different physical disk
drive. If you decide to create additional paging spaces, create them on physical volumes that are more
lightly loaded than the physical volume in rootvg. When allocating paging space blocks, the VMM allocates
four blocks, in turn, from each of the active paging spaces that has space available. While the system is
booting, only the primary paging space (hd6) is active. Consequently, all paging-space blocks allocated
during boot are on the primary paging space. This means that the primary paging space should be
somewhat larger than the secondary paging spaces. The secondary paging spaces should all be of the
same size to ensure that the algorithm performed in turn can work effectively.
The lsps -a command gives a snapshot of the current utilization level of all the paging spaces on a
system. You can also used the psdanger() subroutine to determine how closely paging-space utilization is
approaching critical levels. As an example, the following program uses the psdanger() subroutine to
provide a warning message when a threshold is exceeded:
/* psmonitor.c
Monitors system for paging space low conditions. When the condition is
detected, writes a message to stderr.
Usage:
psmonitor [Interval [Count]]
Default: psmonitor 1 1000000
*/
#include <stdio.h>
#include <signal.h>
main(int argc,char **argv)
{
int interval = 1;
/* seconds */
int count = 1000000;
/* intervals */
int current;
/* interval */
int last;
/* check */
int kill_offset;
/* returned by psdanger() */
int danger_offset;
/* returned by psdanger() */
/* are there any parameters at all? */
if (argc > 1) {
if ( (interval = atoi(argv[1])) < 1 ) {
fprintf(stderr,"Usage: psmonitor [ interval [ count ] ]\n");
exit(1);
}
if (argc > 2) {
if ( (count = atoi( argv[2])) < 1 ) {
fprintf(stderr,"Usage: psmonitor [ interval [ count ] ]\n");
80
Performance Management Guide
}
}
exit(1);
}
}
last = count -1;
for(current = 0; current < count; current++) {
kill_offset = psdanger(SIGKILL); /* check for out of paging space */
if (kill_offset < 0)
fprintf(stderr,
"OUT OF PAGING SPACE! %d blocks beyond SIGKILL threshold.\n",
kill_offset*(-1));
else {
danger_offset = psdanger(SIGDANGER); /* check for paging space low */
if (danger_offset < 0) {
fprintf(stderr,
"WARNING: paging space low. %d blocks beyond SIGDANGER threshold.\n",
danger_offset*(-1));
fprintf(stderr,
"
%d blocks below SIGKILL threshold.\n",
kill_offset);
}
}
if (current < last)
sleep(interval);
}
Performance Implications of Disk Mirroring
If mirroring is being used and Mirror Write Consistency is on (as it is by default), consider locating the
copies in the outer region of the disk, because the Mirror Write Consistency information is always written
in Cylinder 0. From a performance standpoint, mirroring is costly, mirroring with Write Verify is costlier still
(extra disk rotation per write), and mirroring with both Write Verify and Mirror Write Consistency is costliest
of all (disk rotation plus a seek to Cylinder 0). From a fiscal standpoint, only mirroring with writes is
expensive. Although an lslv command will usually show Mirror Write Consistency to be on for non-mirrored
logical volumes, no actual processing is incurred unless the COPIES value is greater than one. Write
Verify defaults to off, because it does have meaning (and cost) for non-mirrored logical volumes.
Beginning in AIX 5.1, a mirror write consistency option called Passive Mirror Write Consistency (MWC) is
available. The default mechanism for ensuring mirror write consistency is Active MWC. Active MWC
provides fast recovery at reboot time after a crash has occurred. However, this benefit comes at the
expense of write performance degradation, particularly in the case of random writes. Disabling Active
MWC eliminates this write-performance penalty, but upon reboot after a crash you must use the syncvg -f
command to manually synchronize the entire volume group before users can access the volume group. To
achieve this, automatic vary-on of volume groups must be disabled.
Enabling Passive MWC not only eliminates the write-performance penalty associated with Active MWC, but
logical volumes will be automatically resynchronized as the partitions are being accessed. This means that
the administrator does not have to synchronize logical volumes manually or disable automatic vary-on.
The disadvantage of Passive MWC is that slower read operations may occur until all the partitions have
been resynchronized.
You can select either mirror write consistency option within SMIT when creating or changing a logical
volume. The selection option takes effect only when the logical volume is mirrored (copies > 1).
Performance Implications of Mirrored Striped LVs
Prior to AIX 4.3.3, logical volumes could not be mirrored and striped at the same time. Logical volume
mirroring and striping combines the data availability of RAID 1 with the performance of RAID 0 entirely
through software. Volume groups that contain striped and mirrored logical volumes cannot be imported into
AIX 4.3.2 or earlier.
Chapter 5. Planning and Implementing for Performance
81
Communications Preinstallation Guidelines
See the summary of communications tuning recommendations in Tuning TCP and UDP Performance and
Tuning mbuf Pool Performance.
For correct placement of adapters and various performance guidelines, see PCI Adapter Placement
Reference.
82
Performance Management Guide
Chapter 6. System Monitoring and Initial Performance
Diagnosis
This chapter describes tools and techniques for monitoring performance-related system activity and
diagnosing performance problems. The major sections are:
v The Case for Continuous Performance Monitoring
v Using the vmstat, iostat, netstat, and sar Commands
v Using the topas Monitor
v Using the Performance Diagnostic Tool
v Using the Performance Toolbox
v Determining the Kind of Performance Problem Reported
v Identifying the Performance-Limiting Resource
v Managing Workload
The Case for Continuous Performance Monitoring
In some installations, performance activity is monitored on a demand basis. When a performance problem
is reported, the performance analyst runs one or more commands in an attempt to determine why the
problem occurred. In some cases, explicit recreation of the problem is needed in order to collect analysis
data. The result is that users experience every performance problem twice.
It is usually more effective to monitor performance continuously, preferably with automatic collection of
additional data if performance deteriorates. The costs of continuous monitoring are outweighed by the
advantages, such as:
v Monitoring can sometimes detect incipient problems before they have an adverse effect.
v Monitoring can detect problems that happen to users who are reluctant to complain, as well as
problems that are not quite severe enough to complain about, but are affecting productivity and morale.
v Monitoring can collect data when a problem occurs for the first time.
Successful monitoring involves the following main activities:
v Periodically obtaining performance-related information from the operating system
v Storing the information for future use in problem diagnosis
v Displaying the information for the benefit of the system administrator
v Detecting situations that require additional data collection or responding to directions from the system
administrator to collect such data, or both
v Collecting and storing the necessary detail data
The following sections discuss several approaches to continuous monitoring. These approaches are not
mutually exclusive, but use of more than one may involve some redundancy.
Using the vmstat, iostat, netstat, and sar Commands
The vmstat, iostat, netstat, and sar commands have functional characteristics that make them useful for
continuous monitoring of system performance. These commands can:
v Produce reports indefinitely at a fixed interval
v Report on activity that varies with different types of load
v Report on activity since the last previous report (except the sar command), so changes in activity are
easy to detect
© Copyright IBM Corp. 1997, 2002
83
The following example shows samples of the periodic reports produced by these programs.
# vmstat 5 2
kthr
memory
page
faults
----- ----------- ------------------------ -----------r b
avm
fre re pi po fr
sr cy in
sy cs
0 1 75318
142
0
0
0
0
0
0 299 1845 315
0 1 75318
141
0
0
0
0
0
0 626 4949 842
cpu
----------us sy id wa
5 2 93 0
8 6 87 0
See The vmstat Command (CPU), The vmstat Command (Memory), and Assessing Disk Performance with
the vmstat Command for detailed discussions of the vmstat command.
# iostat 5 2
tty:
tin
0.0
tty:
tin
0.0
Disks:
hdisk0
hdisk1
cd0
tout
avg-cpu: % user
% sys
% idle
0.5
5.2
1.8
92.7
" Disk history since boot not available. "
tout
41.4
% tm_act
7.1
0.0
0.0
avg-cpu:
Kbps
28.4
0.0
0.0
% user
8.1
tps
6.9
0.0
0.0
% sys
5.5
Kb_read
0
0
0
% idle
84.2
% iowait
0.3
% iowait
2.2
Kb_wrtn
144
0
0
The system maintains a history of disk activity. If the history is disabled (smitty chgsys -> Continuously
maintain DISK I/O history [false]) you see this message when running the iostat command: Disk
history since boot not available. The interval disk I/O statistics are unaffected by this. For detailed
discussion of this command, see The iostat Command and Assessing Disk Performance with the iostat
Command.
# netstat -I tr0 5
input
(tr0)
output
packets errs packets errs colls
725227
0
445748
0
0
0
0
0
0
0
2
0
0
0
0
CTRL C
input
(Total)
output
packets errs packets errs colls
799996
0
520517
0
0
0
0
0
0
0
2
0
0
0
0
Other useful netstat command options to use are -s and -v. See The netstat Command for details.
# sar -P ALL 5 2
AIX rugby 3 4 00058033A100
11:17:41 cpu
11:17:46 0
1
2
3
11:17:51 0
1
2
3
Average
0
1
2
3
-
12/01/99
%usr
0
0
0
0
0
0
0
0
0
0
%sys
0
0
0
0
0
0
0
0
0
0
%wio
0
0
0
0
0
0
0
0
0
0
%idle
100
100
100
100
100
100
100
100
100
100
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
100
100
100
100
100
For details on the sar command, see The sar Command and Assessing Disk Performance with the sar
Command.
84
Performance Management Guide
Remember that the first report from the vmstat, iostat, and netstat commands is for cumulative activity
since the last system boot. The second report shows activity for the first 5-second interval. The sar
command does not report the cumulative activity since the last system boot.
These commands are the basic foundation on which to construct a performance-monitoring mechanism.
Shell scripts can be written to perform data reduction on the command output and warn of performance
problems or record data on the status of the system when a problem is occurring. For example, a shell
script could test the CPU idle percentage for zero and execute another shell script when that
CPU-saturated condition occurred. A script such as:
# ps -ef | egrep -v "STIME|$LOGNAME" | sort +3 -r | head -n 15
would record the 15 active processes that had consumed the most CPU time recently (other than the
processes owned by the user of the script).
Depending on the required level of sophistication, creating such a family of shell scripts can be a
substantial project. Fortunately, there are packages available that require less development and setup and
have considerably more function than the typical installation would want to implement locally.
Using the topas Monitor
The topas command reports vital statistics about the activity on the local system in a character terminal. It
requires AIX 4.3.3 or later with the perfagent.tools fileset, and AIX 5 or later with the bos.perf.tools
fileset installed on the system.
The program extracts and displays statistics from the system with a default interval of two seconds. On the
operating system version 4.3.3, the output consists on two fixed part and one variable section.
The top two lines at the left of the output show the name of the system the topas program runs on, the
date and time of the last observation, and the monitoring interval. Following that is a fixed section which
lists the CPU utilization in both numeric and block-graph format.
The second fixed section fills the rightmost 32 positions of the output. It contains five subsections of
statistics, EVENTS/QUEUES, FILE/TTY, PAGING, MEMORY and PAGING SPACE.
The variable part of topas running on the operating system version 4.3.3, can have one, two or three
subsections. If more than one appears, the subsections are always shown in the following order:
v Network Interfaces
v Physical Disks
v Processes
The three sections present respectively, a sorted list of the busiest network interfaces, disks, and
processes. The following is an example of the output generated by the topas command running on the
operating system 4.3.3:
Topas Monitor for host:
Wed Nov 8 14:19:05 2000
lambic
Interval:
2
Kernel
User
Wait
Idle
0.5
0.0
0.0
99.5
|
|
|
|
|
|
|############################|
Interf
tr0
lo0
KBPS
0.0
0.0
I-Pack
0.0
0.0
Disk
Busy%
hdisk1
0.0
hdisk0
0.0
KBPS
0.0
0.0
O-Pack
0.0
0.0
KB-In
0.0
0.0
EVENTS/QUEUES
Cswitch
20
Syscall
13
Reads
4
Writes
0
Forks
0
Execs
0
Runqueue
0.0
Waitqueue
0.0
KB-Out
0.0
0.0 PAGING
Faults
TPS KB-Read KB-Writ Steals
0.0
0.0
0.0 PgspIn
0.0
0.0
0.0 PgspOut
FILE/TTY
Readch
Writech
Rawin
Ttyout
Igets
Namei
Dirblk
MEMORY
0 Real,MB
0 % Comp
0 % Noncomp
0 % Client
912
42
0
42
0
0
0
159
24.0
8.0
0.0
Chapter 6. System Monitoring and Initial Performance Diagnosis
85
gil
topas
syncd
init
snmpd
sendmail
ksh
inetd
portmap
(1032)
(5436)
(2370)
(1)
(4386)
(3880)
(5944)
(4128)
(3616)
0.5%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
PgSp:
PgSp:
PgSp:
PgSp:
PgSp:
PgSp:
PgSp:
PgSp:
PgSp:
0.0mb
0.3mb
0.1mb
0.6mb
0.7mb
0.7mb
0.3mb
0.3mb
0.5mb
root
root
root
root
root
root
root
root
root
PageIn
PageOut
Sios
0
0
0
PAGING SPACE
Size,MB
128
% Used
2.7
% Free
97.2
Press "h" for help screen.
Press "q" to quit program.
On the operating system Version 5, the topas program has been enhanced to add two alternate screens,
the CPU utilization report has become an optional subsection, and the fixed section includes an additional
subsection with NFS statistics, while the variable section includes a new WLM subsection. The complete
list of optional subsections on the operating system version 5, in the order in which they are displayed is:
v CPU utilization
v Network Interfaces
v Physical Disks
v WLM classes
v Processes
Here is an example of the main screen when the topas program runs on the operating system version 5:
Topas Monitor for host:
Wed Nov 8 12:32:12 2000
mothra
Interval:
2
Kernel
User
Wait
Idle
0.0
0.2
0.0
99.7
|
|
|
|
|
|
|############################|
Network
lo0
tr0
KBPS
0.0
0.0
I-Pack
0.0
0.0
KB-Out
0.0
0.0 PAGING
Faults
Disk
Busy%
KBPS
TPS KB-Read KB-Writ Steals
hdisk0
0.0
0.0
0.0
0.0
0.0 PgspIn
PgspOut
WLM-Class (Active)
CPU%
Mem% Disk-I/O% PageIn
System
0
9
0
PageOut
Shared
0
4
0
Sios
Name
topas
gil
syncd
wlmsched
sendmail
PID
14446
1806
3144
2064
5426
CPU%
0.2
0.0
0.0
0.0
0.0
O-Pack
0.0
0.0
PgSp
0.6
0.0
0.1
0.0
0.7
KB-In
0.0
0.0
EVENTS/QUEUES
Cswitch
25
Syscall
24
Reads
0
Writes
0
Forks
0
Execs
0
Runqueue
0.0
Waitqueue
1.0
Class
System
System
System
System
System
0
0
0
0
0
0
0
FILE/TTY
Readch
Writech
Rawin
Ttyout
Igets
Namei
Dirblk
MEMORY
Real,MB
% Comp
% Noncomp
% Client
0
23
0
0
0
0
0
511
30.0
29.0
0.0
PAGING SPACE
Size,MB
0
% Used
0.6
NFS (calls/sec) % Free
99.3
ServerV2
0
ClientV2
0
Press:
ServerV3
0
"h" for help
ClientV3
0
"q" to quit
Additionaly, except for the Process variable subsection, all subsections can be sorted by any of their
columns by simply moving the cursor on top of the desired column. All variable subsections, except the
Processes list, now have two views, one presenting the top resource users, and another view presenting
the sum of the activity in a one line report: showing the total disk or network thoughput for instance. For
the CPU subsection, the user can select either the list of busiest processors, or the global CPU utilization
as shown in the previous example.
On the operating system version 5, two additional screens are available. The first alternate screen
(reachable with the P command or the -P flag), presents the list of busiest processes, similar to the
processes subsection of the main screen, but with more columns. This screen is sortable by any of its
columns. Here is an example of such a display:
86
Performance Management Guide
Topas Monitor for host:
USER
root
root
root
root
root
root
root
root
root
root
root
root
root
root
root
root
root
root
root
root
root
root
PID
1806
1032
1290
1548
1
2064
2698
3144
3362
3666
3982
4644
4912
5202
5426
5678
5934
6192
6450
6708
0
6990
PPID
0
0
0
0
0
0
1
1
0
1
0
1
1
4912
4912
4912
4912
4912
4912
4912
0
1
PRI
37
16
60
36
60
16
60
60
60
60
60
17
60
60
60
60
60
60
60
60
16
60
mothra
DATA
NI
RES
41
16
41
3
41
4
41
4
20
197
41
4
20
14
20
40
20
4
20
135
20
4
20
6
20
106
20
94
20
195
20
161
20
103
20
217
20
137
20
157
41
3
20
106
Interval:
2
TEXT PAGE
RES SPACE
TIME
3374
16
13:25
3374
3
0:00
3374
4
0:02
3374
4
0:26
9
180
0:24
3374
4
0:04
2
14
0:00
1
36
5:19
3374
4
0:00
23
123
0:00
3374
4
0:01
3374
6
0:00
13
85
0:00
8
84
0:01
76
181
0:12
11
147
0:01
11
88
0:00
61
188
0:21
10
116
0:00
29
139
0:06
3374
3
7:08
10
86
0:06
Wed Nov 8 12:27:34 2000
PGFAULTS
CPU% I/O OTH COMMAND
1.0
0
0 gil
0.0
0
0 lrud
0.0
0
0 xmgc
0.0
0
0 netm
0.0
0
0 init
0.0
0
0 wlmsched
0.0
0
0 shlap
0.0
0
0 syncd
0.0
0
0 lvmbb
0.0
0
0 errdemon
0.0
0
0 rtcmd
0.0
0
0 dog
0.0
0
0 srcmstr
0.0
0
0 syslogd
0.0
0
0 sendmail
0.0
0
0 portmap
0.0
0
0 inetd
0.0
0
0 snmpd
0.0
0
0 dpid2
0.0
0
0 hostmibd
0.0
0
0
0.0
0
0 cron
The second alternate screen (reachable with the W command, or the -W flag), is divided in two sections.
The top section is the same list of busiest WLM classes as presented in the WLM subsection of the main
screen, also sortable by any of its columns. When the user selects one the WLM classes shown using the
arrow keys and the ″f″ key, the second section of the screen will show the list of hot processes within the
selected WLM class. Here is an example of the WLM full screen report:
Topas Monitor for host:
WLM-Class (Active)
System
Shared
Default
Unmanaged
Unclassified
mothra
CPU%
0
0
0
0
0
Interval:
Mem%
0
0
0
0
0
2
Wed Nov
Disk-I/O%
0
0
0
0
0
8 12:30:54 2000
==============================================================================
DATA TEXT PAGE
PGFAULTS
USER
PID PPID PRI NI
RES
RES SPACE
TIME CPU% I/O OTH COMMAND
root
1
0 108 20
197
9
180
0:24 0.0
0
0 init
root
1032
0 16 41
3 3374
3
0:00 0.0
0
0 lrud
root
1290
0 60 41
4 3374
4
0:02 0.0
0
0 xmgc
root
1548
0 36 41
4 3374
4
0:26 0.0
0
0 netm
root
1806
0 37 41
16 3374
16
13:25 0.0
0
0 gil
root
2064
0 16 41
4 3374
4
0:04 0.0
0
0 wlmsched
root
2698
1 108 20
14
2
14
0:00 0.0
0
0 shlap
root
3144
1 108 20
40
1
36
5:19 0.0
0
0 syncd
root
3362
0 108 20
4 3374
4
0:00 0.0
0
0 lvmbb
root
3666
1 108 20
135
23
123
0:00 0.0
0
0 errdemon
root
3982
0 108 20
4 3374
4
0:01 0.0
0
0 rtcmd
The individual command line flags, commands, and sections are described in the formal documentation of
the topas command in the AIX 5L Version 5.2 Commands Reference.
Chapter 6. System Monitoring and Initial Performance Diagnosis
87
Using the Performance Diagnostic Tool
The Performance Diagnostic Tool (PDT) is a tool available in operating system version 4. PDT collects
configuration and performance information and attempts to identify potential problems, both current and
future.
PDT is an optionally installable component of the Base Operating System. Its name is bos.perf.diag_tool.
After PDT has been installed, it must be activated with the /usr/sbin/perf/diag_tool/pdt_config command.
This causes appropriate entries to be made in the crontab file, which causes PDT to run periodically,
recording data and looking for new trends.
In assessing the configuration and the historical record of performance measurements, PDT attempts to
identify:
v Resource imbalances: asymmetrical aspects of configuration or device utilization
v Usage trends: changes in usage levels that will lead to saturation
v New consumers of resources: expensive processes that have not been observed previously
v Inappropriate system parameter values: settings that may cause problems
v Errors: hardware or software problems that may lead to performance problems
PDT is described in detail in Chapter 13. Using Performance Diagnostic Tool (PDT).
Using the Performance Toolbox
The Performance Toolbox (PTX) is a licensed product that allows graphical display of a variety of
performance-related metrics. Among the advantages of PTX over ASCII reporting programs is that it is
much easier to check current performance with a glance at the graphics monitor than by looking at a
screen full of numbers. PTX also facilitates the combination of information from multiple
performance-related commands and allows recording and playback.
PTX contains tools for local and remote system-activity monitoring and tuning. The product consists of two
main components: the PTX Manager and the PTX Agent. The PTX Agent is available as a separate
licensed product called the Performance Aide for AIX. The following figure shows a simplified LAN
configuration in which the PTX Manager is monitoring the activity of several systems.
88
Performance Management Guide
Figure 16. LAN Configuration Using Performance Toolbox. This illustration shows five nodes of a local area network
that are connected using the star topology. Each node on the network has the performance toolbox (PTX) agent
running on it. One of the nodes is the PTX Manager and can monitor the other nodes via the resident agent.
The main purpose of the PTX Manager is to collect and display data from the various systems in the
configuration. The primary program for this purpose is xmperf. The primary program used by the Agent to
collect and transmit data to the Manager is xmservd.
PTX is described in detail in the Performance Toolbox Version 2 and 3 for AIX: Guide and Reference and
Customizing Performance Toolbox and Performance Toolbox Extensions for AIX.
Recording with the Performance Agent
If you need to record information or a playback facility, several tools are included with the manager code
(xmperf, 3dmon, azizo, ptxrlog). Other tools (xmservd, ptxtab, ptxsplit) are packaged with the agent
code.
Best suited for continuous monitoring are the following tools:
v The ptxrlog command can produce recordings in ASCII format, which allows you to print the output or
post-process it, or it can produce a recording file in binary to be viewed with the azizo or xmperf
commands.
v The xmservd daemon can act as a recording facility and is controlled through the xmservd.cf
configuration file. This daemon can simultaneously provide near real-time network-based data
monitoring and local recording on a given node.
v The xmtrend daemon, much like xmservd, acts as a recording facility. The main difference is in the
storage requirements for each daemon. Typical xmservd recordings can consume several megabytes
of disk storage every hour. The xmtrend agent was created to focus on providing manageable 24 x 7
long-term recordings of large metric sets.
v The jazizo tool is a Java version that replaces azizo. Jazizo is a tool for analyzing the long-term
performance characteristics of a system. It analyzes recordings created by the xmtrend daemon, and
provides displays of the recorded data that can be customized.
v The wlmperf tools provide graphical views of Workload Manager (WLM) resource activities by class.
This tool can generate reports from trend recordings made by the PTX daemons covering minutes,
hours, days, weeks, or monthly periods.
Chapter 6. System Monitoring and Initial Performance Diagnosis
89
Determining the Kind of Performance Problem Reported
When a performance problem is reported, determining the kind of performance problem often helps the
performance analyst to narrow the list of possible culprits.
A Particular Program Runs Slowly
Although this situation might seem trivial, there are still questions to be asked:
v Has the program always run slowly?
If the program has just started running slowly, a recent change might be the cause.
v Has the source code been changed or a new version installed?
If so, check with the programmer or vendor.
v Has something in the environment changed?
If a file used by the program (including its own executable program) has been moved, it may now be
experiencing Local Area Network (LAN) delays that did not exist previously. Or, files may be contending
for a single disk accessor that were on different disks previously.
If the system administrator has changed system-tuning parameters, the program may be subject to
constraints that it did not experience previously. For example, if the schedtune -r command has been
used to change the way priority is calculated, programs that used to run rather quickly in the
background may now be slowed down, while foreground programs have speeded up.
v Is the program written in the perl, awk, csh, or some other interpretive language?
While they allow programs to be written quickly, interpretive languages have the problem that they are
not optimized by a compiler. Also, it is easy in a language like perl or awk to request an extremely
compute- or I/O-intensive operation with a few characters. It is often worthwhile to perform a desk check
or informal peer review of such programs with the emphasis on the number of iterations implied by each
operation.
v Does the program always run at the same speed, or is it sometimes faster?
The file system uses some of system memory to hold pages of files for future reference. If a disk-limited
program is run twice in quick succession, it will normally run faster the second time than the first.
Similar phenomena might be observed with programs that use NFS. This can also occur with large
programs, such as compilers. The program’s algorithm might not be disk-limited, but the time needed to
load a large executable program might make the first execution of the program much longer than
subsequent ones.
v If the program has always run slowly, or has slowed down without any obvious change in its
environment, look at its dependency on resources.
Identifying the Performance-Limiting Resource describes techniques for finding the bottleneck.
Everything Runs Slowly at a Particular Time of Day
Most people have experienced the rush-hour slowdown that occurs because a large number of people in
the organization habitually use the system at one or more particular times each day. This phenomenon is
not always simply due to a concentration of load. Sometimes it is an indication of an imbalance that is (at
present) only a problem when the load is high. Other sources of recurring situations in the system should
be considered.
v If you run the iostat and netstat commands for a period that spans the time of the slowdown (or have
previously captured data from your monitoring mechanism), are some disks much more heavily used
than others? Is the CPU idle percentage consistently near zero? Is the number of packets sent or
received unusually high?
– If the disks are unbalanced, see Monitoring and Tuning Disk I/O Use.
– If the CPU is saturated, use the ps or topas commands to identify the programs being run during
this period. The script given in Using the vmstat, iostat, netstat, and sar Commands simplifies the
search for the heaviest CPU users.
90
Performance Management Guide
– If the slowdown is counter-intuitive, such as paralysis during lunch time, look for a pathological
program such as a graphic xlock or game program. Some versions of the xlock program are known
to use huge amounts of CPU time to display graphic patterns on an idle display. It is also possible
that someone is running a program that is a known CPU burner and is trying to run it at the least
intrusive time.
v Unless your /var/adm/cron/cron.allow file is null, you may want to check the contents of the
/var/adm/cron/crontab directory for expensive operations.
If you find that the problem stems from conflict between foreground activity and long-running,
CPU-intensive programs that are, or should be, run in the background, consider using the command
schedtune -r -d to give the foreground higher priority. See Tuning the Thread-Priority-Value Calculation.
Everything Runs Slowly at Unpredictable Times
The best tool for this situation is an overload detector, such as the filtd program (a component of PTX).
The filtd daemon can be set up to execute shell scripts or collect specific information when a particular
condition is detected. You can construct a similar, but more specialized, mechanism using shell scripts
containing the vmstat, iostat, netstat, sar, and ps commands.
If the problem is local to a single system in a distributed environment, there is probably a pathological
program at work, or perhaps two that intersect randomly.
Everything That an Individual User Runs is Slow
Sometimes a system seems to ″single out″ an individual.
v Quantify the problem. Ask the user which commands he uses frequently, and run them with the time
command, as in the following example:
# time cp .profile testjunk
real
0m0.08s
user
0m0.00s
sys
0m0.01s
Then run them under a satisfactory user ID. Is there a difference in the reported real time?
v A program should not show much CPU time (user+sys) difference from run to run, but may show a real
time difference because of more or slower I/O. Are the user’s files on an NFS-mounted directory? On a
disk that has high activity for other reasons?
v Check the user’s .profile file for unusual $PATH specifications. For example, if you always search a
couple of NFS-mounted directories (fruitlessly) before searching /usr/bin, everything will take longer.
A Number of LAN-Connected Systems Slow Down Simultaneously
There are some common problems that arise in the transition from independent systems to distributed
systems. The problems usually result from the need to get a new configuration running as soon as
possible, or from a lack of awareness of the cost of certain functions. In addition to tuning the LAN
configuration in terms of maximum transmission units (MTU) and mbufs (see Monitoring and Tuning
Communications I/O Use), look for LAN-specific pathologies or nonoptimal situations that may have
evolved through a sequence of individually reasonable decisions.
v Use network statistics to ensure that there are no physical network problems. Ensure that commands
such as netstat -v, entstat, tokstat, atmstat, or fddistat do not show excessive errors or collision on
the adapter.
v Some types of software or firmware bugs can sporadically saturate the LAN with broadcast or other
packets.
When a broadcast storm occurs, even systems that are not actively using the network can be slowed by
the incessant interrupts and by the CPU resource consumed in receiving and processing the packets.
These problems are better detected and localized with LAN analysis devices than with the normal
performance tools.
Chapter 6. System Monitoring and Initial Performance Diagnosis
91
v Do you have two LANs connected through a system?
Using a system as a router consumes large amounts of CPU time to process and copy packets. It is
also subject to interference from other work being processed by the system. Dedicated hardware
routers and bridges are usually a more cost-effective and robust solution to the need to connect LANs.
v Is there a clearly defensible purpose for each NFS mount?
At some stages in the development of distributed configurations, NFS mounts are used to give users on
new systems access to their home directories on their original systems. This situation simplifies the
initial transition, but imposes a continuing data communication cost. It is not unknown to have users on
system A interacting primarily with data on system B and vice versa.
Access to files through NFS imposes a considerable cost in LAN traffic, client and server CPU time, and
end-user response time. A general guideline is that user and data should normally be on the same
system. The exceptions are those situations in which an overriding concern justifies the extra expense
and time of remote data. Some examples are a need to centralize data for more reliable backup and
control, or a need to ensure that all users are working with the most current version of a program.
If these and other needs dictate a significant level of NFS client-server interchange, it is better to
dedicate a system to the role of server than to have a number of systems that are part-server,
part-client.
v Have programs been ported correctly (and justifiably) to use remote procedure calls (RPCs)?
The simplest method of porting a program into a distributed environment is to replace program calls with
RPCs on a 1:1 basis. Unfortunately, the disparity in performance between local program calls and RPCs
is even greater than the disparity between local disk I/O and NFS I/O. Assuming that the RPCs are
really necessary, they should be batched whenever possible.
Everything on a Particular Service or Device Slows Down at Times
If everything that uses a particular device or service slows down at times, refer to the topic that covers that
particular device or service:
v
v
v
v
Monitoring
Monitoring
Monitoring
Monitoring
and
and
and
and
Tuning
Tuning
Tuning
Tuning
CPU Use
Memory Use
Disk I/O Use
Communications I/O Use
Make sure you have followed the configuration recommendations in the appropriate subsystem manual
and the recommendations in the appropriate ″Monitoring and Tuning″ chapter of this book.
Identifying the Performance-Limiting Resource
Perhaps the best tool for an overall look at resource utilization while running a multiuser workload is the
vmstat command. The vmstat command reports CPU and disk-I/O activity, as well as memory utilization
data. The following command causes the vmstat command to begin writing a one-line summary report of
system activity every 5 seconds:
# vmstat 5
Because no count was specified following the interval, reporting continues until the command is canceled.
The following vmstat report was made on a system running AIXwindows and several synthetic
applications (some low-activity intervals have been removed for example purposes):
kthr
----r b
0 0
0 0
0 0
0 0
92
memory
page
faults
----------- ------------------------ -----------avm
fre re pi po fr
sr cy in
sy cs
8793
81
0
0
0
1
7
0 125
42 30
8793
80
0
0
0
0
0
0 155 113 79
8793
57
0
3
0
0
0
0 178
28 69
9192
66
0
0 16 81 167
0 151
32 34
Performance Management Guide
cpu
----------us sy id wa
1 2 95 2
14 8 78 0
1 12 81 6
1 6 77 16
0
0
0
0
0
0
0
0
0
0
0
0
0 9193
0 9193
0 9693
0 9693
0 10193
0 11194
0 11194
0 5480
0 5467
1 4797
1 3778
0 3751
65
65
69
69
57
64
63
755
5747
5821
6119
6139
0
0
0
0
0
0
0
3
0
0
0
0
0
0
0
0
0
0
0
0
0 53 100 216
0
0
0
0
0
0
0
0
0 38 201 1080
0
0
0
0
1
0
0
0
3
0
0
0
21
0
0
0
24
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
117
120
168
134
124
168
141
154
167
191
188
145
29
30
27
96
29
29
111
107
39
192
170
24
26
31
57
60
32
57
65
71
68
125
98
54
1
1
1
12
1
2
12
13
1
20
5
1
3
3
4
4
3
8
7
8
16
5
8
10
96
95
63
84
94
62
81
78
79
42
41
89
0
0
33
0
2
29
0
2
5
33
46
0
The columns of interest for this initial assessment are pi and po in the page category and the four columns
in the cpu category.
v Entries pi and po are the paging-space page ins and page outs, respectively. If any paging-space I/O is
taking place, the workload may be approaching (or is beyond) the system’s memory limits.
v If the sum of us and sy (user and system) CPU-utilization percentages is greater than 90 percent in a
given 5-second interval, the workload was approaching the CPU limits of the system during that interval.
v If the wa (I/O wait) percentage is close to zero (and pi and po are zero), some time is being spent
waiting on nonoverlapped file I/O, and some part of the workload is I/O-limited.
By ″approaching″ its limits, we mean that some parts of the workload are already experiencing a
slowdown due to the critical resource. The longer response times might not be subjectively significant yet,
but an increase in that element of the workload will cause a rapid deterioration of performance.
If the vmstat command indicates a significant amount of I/O wait time, the iostat command will give more
detailed information. The following command causes the iostat command to begin writing summary
reports of I/O activity and CPU utilization every 5 seconds:
# iostat 5 3
Because a count of 3 was specified following the interval, reporting will stop after the third report.
The following iostat report was made on a system running the same workload as the vmstat reports
above, but at a different time. The first report is for the cumulative activity since the preceding boot, while
subsequent reports are for activity during the preceding 5-second interval:
tty:
tin
0.0
Disks:
hdisk0
hdisk1
hdisk2
cd0
tty:
% tm_act
0.0
0.0
0.4
0.0
tin
0.0
Disks:
hdisk0
hdisk1
hdisk2
cd0
tty:
Disks:
hdisk0
tout
4.3
tout
30.3
% tm_act
0.2
0.0
0.0
0.0
tin
0.0
tout
8.4
% tm_act
0.0
avg-cpu:
Kbps
0.2
0.0
1.5
0.0
tps
0.0
0.0
0.3
0.0
avg-cpu:
Kbps
0.8
0.0
0.0
0.0
% sys
7.2
Kb_read
4
0
0
0
% user
0.2
tps
0.0
% sys
0.6
Kb_read
7993
2179
67548
0
% user
8.8
tps
0.2
0.0
0.0
0.0
avg-cpu:
Kbps
0.0
% user
0.2
% sys
5.8
Kb_read
0
% idle
98.8
%iowait
0.4
Kb_wrtn
4408
1692
59151
0
% idle
83.9
%iowait
0.2
Kb_wrtn
0
0
0
0
% idle
0.0
%iowait
93.8
Kb_wrtn
0
Chapter 6. System Monitoring and Initial Performance Diagnosis
93
hdisk1
hdisk2
cd0
0.0
98.4
0.0
0.0
575.6
0.0
0.0
61.9
0.0
0
396
0
0
2488
0
The first report, which displays cumulative activity since the last boot, shows that the I/O on this system is
unbalanced. Most of the I/O (86.9 percent of kilobytes read and 90.7 percent of kilobytes written) is to
hdisk2, which contains both the operating system and the paging space. The cumulative CPU utilization
since boot statistic is usually meaningless, unless the system is used consistently 24 hours a day.
The second report shows a small amount of disk activity reading from hdisk0, which contains a separate
file system for the system’s primary user. The CPU activity arises from two application programs and the
iostat command itself. Although the output of the iostat command is redirected to a file, the output is not
voluminous, and the system is not sufficiently memory-constrained to force any output during this interval.
In the third report, we have artificially created a near-thrashing condition by running a program that
allocates, and stores into, a large amount of memory (about 26 MB in this example). In the example,
hdisk2 is active 98.4 percent of the time, which results in 93.8 percent I/O wait.
If the vmstat command indicates that there is a significant amount of CPU idle time when the system
seems subjectively to be running slowly, you might be experiencing delays due to kernel lock contention.
This possibility can be investigated with the lockstat command (only available in AIX Version 4).
Determining the Limiting Factor for a Single Program
If you are the sole user of a system, you can get a general idea of whether a program is I/O or CPU
dependent by using the time command as follows:
# time cp foo.in foo.out
real
user
sys
0m0.13s
0m0.01s
0m0.02s
Note: Examples of the time command here and elsewhere in this guide use the version that is built into
the Korn shell (ksh). The official time command (/usr/bin/time) reports with a lower precision.
In this example, the fact that the real, elapsed time for the execution of the cp program (.13 seconds) is
significantly greater than the sum (.03 seconds) of the user and system CPU times indicates that the
program is I/O bound. This occurs primarily because the foo.in file has not been read recently.
On an SMP the output takes on a new meaning. See time and timex Cautions.
Running the same command a few seconds later against the same file gives the following output:
real
user
sys
0m0.06s
0m0.01s
0m0.03s
Most or all of the pages of the foo.in file are still in memory because there has been no intervening
process to cause them to be reclaimed and because the file is small compared with the amount of RAM
on the system. A small foo.out file would also be buffered in memory, and a program using it as input
would show little disk dependency.
If you are trying to determine the disk dependency of a program, you must be sure that its input is in an
authentic state. That is, if the program will normally be run against a file that has not been accessed
recently, you must make sure that the file used in measuring the program is not in memory. If, on the other
hand, a program is usually run as part of a standard sequence in which it gets its input from the output of
the preceding program, you should prime memory to ensure that the measurement is authentic. For
example, the following command would have the effect of priming memory with the pages of the foo.in
file:
94
Performance Management Guide
# cp foo.in /dev/null
The situation is more complex if the file is large compared to RAM. If the output of one program is the
input of the next and the entire file will not fit in RAM, the second program will read pages at the head of
the file, which displaces pages at the end. Although this situation is very hard to simulate authentically, it is
nearly equivalent to one in which no disk caching takes place.
The case of a file that is (perhaps just slightly) larger than RAM is a special case of the RAM versus disk
analysis discussed in the next section.
Determining Whether the Problem is Related to Disk or Memory
Just as a large fraction of real memory is available for buffering files, the system’s page space is available
as temporary storage for program working data that has been forced out of RAM. Suppose that you have
a program that reads little or no data and yet shows the symptoms of being I/O dependent. Worse, the
ratio of real time to user + system time does not improve with successive runs. The program is probably
memory-limited, and its I/O is to, and possibly from the paging space. A way to check on this possibility is
shown in the following vmstatit (or in AIX 5.1, see the vmstat -I option) shell script:
vmstat -s >temp.file
# cumulative counts before the command
time $1
# command under test
vmstat -s >>temp.file # cumulative counts after execution
grep "pagi.*ins" temp.file >>results
# extract only the data
grep "pagi.*outs" temp.file >>results # of interest
The vmstatit script summarizes the voluminous vmstat -s report, which gives cumulative counts for a
number of system activities since the system was started.
If the shell script is run as follows:
# vmstatit "cp file1 file2"
2>results
the result is as follows:
real
user
sys
0m0.03s
0m0.01s
0m0.02s
2323 paging
2323 paging
4850 paging
4850 paging
space
space
space
space
page
page
page
page
ins
ins
outs
outs
The before-and-after paging statistics are identical, which confirms our belief that the cp command is not
paging-bound. An extended variant of the vmstatit shell script can be used to show the true situation, as
follows:
vmstat -s >temp.file
time $1
vmstat -s >>temp.file
echo "Ordinary Input:"
grep "^[ 0-9]*page ins"
echo "Ordinary Output:"
grep "^[ 0-9]*page outs"
echo "True Paging Output:"
grep "pagi.*outs"
echo "True Paging Input:"
grep "pagi.*ins"
>>results
temp.file >>results
>>results
temp.file >>results
>>results
temp.file >>results
>>results
temp.file >>results
Because file I/O in the operating system is processed through the VMM, the vmstat -s command reports
ordinary program I/O as page ins and page outs. When the previous version of the vmstatit shell script
was run against the cp command of a large file that had not been read recently, the result was as follows:
Chapter 6. System Monitoring and Initial Performance Diagnosis
95
real
0m2.09s
user
0m0.03s
sys
0m0.74s
Ordinary Input:
46416 page ins
47132 page ins
Ordinary Output:
146483 page outs
147012 page outs
True Paging Output:
4854 paging space
4854 paging space
True Paging Input:
2527 paging space
2527 paging space
page outs
page outs
page ins
page ins
The time command output confirms the existence of an I/O dependency. The increase in page ins shows
the I/O necessary to satisfy the cp command. The increase in page outs indicates that the file is large
enough to force the writing of dirty pages (not necessarily its own) from memory. The fact that there is no
change in the cumulative paging-space-I/O counts confirms that the cp command does not build data
structures large enough to overload the memory of the test machine.
The order in which this version of the vmstatit script reports I/O is intentional. Typical programs read file
input and then write file output. Paging activity, on the other hand, typically begins with the writing out of a
working-segment page that does not fit. The page is read back in only if the program tries to access it.
The fact that the test system has experienced almost twice as many paging space page outs as paging
space page ins since it was booted indicates that at least some of the programs that have been run on
this system have stored data in memory that was not accessed again before the end of the program.
Memory-Limited Programs provides more information. See also Monitoring and Tuning Memory Use.
To show the effects of memory limitation on these statistics, the following example observes a given
command in an environment of adequate memory (32 MB) and then artificially shrinks the system using
the rmss command (see Assessing Memory Requirements Through the rmss Command). The command
sequence
# cc -c ed.c
# vmstatit "cc -c ed.c" 2>results
first primes memory with the 7944-line source file and the executable file of the C compiler, then measures
the I/O activity of the second execution:
real
0m7.76s
user
0m7.44s
sys
0m0.15s
Ordinary Input:
57192 page ins
57192 page ins
Ordinary Output:
165516 page outs
165553 page outs
True Paging Output:
10846 paging space
10846 paging space
True Paging Input:
6409 paging space
6409 paging space
page outs
page outs
page ins
page ins
Clearly, this is not I/O limited. There is not even any I/O necessary to read the source code. If we then
issue the following command:
# rmss -c 8
96
Performance Management Guide
to change the effective size of the machine to 8 MB, and perform the same sequence of commands, we
get the following output:
real
0m9.87s
user
0m7.70s
sys
0m0.18s
Ordinary Input:
57625 page ins
57809 page ins
Ordinary Output:
165811 page outs
165882 page outs
True Paging Output:
11010 paging space
11061 paging space
True Paging Input:
6623 paging space
6701 paging space
page outs
page outs
page ins
page ins
The following symptoms of I/O dependency are present:
v Elapsed time longer than total CPU time
v Significant amounts of ordinary I/O on the nth execution of the command
The fact that the elapsed time is longer than in the memory-unconstrained situation, and the existence of
significant amounts of paging-space I/O, make it clear that the compiler is being hampered by insufficient
memory.
Note: This example illustrates the effects of memory constraint. No effort was made to minimize the use
of memory by other processes, so the absolute size at which the compiler was forced to page in
this environment does not constitute a meaningful measurement.
To avoid working with an artificially shrunken machine until the next restart, run
# rmss -r
to release back to the operating system the memory that the rmss command had sequestered, thus
restoring the system to its normal capacity.
Managing Workload
When you have exhausted the program performance-improvement and system-tuning possibilities, and
performance is still unsatisfactory at times, you have three choices:
v Let the situation remain as is.
v Upgrade the performance-limiting resource.
v Adopt workload-management techniques.
If you adopt the first approach, some of your less stoic users will experience increasing frustration and
decreasing productivity. If you choose to upgrade, you will most likely have to justify the expenditure. That
person will undoubtedly want to know if you have exhausted all possibilities with the current system, which
means you need to investigate the possibilities of workload management.
Workload management simply means assessing the components of the workload to determine whether
they are all needed as soon as possible. Usually, there is work that can wait for a while; for example, a
report that is needed first thing in the morning. That report is equally useful when run at 3 a.m. as at 4
p.m. on the preceding day. The difference is that at 3 a.m. it uses CPU cycles and other resources that
would otherwise be idle. Use the at or crontab command to request a program to run at a specific time or
at regular intervals.
Chapter 6. System Monitoring and Initial Performance Diagnosis
97
Similarly, some programs that do have to be run during the day can be run at reduced priority. They will
take longer to complete, but they will be less in competition with really time-critical processes.
A related technique is moving work from one machine to another; for example, running a compilation on
the machine where the source code resides. This kind of workload balancing requires more planning and
monitoring, because reducing the load on the network and increasing the CPU load on a server might
result in a net loss.
Starting with AIX 4.3.3, AIX Workload Manager (WLM) is provided as an integrated part of the operating
system kernel. WLM is designed to give the system administrator greater control over how the scheduler
and virtual memory manager (VMM) allocate CPU and physical memory resources to processes. In AIX
5.1 and subsequent releases, disk usage can also be controlled by WLM. This can be used to prevent
different classes of jobs from interfering with each other and to explicitly apply resources based on the
requirements of different groups of users. For further information, see Server Consolidation on RS/6000.
98
Performance Management Guide
Chapter 7. Monitoring and Tuning CPU Use
This chapter deals with techniques for detecting runaway or CPU-intensive programs and minimizing their
adverse effect on performance.
Readers who are not familiar with CPU scheduling may want to look at Performance Overview of the CPU
Scheduler before continuing.
The following sections cover the different aspects of CPU tuning:
v Monitoring CPU Use
v Using the time Command to Measure CPU Use
v
v
v
v
v
v
v
Identifying CPU-Intensive Programs
Using the tprof Program to Analyze Programs for CPU Use
Using the pprof Command to Measure CPU usage of Kernel Threads
Detecting Instruction Emulation with the emstat Tool
Detecting Alignment Exceptions with the alstat Tool
Restructuring Executable Programs with the fdpr Program
Controlling Contention for the CPU
v CPU-Efficient User ID Administration (The mkpasswd Command)
Monitoring CPU Use
The processing unit is one of the fastest components of the system. It is comparatively rare for a single
program to keep the CPU 100 percent busy (that is, 0 percent idle and 0 percent wait) for more than a few
seconds at a time. Even in heavily loaded multiuser systems, there are occasional 10 milliseconds (ms)
periods that end with all threads in a wait state. If a monitor shows the CPU 100 percent busy for an
extended period, there is a good chance that some program is in an infinite loop. Even if the program is
″merely″ expensive, rather than broken, it needs to be identified and dealt with.
The vmstat Command (CPU)
The first tool to use is the vmstat command, which quickly provides compact information about various
system resources and their related performance problems. The vmstat command reports statistics about
kernel threads in the run and wait queue, memory, paging, disks, interrupts, system calls, context
switches, and CPU activity. The reported CPU activity is a percentage breakdown of user mode, system
mode, idle time, and waits for disk I/O.
Note: If the vmstat command is used without any options or only with the interval and optionally, the
count parameter, such as vmstat 2 10; then the first line of numbers is an average since system
reboot.
As a CPU monitor, the vmstat command is superior to the iostat command in that its one-line-per-report
output is easier to scan as it scrolls and there is less overhead involved if there are a lot of disks attached
to the system. The following example can help you identify situations in which a program has run away or
is too CPU-intensive to run in a multiuser environment.
# vmstat 2
kthr
memory
page
faults
cpu
----- ----------- ------------------------ ------------ ----------r b
avm
fre re pi po fr
sr cy in
sy cs us sy id wa
1 0 22478 1677
0
0
0
0
0
0 188 1380 157 57 32 0 10
1 0 22506 1609
0
0
0
0
0
0 214 1476 186 48 37 0 16
0 0 22498 1582
0
0
0
0
0
0 248 1470 226 55 36 0 9
2
0 22534
1465
0
0
0
© Copyright IBM Corp. 1997, 2002
0
0
0 238
903 239 77 23
0
0
99
2 0
2 0
3 0
2 1
2
1
1
22534 1445
22534 1426
22534 1410
22557 1365
0 22541
0 22524
0 22546
1356
1350
1293
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
209
189
255
383
1142
1220
1704
977
205
212
268
216
72
74
70
72
28 0 0
26 0 0
30 0 0
28 0 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 237 1418 209 63 33
0 241 1348 179 52 32
0 217 1473 180 51 35
0 4
0 16
0 14
This output shows the effect of introducing a program in a tight loop to a busy multiuser system. The first
three reports (the summary has been removed) show the system balanced at 50-55 percent user, 30-35
percent system, and 10-15 percent I/O wait. When the looping program begins, all available CPU cycles
are consumed. Because the looping program does no I/O, it can absorb all of the cycles previously
unused because of I/O wait. Worse, it represents a process that is always ready to take over the CPU
when a useful process relinquishes it. Because the looping program has a priority equal to that of all other
foreground processes, it will not necessarily have to give up the CPU when another process becomes
dispatchable. The program runs for about 10 seconds (five reports), and then the activity reported by the
vmstat command returns to a more normal pattern.
The CPU statistics can be somewhat distorted on systems with very high device-interrupt load. This
situation is due to the fact that the tool samples on timer interrupts. The timer is the lowest priority device
and therefore it can easily be preempted by other interrupts. To eliminate this distortion, operating system
versions later than AIX 4.3.3 use a different method to sample the timer.
Note: For SMP systems the us, sy, id and wa columns are only averages over the processors (the sar
command can report per-processor statistics). An I/O wait is distinguished from idle time only by the
state of a pending I/O. If there is any pending disk I/O, and the processor is not busy, then it is an
I/O wait time. AIX 4.3.3 and later contains an enhancement to the method used to compute the
percentage of CPU time spent waiting on disk I/O (wio time). See Wait I/O Time Reporting for more
details.
Optimum use would have the CPU working 100 percent of the time. This holds true in the case of a
single-user system with no need to share the CPU. Generally, if us + sy time is below 90 percent, a
single-user system is not considered CPU constrained. However, if us + sy time on a multiuser system
exceeds 80 percent, the processes may spend time waiting in the run queue. Response time and
throughput might suffer.
To check if the CPU is the bottleneck, consider the four cpu columns and the two kthr (kernel threads)
columns in the vmstat report. It may also be worthwhile looking at the faults column:
v cpu
Percentage breakdown of CPU time usage during the interval. The cpu columns are as follows:
– us
The us column shows the percent of CPU time spent in user mode. A UNIX process can execute in
either user mode or system (kernel) mode. When in user mode, a process executes within its
application code and does not require kernel resources to perform computations, manage memory,
or set variables.
– sy
The sy column details the percentage of time the CPU was executing a process in system mode.
This includes CPU resource consumed by kernel processes (kprocs) and others that need access to
kernel resources. If a process needs kernel resources, it must execute a system call and is thereby
switched to system mode to make that resource available. For example, reading or writing of a file
requires kernel resources to open the file, seek a specific location, and read or write data, unless
memory mapped files are used.
– id
100
Performance Management Guide
The id column shows the percentage of time which the CPU is idle, or waiting, without pending local
disk I/O. If there are no threads available for execution (the run queue is empty), the system
dispatches a thread called wait, which is also known as the idle kproc. On an SMP system, one
wait thread per processor can be dispatched. The report generated by the ps command (with the -k
or -g 0 option) identifies this as kproc or wait. If the ps report shows a high aggregate time for this
thread, it means there were significant periods of time when no other thread was ready to run or
waiting to be executed on the CPU. The system was therefore mostly idle and waiting for new tasks.
If there are no I/Os pending, all time charged to wait is classified as idle time. In operating system
version 4.3.2 and earlier, an access to remote disks (NFS-mounted disks) is treated as idle time
(with a small amount of sy time to execute the NFS requests) because there is no pending I/O
request to a local disk. With AIX 4.3.3 and later NFS goes through the buffer cache, and waits in
those routines are accounted for in the wa statistics.
– wa
The wa column details the percentage of time the CPU was idle with pending local disk I/O (in AIX
4.3.3 and later this is also true for NFS-mounted disks). If there is at least one outstanding I/O to a
disk when wait is running, the time is classified as waiting for I/O. Unless asynchronous I/O is being
used by the process, an I/O request to disk causes the calling process to block (or sleep) until the
request has been completed. Once an I/O request for a process completes, it is placed on the run
queue. If the I/Os were completing faster, more CPU time could be used.
A wa value over 25 percent could indicate that the disk subsystem might not be balanced properly, or
it might be the result of a disk-intensive workload.
For information on the change made to wa, see Wait I/O Time Reporting.
v kthr
Number of kernel threads in various queues averaged per second over the sampling interval. The kthr
columns are as follows:
– r
Average number of kernel threads that are runnable, which includes threads that are running and
threads that are waiting for the CPU. If this number is greater than the number of CPUs, there is at
least one thread waiting for a CPU and the more threads there are waiting for CPUs, the greater the
likelihood of a performance impact.
– b
Average number of kernel threads in the VMM wait queue per second. This includes threads that are
waiting on filesystem I/O or threads that have been suspended due to memory load control.
If processes are suspended due to memory load control, the blocked column (b) in the vmstat report
indicates the increase in the number of threads rather than the run queue.
– p
For vmstat -I The number of threads waiting on I/Os to raw devices per second.Threads waiting on
I/Os to filesystems would not be included here.
v faults
Information about process control, such as trap and interrupt rate. The faults columns are as follows:
– in
Number of device interrupts per second observed in the interval. Additional information can be found
in Assessing Disk Performance with the vmstat Command.
– sy
The number of system calls per second observed in the interval. Resources are available to user
processes through well-defined system calls. These calls instruct the kernel to perform operations for
the calling process and exchange data between the kernel and the process. Because workloads and
applications vary widely, and different calls perform different functions, it is impossible to define how
many system calls per-second are too many. But typically, when the sy column raises over 10000
calls per second on a uniprocessor, further investigations is called for (on an SMP system the
Chapter 7. Monitoring and Tuning CPU Use
101
number is 10000 calls per second per processor). One reason could be ″polling″ subroutines like the
select() subroutine. For this column, it is advisable to have a baseline measurement that gives a
count for a normal sy value.
– cs
Number of context switches per second observed in the interval. The physical CPU resource is
subdivided into logical time slices of 10 milliseconds each. Assuming a thread is scheduled for
execution, it will run until its time slice expires, until it is preempted, or until it voluntarily gives up
control of the CPU. When another thread is given control of the CPU, the context or working
environment of the previous thread must be saved and the context of the current thread must be
loaded. The operating system has a very efficient context switching procedure, so each switch is
inexpensive in terms of resources. Any significant increase in context switches, such as when cs is a
lot higher than the disk I/O and network packet rate, should be cause for further investigation.
The iostat Command
The iostat command is the fastest way to get a first impression, whether or not the system has a disk
I/O-bound performance problem (see Assessing Disk Performance with the iostat Command). The tool
also reports CPU statistics.
The following example shows a part of an iostat command output. The first stanza shows the summary
statistic since system startup.
# iostat -t 2 6
tty:
tin
0.0
0.0
0.0
0.0
0.0
0.0
tout
0.8
80.2
40.5
40.5
40.5
40.5
avg-cpu:
% user
8.4
4.5
7.0
9.0
7.5
10.0
% sys
2.6
3.0
4.0
2.5
1.0
3.5
% idle
88.5
92.1
89.0
88.5
91.5
80.5
% iowait
0.5
0.5
0.0
0.0
0.0
6.0
The CPU statistics columns (% user, % sys, % idle, and % iowait) provide a breakdown of CPU usage.
This information is also reported in the vmstat command output in the columns labeled us, sy, id, and wa.
For a detailed explanation for the values, see The vmstat Command. Also note the change made to
%iowait described in Wait I/O Time Reporting.
The sar Command
The sar command gathers statistical data about the system. Though it can be used to gather some useful
data regarding system performance, the sar command can increase the system load that can exacerbate
a pre-existing performance problem if the sampling frequency is high. But compared to the accounting
package, the sar command is less intrusive. The system maintains a series of system activity counters
which record various activities and provide the data that the sar command reports. The sar command
does not cause these counters to be updated or used; this is done automatically regardless of whether or
not the sar command runs. It merely extracts the data in the counters and saves it, based on the sampling
rate and number of samples specified to the sar command.
With its numerous options, the sar command provides queuing, paging, TTY, and many other statistics.
One important feature of the sar command is that it reports either systemwide (global among all
processors) CPU statistics (which are calculated as averages for values expressed as percentages, and
as sums otherwise), or it reports statistics for each individual processor. Therefore, this command is
particularly useful on SMP systems.
There are three situations to use the sar command:
Real-time sampling and display
To collect and display system statistic reports immediately, use the following command:
102
Performance Management Guide
# sar -u 2 5
AIX texmex 3 4 000691854C00
17:58:15
17:58:17
17:58:19
17:58:21
17:58:23
17:58:25
Average
01/27/00
%usr
43
35
36
21
85
%sys
9
17
22
17
12
%wio
1
3
20
0
3
%idle
46
45
23
63
0
44
15
5
35
This example is from a single user workstation and shows the CPU utilization.
Display previously captured data
The -o and -f options (write and read to/from user given data files) allow you to visualize the behavior of
your machine in two independent steps. This consumes less resources during the problem-reproduction
period. You can use a separate machine to analyze the data by transferring the file because the collected
binary file keeps all data the sar command needs.
# sar -o /tmp/sar.out 2 5 > /dev/null
The above command runs the sar command in the background, collects system activity data at 2-second
intervals for 5 intervals, and stores the (unformatted) sar data in the /tmp/sar.out file. The redirection of
standard output is used to avoid a screen output.
The following command extracts CPU information from the file and outputs a formatted report to standard
output:
# sar -f/tmp/sar.out
AIX texmex 3 4 000691854C00
18:10:18
18:10:20
18:10:22
18:10:24
18:10:26
18:10:28
Average
01/27/00
%usr
9
13
37
8
20
%sys
2
10
4
2
3
%wio
0
0
0
0
0
%idle
88
76
59
90
77
18
4
0
78
The captured binary data file keeps all information needed for the reports. Every possible sar report could
therefore be investigated. This also allows to display the processor-specific information of an SMP system
on a single processor system.
System activity accounting via cron daemon
The sar command calls a process named sadc to access system data. Two shell scripts (/usr/lib/sa/sa1
and /usr/lib/sa/sa2) are structured to be run by the cron daemon and provide daily statistics and reports.
Sample stanzas are included (but commented out) in the /var/spool/cron/crontabs/adm crontab file to
specify when the cron daemon should run the shell scripts.
The following lines show a modified crontab for the adm user. Only the comment characters for the data
collections were removed:
#=================================================================
#
SYSTEM ACTIVITY REPORTS
# 8am-5pm activity reports every 20 mins during weekdays.
# activity reports every an hour on Saturday and Sunday.
# 6pm-7am activity reports every an hour during weekdays.
# Daily summary prepared at 18:05.
#=================================================================
0 8-17 * * 1-5 /usr/lib/sa/sa1 1200 3 &
Chapter 7. Monitoring and Tuning CPU Use
103
0 * * * 0,6 /usr/lib/sa/sa1 &
0 18-7 * * 1-5 /usr/lib/sa/sa1 &
5 18 * * 1-5 /usr/lib/sa/sa2 -s 8:00 -e 18:01 -i 3600 -ubcwyaqvm &
#=================================================================
Collection of data in this manner is useful to characterize system usage over a period of time and to
determine peak usage hours.
Useful CPU Options: The most useful CPU-related options for the sar command are:
v sar -P
The -P option reports per-processor statistics for the specified processors. By specifying the ALL
keyword, statistics for each individual processor and an average for all processors is reported. Of the
flags which specify the statistics to be reported, only the -a, -c, -m, -u, and -w flags are meaningful with
the -P flag.
The following example shows the per-processor statistic while a CPU-bound program was running on
processor number 0:
# sar -P ALL 2 3
AIX rugby 3 4 00058033A100
01/27/00
17:30:50 cpu
17:30:52 0
1
2
3
17:30:54 0
1
2
3
17:30:56 0
1
2
3
-
%usr
8
0
0
0
2
12
0
0
0
3
11
0
0
0
3
%sys
92
4
1
0
24
88
3
1
0
23
89
3
0
0
23
%wio
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Average 0
1
2
3
-
10
0
0
0
3
90
4
1
0
24
0
0
0
0
0
%idle
0
96
99
100
74
0
97
99
100
74
0
97
100
100
74
0
96
99
100
74
The last line of every stanza, which starts with a dash (-) in the cpu column, is the average for all
processors. An average (-) line displays only if the -P ALL option is used. It is removed if processors
are specified. The last stanza, labeled with the word Average instead of a time stamp, keeps the
averages for the processor-specific rows over all stanzas.
The following example shows the vmstat output during this time:
# vmstat 2 5
kthr
memory
----- ----------r b
avm
fre
0 0 5636 16054
1 1 5733 15931
1 1 5733 15930
1 1 5733 15930
1 1 5733 15930
page
faults
------------------------ -----------re pi po fr
sr cy in
sy cs
0
0
0
0
0
0 116 266
5
0
0
0
0
0
0 476 50781 35
0
0
0
0
0
0 476 49437 27
0
0
0
0
0
0 473 48923 31
0
0
0
0
0
0 466 49383 27
cpu
-----------us sy id wa
0 1 99 0
2 27 70 0
2 24 74 0
3 23 74 0
3 23 74 0
The first numbered line is the summary since startup of the system. The second line reflects the start of
the sar command, and with the third row, the reports are comparable. The vmstat command can only
104
Performance Management Guide
display the average CPU utilization over all processors. This is comparable with the dashed (-) rows
from the CPU utilization output from the sar command.
v sar -u
This displays the CPU utilization. It is the default if no other flag is specified. It shows the same
information as the CPU statistics of the vmstat or iostat commands.
During the following example, a copy command was started:
# sar -u -P ALL 1 5
AIX rugby 3 4 00058033A100
13:33:42 cpu
13:33:43 0
1
2
3
13:33:44 0
1
2
3
13:33:45 0
1
2
3
13:33:46 0
1
2
3
13:33:47 0
1
2
3
Average
0
1
2
3
-
10/07/99
%usr
0
0
0
0
0
2
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
%sys
0
0
0
0
0
66
1
0
1
17
52
1
4
0
14
8
0
0
1
2
7
0
1
0
2
%wio
0
0
0
0
0
0
0
0
0
0
44
0
0
0
11
91
0
0
0
23
93
0
0
0
23
%idle
100
100
100
100
100
32
99
100
99
82
3
99
96
100
74
1
100
100
99
75
0
100
99
100
75
1
0
0
0
0
27
0
1
0
7
46
0
0
0
11
27
100
99
100
81
The cp command is working on processor number 0, and the three other processors are idle. This
reflects the change with operating system version 4.3.3 (see Wait I/O Time Reporting).
v sar -c
The -c option shows the system call rate.
# sar -c 1 3
19:28:25 scall/s sread/s swrit/s
19:28:26
134
36
1
19:28:27
46
34
1
19:28:28
46
34
1
Average
75
35
1
fork/s
0.00
0.00
0.00
0.00
exec/s
0.00
0.00
0.00
rchar/s wchar/s
2691306
1517
2716922
1531
2716922
1531
0.00 2708329
1527
While the vmstat command shows system call rates as well, the sar command can also show if these
system calls are read(), write(), fork(), exec(), and others. Pay particular attention to the fork/s
column. If this is high, then further investigation might be needed using the accounting utilities, the trace
command, or the tprof command.
v sar -q
The -q option shows the run-queue size and the swap-queue size.
Chapter 7. Monitoring and Tuning CPU Use
105
# sar -q 5 3
19:31:42 runq-sz %runocc swpq-sz %swpocc
19:31:47
1.0
100
1.0
100
19:31:52
2.0
100
1.0
100
19:31:57
1.0
100
1.0
100
Average
1.3
95
1.0
95
runq-sz
The average number of threads that are runnable per second and the percentage of time that
the run queue was occupied (the % field is subject to error).
swapq-sz
The average number of threads in the VMM wait queue and the % of time that the swap queue
was occupied. (The % field is subject to error.)
The -q option can indicate whether you have too many jobs running (runq-sz) or have a potential
paging bottleneck. In a highly transactional system, for example Enterprise Resource Planning (ERP),
the run queue can be in the hundreds, because each transaction uses small amounts of CPU time. If
paging is the problem, run the vmstat command. High I/O wait indicates that there is significant
competing disk activity or excessive paging due to insufficient memory.
The xmperf Program
Using the xmperf program displays CPU use as a moving skyline chart. The xmperf program is described
in detail in the Performance Toolbox Version 2 and 3 for AIX: Guide and Reference.
Using the time Command to Measure CPU Use
Use the time command to understand the performance characteristics of a single program and its
synchronous children. It reports the real time, that is the elapsed time from beginning to end of the
program. It also reports the amount of CPU time used by the program. The CPU time is divided into user
and sys. The user value is the time used by the program itself and any library subroutines it calls. The sys
value is the time used by system calls invoked by the program (directly or indirectly).
The sum of user + sys is the total direct CPU cost of executing the program. This does not include the
CPU costs of parts of the kernel that can be said to run on behalf of the program, but which do not
actually run on its thread. For example, the cost of stealing page frames to replace the page frames taken
from the free list when the program started is not reported as part of the program’s CPU consumption.
On a uniprocessor, the difference between the real time and the total CPU time, that is:
real - (user + sys)
is the sum of all of the factors that can delay the program, plus the program’s own unattributed costs. On
an SMP, an approximation would be as follows:
real * number_of_processors - (user + sys)
In approximately the order of diminishing size, the factors can be:
v I/O required to bring in the program’s text and data
v I/O required to acquire real memory for the program’s use
v CPU time consumed by other programs
v CPU time consumed by the operating system
In the following example, the program used in the preceding section has been compiled with -O3 to make
it run more quickly. There is very little difference between the real (wall-clock) time required to run the
106
Performance Management Guide
program and the sum of its user and system CPU times. The program is getting all the time it wants,
probably at the expense of other programs in the system.
# time looper
real
0m3.58s
user
0m3.16s
sys
0m0.04s
In the next example, we run the program at a less favorable priority by adding 10 to its nice value. It takes
almost twice as long to run, but other programs are also getting a chance to do their work:
# time nice -n 10 looper
real
0m6.54s
user
0m3.17s
sys
0m0.03s
Note that we placed the nice command within the time command, rather than the reverse. If we had
entered
# nice -n 10 time looper
we would have gotten a different time command (/usr/bin/time) with a lower-precision report, rather than
the version of the time command we have been using, which is built into the ksh shell. If the time
command comes first, you get the built-in version, unless you specify the fully qualified name of
/usr/bin/time. If the time command is invoked from another command, you get /usr/bin/time.
time and timex Cautions
Take several considerations into account when you use either the time or the timex command:
v The use of the /usr/bin/time and /usr/bin/timex commands is not recommended. When possible, use
the time subcommand of the Korn or C shell.
v The timex -s command uses the sar command to acquire additional statistics. Because the sar
command is intrusive, the timex -s command is also. Especially for brief runs, the data reported by the
timex -s command may not precisely reflect the behavior of a program in an unmonitored system.
v Because of the length of the system clock tick (10 milliseconds) and the rules used by the scheduler in
attributing CPU time use to threads, the results of the time command are not completely deterministic.
Because the time is sampled, there is a certain amount of unavoidable variation between successive
runs. This variation is in terms of clock ticks. The shorter the run time of the program, the larger the
variation as a percentage of the reported result (see Accessing the Processor Timer).
v Use of the time or timex command (whether from /usr/bin or through the built-in shell time function) to
measure the user or system time of a sequence of commands connected by pipes, entered on the
command line, is not recommended. One potential problem is that syntax oversights can cause the time
command to measure only one of the commands, without any indication of a user error. The syntax is
technically correct; it just does not produce the answer that the user intended.
v Although the time command syntax did not change, its output takes on a new meaning in an SMP
environment:
On an SMP the real, or elapsed time may be smaller than the user time of a process. The user time is
now the sum of all the times spent by the threads or the process on all processors.
If a process has four threads, running it on a uniprocessor (UP) system shows that the real time is
greater than the user time:
# time 4threadedprog
real
0m11.70s
user
0m11.09s
sys
0m0.08s
Running it on a 4-way SMP system could show that the real time is only about 1/4 of the user time. The
following output shows that the multithreaded process distributed its workload on several processors
and improved its real execution time. The throughput of the system was therefore increased.
Chapter 7. Monitoring and Tuning CPU Use
107
# time 4threadedprog
real
0m3.40s
user
0m9.81s
sys
0m0.09s
Identifying CPU-Intensive Programs
To locate the processes dominating CPU usage, there are two standard tools, the ps command and the
acctcom command. Another tool to use is the topas monitor, which is described in Using the topas
Monitor.
Using the ps Command
The ps command is a flexible tool for identifying the programs that are running on the system and the
resources they are using. It displays statistics and status information about processes on the system, such
as process or thread ID, I/O activity, CPU and memory utilization. In this chapter, we discuss only the
options and output fields that are relevant for CPU.
Three of the possible ps output columns report CPU use, each in a different way.
Column
Value Is:
C
Recently used CPU time for the process (in units of clock ticks).
TIME
Total CPU time used by the process since it started (in units of minutes and seconds).
%CPU Total CPU time used by the process since it started, divided by the elapsed time since the process
started. This is a measure of the CPU dependence of the program.
CPU Intensive
The following shell script:
# ps -ef | egrep -v "STIME|$LOGNAME" | sort +3 -r | head -n 15
is a tool for focusing on the highest recently used CPU-intensive user processes in the system (the header
line has been reinserted for clarity):
UID
mary
root
root
root
root
root
bick
bick
luc
PID
45742
52122
4250
38812
27036
47418
37652
43538
60062
PPID
C
STIME
TTY TIME CMD
54702 120 15:19:05 pts/29 0:02 ./looper
1 11 15:32:33 pts/31 58:39 xhogger
1
3 15:32:33 pts/31 26:03 xmconsole allcon
4250
1 15:32:34 pts/31 8:58 xmconstats 0 3 30
6864
1 15:18:35
- 0:00 rlogind
25926
0 17:04:26
- 0:00 coelogin <d29dbms:0>
43538
0 16:58:40 pts/4 0:00 /bin/ksh
1
0 16:58:38
- 0:07 aixterm
27036
0 15:18:35 pts/18 0:00 -ksh
Recent CPU use is the fourth column (C). The looping program’s process easily heads the list. Observe
that the C value may understate the looping process’ CPU usage, because the scheduler stops counting at
120.
CPU Time Ratio
The ps command, run periodically, displays the CPU time under the TIME column and the ratio of CPU
time to real time under the %CPU column. Look for the processes that dominate usage. The au and v
options give similar information on user processes. The options aux and vg display both user and system
processes.
The following example is taken from a four-way SMP system:
# ps au
USER
PID %CPU %MEM
root
19048 24.6 0.0
108
SZ
28
RSS
44
Performance Management Guide
TTY STAT
STIME TIME COMMAND
pts/1 A
13:53:00 2:16 /tmp/cpubound
root
root
root
root
root
root
root
19388
15348
20418
16178
16780
18516
15746
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
372
372
368
292
364
360
212
460
460
452
364
392
412
268
pts/1
pts/4
pts/3
0
pts/2
pts/0
pts/1
A
A
A
A
A
A
A
Feb 20
Feb 20
Feb 20
Feb 19
Feb 19
Feb 20
13:55:18
0:02
0:01
0:01
0:00
0:00
0:00
0:00
-ksh
-ksh
-ksh
/usr/sbin/getty
-ksh
-ksh
ps au
The %CPU is the percentage of CPU time that has been allocated to that process since the process was
started. It is calculated as follows:
(process CPU time / process duration) * 100
Imagine two processes: The first starts and runs five seconds, but does not finish; then the second starts
and runs five seconds but does not finish. The ps command would now show 50 percent %CPU for the first
process (five seconds CPU for 10 seconds of elapsed time) and 100 percent for the second (five seconds
CPU for five seconds of elapsed time).
On an SMP, this value is divided by the number of available CPUs on the system. Looking back at the
previous example, this is the reason why the %CPU value for the cpubound process will never exceed 25,
because the example is run on a four-way processor system. The cpubound process uses 100 percent of
one processor, but the %CPU value is divided by the number of available CPUs.
The THREAD Option
The ps command can display threads and the CPUs that threads or processes are bound to by using the
ps -mo THREAD command. The following is an example:
# ps
USER
root
-
-mo THREAD
PID
PPID TID
ST CP PRI SC WCHAN F
TT
BND COMMAND
20918 20660 A 0 60 1 240001 pts/1 -ksh
20005 S 0 60 1 400
-
The TID column shows the thread ID, the BND column shows processes and threads bound to a processor.
It is normal to see a process named kproc (PID of 516 in operating system version 4) using CPU time.
When there are no threads that can be run during a time slice, the scheduler assigns the CPU time for
that time slice to this kernel process (kproc), which is known as the idle or wait kproc. SMP systems will
have an idle kproc for each processor. With operating system versions later than AIX 4.3.3 the name
shown in the output is wait.
For complete details about the ps command, see the AIX 5L Version 5.2 Commands Reference.
Using the acctcom Command
The acctcom command displays historical data on CPU usage if the accounting system is activated.
Starting the accounting system puts a measurable overhead on the system. Therefore, activate accounting
only if absolutely needed. To activate the accounting system, do the following:
1. Create an empty accounting file:
# touch acctfile
2. Turn on accounting:
# /usr/sbin/acct/accton acctfile
3. Allow accounting to run for a while and then turn off accounting:
# /usr/sbin/acct/accton
4. Display what accounting captured, as follows:
# /usr/sbin/acct/acctcom acctfile
COMMAND
START
NAME
USER
TTYNAME TIME
#accton
root
pts/2
19:57:18
#ps
root
pts/2
19:57:19
END
TIME
19:57:18
19:57:19
REAL
(SECS)
0.02
0.19
CPU
(SECS)
0.02
0.17
MEAN
SIZE(K)
184.00
35.00
Chapter 7. Monitoring and Tuning CPU Use
109
#ls
#ps
#accton
#who
root
root
root
root
pts/2
pts/2
pts/2
pts/2
19:57:20
19:57:22
20:04:17
20:04:19
19:57:20
19:57:22
20:04:17
20:04:19
0.09
0.19
0.00
0.02
0.03
0.17
0.00
0.02
109.00
34.00
0.00
0.00
If you reuse the same file, you can see when the newer processes were started by looking for the accton
process (this was the process used to turn off accounting the first time).
Using the tprof Program to Analyze Programs for CPU Use
The typical program execution is a variable mixture of application code, library subroutines, and kernel
services. Frequently, a program that has not yet been tuned is found to expend most of its CPU cycles in
a few statements or subroutines. Often these hot spots are a surprise to the programmer. They often can
be considered performance problems. Use the tprof command to pinpoint any hot spots (for additional
information see The tprof Command). The tprof command can profile any program produced by one of the
compilers: C, C++, and FORTRAN.
To determine whether the tprof program is installed and available, run the following command:
# lslpp -lI bos.perf.tools
The raw data for the tprof program is obtained through the trace facility (see Analyzing Performance with
the Trace Facility). When a program is profiled, the trace facility is activated and instructed to collect data
from the trace hook (hook ID 234) that records the contents of the Instruction Address Register when a
system-clock interrupt occurs (100 times a second per processor). Several other trace hooks are also
activated to allow the tprof program to track process and dispatch activity. The trace records are not
written to a disk file; they are written to a pipe that is read by a program that builds a table of the unique
program addresses that have been encountered and the number of times each one occurred. When the
workload being profiled is complete, the table of addresses and their occurrence counts is written to disk.
The data-reduction component of the tprof program then correlates the instruction addresses that were
encountered with the ranges of addresses occupied by the various programs and reports the distribution of
address occurrences (ticks) across the programs involved in the workload.
The distribution of ticks is roughly proportional to the CPU time spent in each program (10 milliseconds per
tick). After the high-use programs have been identified, the programmer can take action to restructure their
hot spots or minimize their use.
A tprof Example for releases prior to AIX 5.2
The following C program initializes each byte of a large array of integers to 0x01, increments each integer
by a random constant, and prints out a randomly selected integer. The program is representative of
programs that process large arrays.
/* Array Incrementer -- Version 1 */
#include <stdlib.h>
#define Asize 1024
#define RowDim InnerIndex
#define ColDim OuterIndex
main()
{
int Increment;
int OuterIndex;
int InnerIndex;
int big [Asize][Asize];
/* initialize every byte of the array to 0x01 */
for(OuterIndex=0; OuterIndex<Asize; OuterIndex++)
{
for (InnerIndex=0; InnerIndex<Asize; InnerIndex++)
big[RowDim][ColDim] = 0x01010101;
}
Increment = rand();
/* increment every element in the array */
110
Performance Management Guide
}
for(OuterIndex=0; OuterIndex<Asize; OuterIndex++)
{
for (InnerIndex=0; InnerIndex<Asize; InnerIndex++)
{
big[RowDim][ColDim] += Increment;
if (big[RowDim][ColDim] < 0)
printf("Negative number. %d\n",big[RowDim][ColDim]);
}
}
printf("Version 1 Check Num: %d\n",
big[rand()%Asize][rand()%Asize]);
return(0);
The program was compiled with the following command:
# xlc -g version1.c -o version1
The -g parameter causes the C compiler to generate the object module with symbolic debugging
information for use by the tprof program. Although the tprof program can profile optimized modules, the
-O parameter has been omitted to make the line numbers that the tprof program uses more precise.
When the C compiler is optimizing, it often does enough rearrangement of code to make the output of the
tprof program harder to interpret. On the test system, this program runs in about 5.97 seconds of elapsed
time, of which more than 5.9 seconds is user CPU time. The program clearly meets its objective of being
CPU-limited.
We can profile the program with the following command (include the -m option on operating ystem
versions later than AIX 4.3.3):
# tprof -p version1 -x version1
A file called __version1.all (shown below) is created. It reports how many CPU ticks each of the programs
involved in the execution consumed.
Process
PID
=======
===
version1
30480
ksh
32582
/etc/init
1
/etc/syncd
3854
tprof
5038
rlogind
11344
PID.771
770
tprof
11940
tprof
11950
tprof
13986
ksh
16048
=======
===
Total
Process
FREQ
=======
===
version1
1
ksh
2
/etc/init
1
/etc/syncd
1
tprof
4
rlogind
1
PID.771
1
=======
===
Total
11
Total Ticks For version1(
Subroutine
Ticks
============= ======
.main
763
TID
===
30481
32583
459
4631
5019
15115
771
11941
11951
15115
7181
===
Total
Kernel
=====
======
793
30
8
8
6
0
5
5
4
2
2
2
1
1
1
1
1
1
1
1
1
1
=====
======
823
52
Total
Kernel
User
=====
======
====
793
30
763
9
9
0
6
0
6
5
5
0
7
5
2
2
2
0
1
1
0
=====
======
====
823
52
771
USER) =
763
%
Source
======
=======
92.7
version1.c
User
====
763
0
6
0
2
0
0
0
0
0
0
====
771
Shared
======
0
0
0
0
0
0
0
======
0
Shared
======
0
0
0
0
0
0
0
0
0
0
0
======
0
Other
=====
0
0
0
0
0
0
0
=====
0
Other
=====
0
0
0
0
0
0
0
0
0
0
0
=====
0
Address Bytes
======= =====
632
560
Chapter 7. Monitoring and Tuning CPU Use
111
The first section of the tprof report shows the number of ticks consumed by, or on behalf of, each
process. The program version1 used 763 ticks itself, and 30 ticks occurred in the kernel on behalf of
version1’s process. Two processes running the Bourne shell were involved in the execution of version1.
Four processes were running tprof-related code. The init process, the sync daemon, an rlogin process,
and one other process accounted for 14 ticks.
Remember that the program associated with a given numerical process ID changes with each exec()
subroutine call. If one application program uses the exec() subroutine to execute another, both program
names will appear in the tprof output associated with the same process ID.
The second section of the report summarizes the results by program, regardless of process ID. It shows
the number (FREQ) of different processes that ran each program at some point.
The third section breaks down the user ticks associated with the executable program being profiled. It
reports the number of ticks used by each function in the executable program, and the percentage of the
total run’s CPU ticks (823) that each function’s ticks represent.
Up to this point, none of the tprof processing has required access to the specially compiled version of the
program. You could have done the preceding analysis on a program for which you did not have access to
the source code.
It is clear from this report that the main CPU consumption (92.7 percent) is in the program itself, not in the
kernel nor in library subroutines that the program uses. You must examine the program itself more closely.
Because you compiled version1.c with the -g option, the object file contains information that relates
offsets in the program text to lines of source code. Consequently, the tprof program created an annotated
version of the source file version1.c, called __t.version1.c, based on the offsets and line number
information in the object module. The first column is the line number. The second column is the number of
times the trace hook reported that the timer interrupt occurred while the system was executing one of the
instructions associated with that line.
Ticks Profile for main in version1.c
Line
Ticks
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
34
40
261
70
69
50
239
29
30
31
32
33
34
-
Source
for(OuterIndex=0; OuterIndex<Asize; OuterIndex++)
{
for (InnerIndex=0; InnerIndex<Asize; InnerIndex++)
big[RowDim][ColDim] = 0x01010101;
}
Increment = rand();
}
/* increment every element in the array */
for(OuterIndex=0; OuterIndex<Asize; OuterIndex++)
{
for (InnerIndex=0; InnerIndex<Asize; InnerIndex++)
{
big[RowDim][ColDim] += Increment;
if (big[RowDim][ColDim] < 0)
printf("Negative number.%d\n",
big[RowDim][ColDim]);
}
}
printf("Version 1 Check Num: %d\n",
big[rand()%Asize][rand()%Asize]);
return(0);
763 Total Ticks for main in version1.c
112
Performance Management Guide
This file shows that the largest numbers of ticks are associated with accessing elements of the array big,
so you should be able to enhance performance significantly by concentrating on the inner for loops. The
first (initialization) for loop is a case of inefficient programming, because it initializes the array one element
at a time. If you were setting the array to 0, use the bzero() subroutine. Because you are setting each
byte to a specific character, use the memset() subroutine to replace the first for loop. (The efficient
bzero() and memset() functions, like the str*() functions, are written in assembler language and use
hardware instructions that have no direct equivalent in the C language.)
You must access the array one element at a time to increment the values, but ensure that the pattern of
memory reference is to consecutive addresses, to maximize cache use. In this case, you have the row
dimension changing faster than the column dimension. Because C arrays are arranged in row-major order,
you are skipping over a complete row with each successive memory reference. Because the rows are
1024 integers long (4096 bytes), you are changing pages on every reference. The size of the array greatly
exceeds both the data cache and data Translation Lookaside Buffer (TLB) capacities, so you have written
a program for maximum cache and TLB thrashing. To fix this problem, transpose the two #define
statements to reverse the values of RowDim and ColDim.
The unoptimized form of the resulting program (version2.c) consumes about 2.7 CPU seconds, compared
with 7.9 CPU seconds for program version1.
The following file, __t.version2.c, is the result of a tprof run against the unoptimized form:
Ticks Profile for main in version2.c
Line
Ticks
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
60
67
60
43
-
Source
memset(big,0x01,sizeof(big));
Increment = rand();
}
/* increment in memory order */
for(OuterIndex=0; OuterIndex<Asize; OuterIndex++)
{
for (InnerIndex=0; InnerIndex<Asize; InnerIndex++)
{
big[RowDim][ColDim] += Increment;
if (big[RowDim][ColDim] < 0)
printf("Negative number. %d\n",big[RowDim][ColDim]);
}
}
printf("Version 2 Check Num: %d\n",
big[rand()%Asize][rand()%Asize]);
return(0);
230 Total Ticks for main in version2.c
By knowing its CPU use pattern, you have improved the CPU speed of this program by a factor of almost
three, for the unoptimized case. When you compile version1.c and version2.c with optimization and
compare their performance, the ″before and after″ improvement due to the changes is a factor of 7.
In many cases, most of a program’s CPU use will occur in the library subroutines it uses rather than in the
program itself. If you take version2.c and remove the conditional test on line 24 and the printf() entry on
line 28, to create a version3.c that reads as follows:
#include <string.h>
#include <stdlib.h>
#define Asize 256
#define RowDim OuterIndex
#define ColDim InnerIndex
main()
{
int Increment;
Chapter 7. Monitoring and Tuning CPU Use
113
int OuterIndex;
int InnerIndex;
int big [Asize][Asize];
}
/* Initialize every byte to 0x01 */
memset(big,0x01,sizeof(big));
Increment = rand();
/* increment in memory order */
for(OuterIndex=0; OuterIndex<Asize; OuterIndex++)
{
for (InnerIndex=0; InnerIndex<Asize; InnerIndex++)
{
big[RowDim][ColDim] += Increment;
printf("RowDim=%d, ColDim=%d, Number=%d\n",
RowDim, ColDim, big[RowDim][ColDim]);
}
}
return(0);
the execution time becomes dominated by the printf() statement. The command:
# tprof -v -s -k -p version3 -x version3 >/dev/null
produces a __version3.all that includes profiling data for the kernel and the shared subroutine library
libc.a (the only shared library this program uses):
Process
=======
version3
ksh
tprof
tprof
tprof
ksh
=======
Total
PID
===
28372
27348
15986
7784
12904
13940
===
TID
===
28373
27349
19785
8785
13657
13755
===
Total
=====
818
5
3
1
1
1
=====
829
Kernel
======
30
5
1
1
1
1
======
39
User
====
19
0
2
0
0
0
====
21
Shared
======
769
0
0
0
0
0
======
769
Process
=======
version3
ksh
tprof
=======
Total
FREQ
===
1
2
3
===
6
Total
=====
818
6
5
=====
829
Kernel
======
30
6
3
======
39
User
====
19
0
2
====
21
Shared
======
769
0
0
======
769
Other
=====
0
0
0
=====
0
Total Ticks For version3(
Subroutine
=============
.main
.printf
Ticks
======
11
8
Total Ticks For version3(
Subroutine
=============
.sc_flih
.i_enable
.vmcopyin
.xix_setattr
.isreadonly
.lockl
.v_pagein
.curtime
.trchook
.vmvcs
114
Ticks
======
7
5
3
2
2
2
1
1
1
1
USER) =
%
======
1.3
1.0
KERNEL) =
%
======
0.8
0.6
0.4
0.2
0.2
0.2
0.1
0.1
0.1
0.1
Performance Management Guide
19
Source
=======
version3.c
glink.s
Address Bytes
======= =====
632
320
1112
36
30
Source
=======
low.s
low.s
vmmove.c
xix_sattr.c
disubs.c
lockl.s
v_getsubs1.c
clock.s
noname
vmvcs.s
Address Bytes
======= =====
13832
1244
21760
256
414280
668
819368
672
689016
60
29300
208
372288
1044
27656
76
48168
856
29744
2304
Other
=====
0
0
0
0
0
0
=====
0
.spec_rdwr
.rdwr
.imark
.nodev
.ld_findfp
1
1
1
1
1
0.1
0.1
0.1
0.1
0.1
spec_vnops.c
rdwr.c
isubs.c
devsw_pin.c
ld_libld.c
Total Ticks For version3( SH-LIBs) =
Shared Object
=============
libc.a/shr.o
Ticks
======
769
Profile: /usr/lib/libc.a
%
======
92.0
240
492
184
32
240
769
Source
=======
/usr/lib
Address Bytes
======= =====
794624 724772
shr.o
Total Ticks For version3(/usr/lib/libc.a) =
Subroutine
=============
._doprnt
.fwrite
.strchr
.printf
._moveeq
.strlen
.isatty
._xwrite
.__ioctl
629596
658460
672024
135864
736084
Ticks
======
476
205
41
18
16
10
1
1
1
%
======
56.9
24.5
4.9
2.2
1.9
1.2
0.1
0.1
0.1
769
Source
=======
doprnt.c
fwrite.c
strchr.s
printf.c
memcmp.s
strerror.c
isatty.c
flsbuf.c
ioctl.c
Address Bytes
======= =====
36616
7052
50748
744
31896
196
313796
144
36192
184
46800
124
62932
112
4240
280
57576
240
This report confirms that most of the ticks are being used by the shared libraries (libc.a in this case). The
profile of libc.a shows that most of those ticks are being consumed by the _doprnt() subroutine.
The _doprnt() subroutine is the processing module for the printf(), sprintf(), and other subroutines. With
a simple change, you have increased the run time from 2.7 seconds to 8.6 seconds, and the formatted
printing now consumes about 60 percent of the CPU time. This illustrates why formatting should be used
judiciously. The performance of the _doprnt() subroutine is also affected by the locale. See Appendix E.
National Language Support: Locale versus Speed. These tests were run in the C locale, which is the most
efficient.
A tprof Example for releases starting with AIX 5.2
The tprof command has been completely rewritten for AIX 5.2 such that now it is much faster than before
and provides more functionality. The syntax is different than the previous tprof, and you can view the
complete documentation of tprof in AIX 5L Version 5.2 Commands Reference.
The following is an example of how to collect a CPU tick profile of the version1 program using the new
tprof command, executed on a 4–way SMP system. Since this machine is a very fast running system, the
command finished in less than a second. To make this program run longer, the array size was changed to
4096 instead of 1024, which is the value of the Asize variable in version1.c:
# tprof -z -u -p version1 -x version1
A file called version1.prof (shown below) is created in the current directory which reports how many CPU
ticks for each of the programs running on the system while version1 was running.
Process
=======
wait
./version1
/usr/bin/tprof
/etc/syncd
/usr/bin/sh
swapper
/usr/bin/trcstop
rmcd
Freq
====
4
1
2
1
2
1
1
1
Total
=====
5810
1672
15
2
2
1
1
1
Kernel
======
5810
35
13
2
2
1
1
1
User
====
0
1637
0
0
0
0
0
0
Shared
======
0
0
2
0
0
0
0
0
Other
=====
0
0
0
0
0
0
0
0
Chapter 7. Monitoring and Tuning CPU Use
115
=======
Total
===
13
Process
PID
=======
===
wait
16392
wait
12294
wait
20490
./version1
245974
wait
8196
/usr/bin/tprof
291002
/usr/bin/tprof
274580
/etc/syncd
73824
/usr/bin/sh
245974
/usr/bin/sh
245976
/usr/bin/trcstop
245976
swapper
0
rmcd
155876
=======
===
Total
=====
7504
TID
===
16393
12295
20491
606263
8197
643291
610467
110691
606263
606265
606263
3
348337
===
Total Samples = 7504
======
5865
Total
=====
1874
1873
1860
1672
203
13
2
2
1
1
1
1
1
=====
7504
====
1637
Kernel
======
1874
1873
1860
35
203
13
0
2
1
1
1
1
1
======
5865
======
2
User
====
0
0
0
1637
0
0
0
0
0
0
0
0
0
====
1637
Shared
======
0
0
0
0
0
0
2
0
0
0
0
0
0
======
2
=====
0
Other
=====
0
0
0
0
0
0
0
0
0
0
0
0
0
=====
0
Total Elapsed Time = 18.76s
Profile: ./version1
Total Ticks For All Processes (./version1) = 1637
Subroutine
=============
.main
Ticks
======
1637
%
Source
======
=======
21.82 version1.c
Address Bytes
======= =====
350
536
Profile: ./version1
Total Ticks For ./version1[245974] (./version1) = 1637
Subroutine
=============
.main
Ticks
======
1637
%
Source
======
=======
21.82 version1.c
Address Bytes
======= =====
350
536
The first section of the report summarizes the results by program, regardless of the process ID. It shows
the number of different processes (Freq) that ran each program at some point.
The second section of the tprof report shows the number of ticks consumed by, or on behalf of, each
process. The program version1 used 1637 ticks itself and 35 ticks occurred in the kernel on behalf of
theversion1 process.
The third section breaks down the user ticks associated with the executable program being profiled. It
reports the number of ticks used by each function in the executable program, and the percentage of the
total run’s CPU ticks (7504) that each function’s ticks represent. Since this system’s CPUs were mostly
idle, most of the 7504 ticks are idle ticks. To see what percentage of the busy time this program took,
subtract the wait thread’s CPU ticks (these are the idle CPU ticks) from the total and then subtract that
from the total number of ticks. So, 7504–5810 gives us 1694. This is the total number of ticks for actual
work done on the system. If we divide the version1 program’s user ticks (1637) by 1694, we find that its
percentage of system busy ticks is 1637/1694*100, which is 96.6%.
Offline Processing with the tprof Command prior to AIX 5.2
The -i Trace_File flag allows for offline processing by the tprof command of trace data files created by the
system trace command. The -n flag allows you to specify a Gennames_File to be used when processing
an offline file. These flags are useful when it is necessary to postprocess a trace file from a remote
machine or perform the trace data collection at one time and postprocess it at another time. In this case -n
116
Performance Management Guide
with a Gennames_File must be used from the machine that the trace came from. The flags are also useful
when system loading is high and trace hooks are being missed by the tprof command. The offline option
relieves this problem.
Trace hooks relevant to the tprof command must be collected by the trace command and are specified by
the trace -j flag. The gennames command is then executed to collect additional information for the tprof
command. After the trace and gennames Gennames_File commands have executed, the trcrpt -r
command must be executed on the trace logfile and redirected to another file. At this point an adjusted
trace logfile and a Gennames_File is input to the tprof command.
For example:
#
#
#
#
#
#
trace -af -T 1000000 -L 10000000 -o trace.out -j 000,001,002,003,005,006,234,106,10C,134,139,00A,465
workload
trcoff
gennames > gennames.out
trcstop
trcrpt -r
trace.out > trace.rpt
Next run the tprof command with at least the -i and -n flags, as follows:
# tprof -i trace.rpt -n gennames.out -s -k -e
On systems with many CPUs, it is better to run the trace and trcrpt commands with the -C all flag (see
Formatting a Report from trace -C Output).
Offline Processing with the tprof Command starting with AIX 5.2
The new tprof command can do offline processing of trace files, but it requires that filenames be specified
with a rootstring name. This can be whatever you want to call it. Also, there are certain suffixes required
for the input files that tprof will use. For example, the trace binary file should end in .trc. The trace binary
file does not have to be post-processed with trcrpt -r any more. Also, instead of using gennames output,
you need to collect gensyms output and put this in a file called rootstring.syms.
Let us call our rootstring trace1. Then to collect a trace, we can trace using all of the hooks or at least the
following hooks:
#
#
#
#
#
#
trace -af -T 1000000 -L 10000000 -o trace1.trc -j 000,001,002,003,005,006,234,106,10C,134,139,00A,465
workload
trcoff
gensyms > trace1.syms
trcstop
trcrpt -r trace1 -k -u -s -z
This creates a trace1.prof file which gives you a CPU profile of the system while trace was running.
Using the pprof Command to Measure CPU usage of Kernel Threads
The pprof command reports CPU usage on all kernel threads running within an interval using the trace
utility. The raw process information is saved to pprof.flow and five reports are generated. If no flags are
specified, all reports are generated.
To determine whether the pprof program (available with AIX 4.3.3 and later) is installed and available, run
the following command:
# lslpp -lI fileset_name
where fileset_name is perfagent.tools in AIX 4.3.3 or bos.perf.tools in AIX 5 or later.
The types of reports are as follows:
Chapter 7. Monitoring and Tuning CPU Use
117
pprof.cpu
Lists all kernel level threads sorted by actual CPU time. Contains: Process Name, Process ID,
Parent Process ID, Process State at Beginning and End, Thread ID, Parent Thread ID, Actual
CPU Time, Start Time, Stop Time, Stop - Start.
pprof.famcpu
Lists the information for all families (processes with a common ancestor). The Process Name and
Process ID for the family is not necessarily the ancestor. Contains: Start Time, Process Name,
Process ID, Number of Threads, Total CPU Time.
pprof.famind
Lists all processes grouped by families (processes with a common ancestor). Child process names
are indented with respect to the parent. Contains: Start Time, Stop Time, Actual CPU Time,
Process ID, Parent Process ID, Thread ID, Parent Thread ID, Process State at Beginning and
End, Level, Process Name.
pprof.namecpu
Lists information about each type of kernel thread (all executable with the same name). Contains:
Process Name, Number of Threads, CPU Time, % of Total CPU Time.
pprof.start
Lists all kernel threads sorted by start time that were dispatched during the the pprof command
interval. Contains: Process Name, Process ID, Parent Process ID, Process State Beginning and
End, Thread ID, Parent Thread ID, Actual CPU Time, Start Time, Stop Time, Stop - Start.
Following is a sample pprof.namecpu file that was generated by running the tthreads32 program, which
forks off four threads, which in turn each fork off a process of their own. These processes then execute
several ksh and sleep programs:
Pprof
Sorted
PROCESS
by
CPU
NAME
Report
Time
From: Thu Oct 19 17:53:07 2000
To:
Thu Oct 19 17:53:22 2000
Pname #ofThreads CPU_Time
%
======== ========== ======== ========
tthreads32
13
0.116
37.935
sh
8
0.092
30.087
Idle
2
0.055
17.987
ksh
12
0.026
8.503
trace
3
0.007
2.289
java
3
0.006
1.962
kproc
5
0.004
1.308
xmservd
1
0.000
0.000
trcstop
1
0.000
0.000
swapper
1
0.000
0.000
gil
1
0.000
0.000
ls
4
0.000
0.000
sleep
9
0.000
0.000
ps
4
0.000
0.000
syslogd
1
0.000
0.000
nfsd
2
0.000
0.000
========== ======== ========
70
0.306
100.000
The corresponding pprof.cpu is as follows:
Pprof CPU Report
Sorted by Actual CPU Time
From: Thu Oct 19 17:53:07 2000
118
Performance Management Guide
To:
Thu Oct 19 17:53:22 2000
E = Exec’d
F = Forked
X = Exited
A = Alive (when traced started or stopped)
C = Thread Created
:
:
:
Pname
=====
Idle
tthreads32
sh
sh
sh
ksh
tthreads32
tthreads32
tthreads32
ksh
tthreads32
tthreads32
tthreads32
tthreads32
tthreads32
trace
kproc
Idle
java
java
trace
sh
trace
ksh
kproc
ps
tthreads32
sh
ps
sh
tthreads32
ls
PID
=====
774
5490
11396
14106
13792
5490
5490
5490
14506
11982
13792
5490
11396
5490
14106
5488
1548
516
11612
11612
12544
14506
12544
4930
6478
14108
13794
13794
13794
14108
14108
13792
PPID
=====
0
11982
5490
5490
5490
11982
11982
11982
5490
13258
5490
11982
5490
11982
5490
11982
0
0
11106
11106
5488
5490
5488
2678
0
5490
5490
5490
5490
5490
5490
5490
BE
===
AA
EX
EE
EE
EE
FE
CX
CX
FE
AA
FE
CX
FE
CX
FE
AX
AA
AA
AA
AA
AA
EE
CA
AA
AA
EX
FE
EE
EX
EE
FE
EX
TID
=====
775
18161
21917
16999
20777
18161
5093
18179
17239
22435
20777
18867
21917
10133
16999
18159
2065
517
14965
14707
20507
17239
19297
5963
3133
17001
20779
20779
20779
17001
17001
20777
PTID ACC_time STT_time STP_time STP-STT
===== ======== ======== ======== ========
0
0.052
0.000
0.154
0.154
22435
0.040
0.027
0.154
0.126
5093
0.035
0.082
0.154
0.072
18867
0.028
0.111
0.154
0.043
18179
0.028
0.086
0.154
0.068
22435
0.016
0.010
0.027
0.017
18161
0.011
0.056
0.154
0.098
18161
0.010
0.054
0.154
0.099
10133
0.010
0.128
0.143
0.015
0
0.010
0.005
0.154
0.149
18179
0.010
0.059
0.086
0.027
18161
0.010
0.057
0.154
0.097
5093
0.009
0.069
0.082
0.013
18161
0.008
0.123
0.154
0.030
18867
0.008
0.088
0.111
0.023
0
0.006
0.001
0.005
0.003
0
0.004
0.071
0.154
0.082
0
0.003
0.059
0.154
0.095
0
0.003
0.010
0.154
0.144
0
0.003
0.010
0.154
0.144
0
0.001
0.000
0.001
0.001
10133
0.001
0.143
0.154
0.011
20507
0.000
0.001
0.154
0.153
0
0.000
0.154
0.154
0.000
0
0.000
0.154
0.154
0.000
18867
0.000
0.154
0.154
0.000
18179
0.000
0.154
0.154
0.000
18179
0.000
0.154
0.154
0.000
18179
0.000
0.154
0.154
0.000
18867
0.000
0.154
0.154
0.000
18867
0.000
0.154
0.154
0.000
18179
0.000
0.154
0.154
0.000
Detecting Instruction Emulation with the emstat Tool
The POWER-based architecture deleted (that is, no longer supports) 35 POWER instructions (5 for the
PowerPC 601 RISC Microprocessor ″bridge″). To maintain compatibility with older binaries (which may
contain these deleted instructions) the AIX Version 4 kernel includes emulation routines that provide
support for the deleted instructions. Attempting to execute a deleted instruction results in an illegal
instruction exception. The kernel decodes the illegal instruction, and if it is a deleted instruction, the kernel
runs an emulation routine that functionally emulates the instruction.
However, depending upon the execution frequency of deleted instructions and the deleted instruction
emulation path lengths, emulation can result in varying degrees of performance degradation due to kernel
context switch and instruction emulation overhead. Even a very small percentage of emulation might result
in a big performance difference. The following table shows estimated instruction path lengths for several of
the deleted instructions:
Chapter 7. Monitoring and Tuning CPU Use
119
Instruction
Emulated in
Estimated Path Length (instructions)
abs
assembler
117
doz
assembler
120
mul
assembler
127
rlmi
C
425
sle
C
447
clf
C
542
div
C
1079
Most emulation problems are usually seen on PowerPC 604 RISC Microprocessor systems. A typical
example is a PowerPC 601 RISC Microprocessor system that gets upgraded to a PowerPC 604 RISC
Microprocessor system. If performance slows down instead of speeding up, it is most likely due to
emulation.
The solution to emulation is to recompile the application in common mode. The default architecture
platform for compilations on AIX Version 4 is common architecture. However, the default architecture on
AIX Version 3 was for POWER family, POWER2, and PowerPC 601 RISC Microprocessor. If these
binaries ran on a PowerPC 604 RISC Microprocessor, some instructions could get emulated.
If binaries need to be compiled on AIX Version 3, then the Full Common Mode APAR must be installed on
that AIX Version 3 system, and the application must be compiled with the following options:
-qarch=com -qdebug=useabs -bI:/lib/FCM/lowsys.exp
Instructions that are not common on all platforms must be removed from code written in assembler,
because recompilation is only effective for high-level source code. Routines coded in assembler must be
changed so that they do not use missing instructions, because recompilation has no effect in this case.
The first step is to determine if instruction emulation is occurring by using the emstat tool.
To determine whether the emstat program is installed and available, run the following command:
# lslpp -lI bos.adt.samples
This tool resides in the /usr/samples/kernel directory.
With operating system versions later than AIX 4.3.3, this command ships with the bos.perf.tools fileset
and has a slightly modified output. To determine whether the emstat program is installed and available,
run the following command:
# lslpp -lI bos.perf.tools
The emstat command works similarly to the vmstat command in that you specify an interval time in
seconds, and optionally, the number of intervals. The value in the first column is the cumulative count
since system boot, while the value in the second column is the number of instructions emulated during that
interval. Emulations on the order of many thousands per second can have an impact on performance.
The following is an example of output from issuing the emstat 1 command in AIX 4.3.3:
# /usr/samples/kernel/emstat 1
emstat total count
21951937
21967488
21974877
120
emstat interval count
21951937
15551
7389
Performance Management Guide
21994354
22007329
22018171
22018171
19477
12975
10842
0
The following is an example of output from issuing the emstat 1 command in AIX 5.1 or later:
# emstat 1
Emulation
SinceBoot
0
0
0
Emulation
Delta
0
0
0
Once emulation has been detected, the next step is to determine which application is emulating
instructions. This is much harder to determine. One way is to run only one application at a time and
monitor it with the emstat program. Sometimes certain emulations cause a trace hook to be encountered.
This can be viewed in the ASCII trace report file with the words PROGRAM CHECK. The process/thread
associated with this trace event is emulating instructions either due to its own code emulating instructions,
or it is executing library or code in other modules that are emulating instructions.
Detecting Alignment Exceptions with the alstat Tool
AIX compilers perform natural alignment of data types. For example, data of type short, which is 2 bytes
long, is padded automatically to 4 bytes by the compiler. Common programming practices such as
typecasting and usage of alignment pragmas can cause application data to be aligned incorrectly.
POWER-based optimization assumes correct alignment of data. Thus, fetching misaligned data may
require multiple memory accesses where a single access should have sufficed. Misalignment of data can
cause the hardware to generate an alignment exception, which would force the kernel to simulate the
needed memory accesses. As with the case of instrution emulation, this can degrade application
performance.
Starting with AIX 5.1, the alstat tool packaged with bos.perf.tools can be used to detect if alignment
exceptions are occurring. To show alignment exceptions on a per-CPU basis, use the -v option.
Because alstat and emstat are the same binary, either of these tools can be used to sho instruction
emulation and alignment exceptions. To show instruction emulation, use the -e option on alstat. To show
alignment exceptions, use the -a option on emstat.
The output for alstat looks similar to the following:
# alstat -e 1
Alignment Alignment Emulation
SinceBoot
Delta SinceBoot
0
0
0
0
0
0
0
0
0
Emulation
Delta
0
0
0
Restructuring Executable Programs with the fdpr Program
The fdpr (feedback-directed program restructuring) program optimizes executable modules for faster
execution and more efficient use of real memory. To determine whether the fdpr program is installed and
available on your system, run the following command:
# lslpp -lI perfagent.tools
The fdpr command is a performance-tuning utility that can improve both performance and real memory
utilization of user-level application programs. The source code is not necessary as input to the fdpr
program. However, stripped executable programs are not supported. If source code is available, programs
built with the -qfdpr compiler flag contain information to assist the fdpr program in producing reordered
Chapter 7. Monitoring and Tuning CPU Use
121
programs with guaranteed functionality. If the -qfdpr flag is used, it should be used for all object modules
in a program. Static linking will not enhance performance if the -qfdpr flag is used.
The fdpr tool reorders the instructions in an executable program to improve instruction cache, Translation
Lookaside Buffer (TLB), and real memory utilization by doing the following:
v Packing together highly executed code sequences (as determined through profiling)
v Recoding conditional branches to improve hardware branch prediction
v Moving infrequently executed code out of line
For example, given an ″if-then-else″ statement, the fdpr program might conclude that the program uses
the else branch more often than the if branch. It will then reverse the condition and the two branches as
shown in the following figure.
Figure 17. Example of Conditional Branch Recoding. The illustration shows how conditional branch recoding changes
certain code. For example, the code If (condition) would become If (! condition); the code then stays then; instructions
becomes other instructions; else stays else; other instructions become instruction; and endif stays endif.
Large applications (larger than 5 MB) that are CPU-bound can improve execution time up to 23 percent,
but typically the performance is improved between 5 and 20 percent. The reduction of real memory
requirements for text pages for this type of program can reach 70 percent. The average is between 20 and
50 percent. The numbers depend on the application’s behavior and the optimization options issued when
using the fdpr program.
The fdpr processing takes place in three stages:
1. The executable module to be optimized is instrumented to allow detailed performance-data collection.
2. The instrumented executable module is run in a workload provided by the user, and performance data
from that run is recorded.
3. The performance data is used to drive a performance-optimization process that results in a
restructured executable module that should perform the workload that exercised the instrumented
executable program more efficiently. It is critically important that the workload used to drive the fdpr
program closely match the actual use of the program. The performance of the restructured executable
program with workloads that differ substantially from that used to drive the fdpr program is
unpredictable, but can be worse than that of the original executable program.
As an example, the command
# fdpr -p ProgramName -R3 -x test.sh
would use the testcase test.sh to run an instrumented form of program ProgramName. The output of that
run would be used to perform the most aggressive optimization (-R3) of the program to form a new
module called, by default, ProgramName.fdpr. The degree to which the optimized executable program
performed better in production than its unoptimized predecessor would depend largely on the degree to
which the testcase test.sh successfully imitated the production workload.
Note: The fdpr program incorporates advanced optimization algorithms that sometimes result in optimized
executable programs that do not function in the same way as the original executable module. It is
122
Performance Management Guide
absolutely essential that any optimized executable program be thoroughly tested before being used
in any production situation; that is, before its output is trusted.
In
v
v
v
summary, users of the fdpr program should adhere to the following:
Take pains to use a workload to drive the fdpr program that is representative of the intended use.
Thoroughly test the functioning of the resulting restructured executable program.
Use the restructured executable program only on the workload for which it has been tuned.
Controlling Contention for the CPU
AIX Version 4 introduced the use of threads to control processor time consumption, but most of the system
management tools still refer to the process in which a thread is running, rather than the thread itself.
Controlling the Priority of User Processes
User-process priorities can be manipulated using the nice or renice command or the setpri() subroutine,
and displayed with the ps command. An overview of priority is provided in Process and Thread Priority.
Priority calculation is employed to accomplish the following:
v Share the CPU among threads
v Prevent starvation of any thread
v Penalize compute-bound threads
v Increase continuous discrimination between threads over time
The degree to which the user priorities can be manipulated is release-dependent. The algorithm for
calculating priority value in AIX 4.3.1 and prior releases is more limiting than the current algorithm. The
algorithm was changed in AIX 4.3.2 so that user priorities can be manipulated more than in previous
releases. There is improved distinction between foreground and background processes. Using a given
nice value will have a greater effect on CPU utilization. See “Priority Calculation” on page 126.
Running a Command with the nice Command
Any user can run a command at a less-favorable-than-normal priority by using the nice command. Only
the root user can use the nice command to run commands at a more-favorable-than-normal priority. In this
case, the nice command values range between -20 and 19.
With the nice command, the user specifies a value to be added to or subtracted from the standard nice
value. The modified nice value is used for the process that runs the specified command. The priority of
the process is still non-fixed; that is, the priority value is still recalculated periodically based on the CPU
usage, nice value, and minimum user-process-priority value.
The standard nice value of a foreground process is 20 (24 for a ksh background process). The following
command would cause the vmstat command to be run in the foreground with a nice value of 25 (instead
of the standard 20), resulting in a less favorable priority.
# nice -n 5 vmstat 10 3 > vmstat.out
If you use the root login, the vmstat command can be run at a more favorable priority with the following:
# nice -n -5 vmstat 10 3 > vmstat.out
If you were not using root login and issued the preceeding example nice command, the vmstat command
would still be run but at the standard nice value of 20, and the nice command would not issue any error
message.
Setting a Fixed Priority with the setpri Subroutine
An application that runs under the root user ID can use the setpri() subroutine to set its own priority or
that of another process. For example:
Chapter 7. Monitoring and Tuning CPU Use
123
retcode = setpri(0,59);
would give the current process a fixed priority of 59. If the setpri() subroutine fails, it returns -1.
The following program accepts a priority value and a list of process IDs and sets the priority of all of the
processes to the specified value.
/*
*/
fixprocpri.c
Usage: fixprocpri priority PID . . .
#include <sys/sched.h>
#include <stdio.h>
#include <sys/errno.h>
main(int argc,char **argv)
{
pid_t ProcessID;
int Priority,ReturnP;
if( argc < 3 ) {
printf(" usage - setpri priority pid(s) \n");
exit(1);
}
argv++;
Priority=atoi(*argv++);
if ( Priority < 50 ) {
printf(" Priority must be >= 50 \n");
exit(1);
}
}
while (*argv) {
ProcessID=atoi(*argv++);
ReturnP = setpri(ProcessID, Priority);
if ( ReturnP > 0 )
printf("pid=%d new pri=%d old pri=%d\n",
(int)ProcessID,Priority,ReturnP);
else {
perror(" setpri failed ");
exit(1);
}
}
Displaying Process Priority with the ps Command
The -l (lowercase L) flag of the ps command displays the nice values and current priority values of the
specified processes. For example, you can display the priorities of all of the processes owned by a given
user with the following:
# ps -lu
F
241801
200801
241801
hoetzel
S UID PID
S 200 7032
S 200 7568
S 200 8544
PPID
7286
7032
6494
C PRI NI ADDR
0 60 20 1b4c
0 70 25 2310
0 60 20 154b
SZ
108
88
108
WCHAN
TTY TIME CMD
pts/2 0:00 ksh
5910a58 pts/2 0:00 vmstat
pts/0 0:00 ksh
The output shows the result of the nice -n 5 command described previously. Process 7568 has an inferior
priority of 70. (The ps command was run by a separate session in superuser mode, hence the presence of
two TTYs.)
If one of the processes had used the setpri(10758, 59) subroutine to give itself a fixed priority, a sample
ps -l output would be as follows:
124
Performance Management Guide
F S UID
PID PPID
200903 S
0 10758 10500
C PRI NI ADDR
0 59 -- 3438
SZ
40
WCHAN
4f91f98
TTY TIME CMD
pts/0 0:00 fixpri
Modifying the Priority with the renice Command
The renice command alters the nice value, and thus the priority, of one or more processes that are
already running. The processes are identified either by process ID, process group ID, or the name of the
user who owns the processes.
For AIX Version 4, the syntax of the renice command has been changed to complement the alternative
syntax of the nice command, which uses the -n flag to identify the nice-value increment.
The renice command cannot be used on fixed-priority processes. A non-root user can specify a value to
be added to, but not subtracted from the nice value of one or more running processes. The modification is
done to the nice values of the processes. The priority of these processes is still non-fixed. Only the root
user can use the renice command to alter the priority value within the range of -20 to 20, or subtract from
the nice value of one or more running processes.
To continue the example, use the renice command to alter the nice value of the vmstat process that you
started with nice.
# renice -n -5 7568
# ps -lu hoetzel
F S UID PID
241801 S 200 7032
200801 S 200 7568
241801 S 200 8544
PPID
7286
7032
6494
C PRI NI ADDR
0 60 20 1b4c
0 60 20 2310
0 60 20 154b
SZ
108
92
108
WCHAN
TTY TIME CMD
pts/2 0:00 ksh
5910a58 pts/2 0:00 vmstat
pts/0 0:00 ksh
Now the process is running at a more favorable priority that is equal to the other foreground processes. To
undo the effects of this, you could issue the following:
# renice -n 5 7569
# ps -lu hoetzel
F S UID PID
241801 S 200 7032
200801 S 200 7568
241801 S 200 8544
PPID
7286
7032
6494
C PRI NI ADDR
0 60 20 1b4c
1 70 25 2310
0 60 20 154b
SZ
108
92
108
WCHAN
TTY TIME CMD
pts/2 0:00 ksh
5910a58 pts/2 0:00 vmstat
pts/0 0:00 ksh
In these examples, the renice command was run by the root user. When run by an ordinary user ID, there
are two major limitations to the use of the renice command:
v Only processes owned by that user ID can be specified.
v The nice value of the process cannot be decreased, not even to return the process to the default
priority after making its priority less favorable with the renice command.
Clarification of the nice and renice Command Syntax
The nice and renice commands have different ways of specifying the amount that is to be added to the
standard nice value of 20.
For AIX Version 4, the syntax of the renice command has been changed to complement the alternative
syntax of the nice command, which uses the -n flag to identify the nice-value increment.
Resulting nice
Value
Command
Command
Best Priority Value in
AIX 4.3.1
Best Priority Value in
AIX 4.3.2 and
Subsequent Releases
nice -n 5
renice -n 5
25
65
70
nice -n +5
renice -n +5
25
65
70
nice -n -5
renice -n -5
15
55
55
Chapter 7. Monitoring and Tuning CPU Use
125
Tuning the Thread-Priority-Value Calculation
This section discusses tuning using the following:
v Priority Calculation
v The schedutune command
The schedtune command and several enhancements to the CPU scheduler permit changes to the
parameters used to calculate the priority value for each thread. See Process and Thread Priority for
background information on priority.
To determine whether the schedtune program is installed and available, run the following command:
# lslpp -lI bos.adt.samples
Priority Calculation
The formula for calculating the priority value is:
priority value = base priority + nice penalty + (CPU penalty based on recent CPU usage)
The recent CPU usage value of a given thread is incremented by 1 each time that thread is in control of
the CPU when the timer interrupt occurs (every 10 milliseconds). The recent CPU usage value is
displayed as the C column in the ps command output. The maximum value of recent CPU usage is 120.
The default algorithm calculates the CPU penalty by dividing recent CPU usage by 2. The
CPU-penalty-to-recent-CPU-usage ratio is therefore 0.5. This ratio is controlled by a value called R (the
default is 16). The formula is as follows:
CPU_penalty = C * R/32
Once a second, the default algorithm divides the recent CPU usage value of every thread by 2. The
recent-CPU-usage-decay factor is therefore 0.5. This factor is controlled by a value called D (the default is
16). The formula is as follows:
C = C * D/32
For some users, the algorithm in AIX 4.3.1 does not allow enough distinction between foreground and
background processes. The algorithm for calculating priority value was changed in AIX 4.3.2 to increase
the impact on priorities when the nice value is changed. As the units of CPU time increase, the priority
decreases with the nice effect. Using schedtune -r -d can give additional control over the priority
calculation by setting new values for R and D. See “Tuning with the schedtune Command” on page 127 for
further information.
The algorithm guarantees that threads whose nice values are the default of 20 will behave exactly as they
did in AIX 4.3.1 and prior releases. When the nice value is not equal to the default, the priority will be
affected more due to the use of two formulas.
Begin with the following equation:
p_nice = base priority + nice value
Now use the following formula:
If p_nice > 60,
then x_nice = (p_nice * 2) - 60,
else x_nice = p_nice.
If the nice value is greater than 20, then it has double the impact on the priority value than if it was less
than or equal to 20. The new priority calculation (ignoring integer truncation) is as follows:
priority value = x_nice + [(x_nice + 4)/64 * C*(R/32)]
Note: If the nice value is the default, the formulas for AIX 4.3.1 and AIX 4.3.2 yield the same results.
126
Performance Management Guide
Tuning with the schedtune Command
Tuning is accomplished through two options of the schedtune command: -r and -d. Each option specifies
a parameter that is an integer from 0 through 32. The parameters are applied by multiplying by the
parameter’s value and then dividing by 32. The default r and d values are 16, which yields the same
behavior as the original algorithm [(D=R=16)/32=0.5]. The new range of values permits a far wider
spectrum of behaviors. For example:
# /usr/samples/kernel/schedtune -r 0
[(R=0)/32=0, (D=16)/32=0.5] would mean that the CPU penalty was always 0, making priority a function of
the nice value only. No background process would get any CPU time unless there were no dispatchable
foreground processes at all. The priority values of the threads would effectively be constant, although they
would not technically be fixed-priority threads.
# /usr/samples/kernel/schedtune -r 5
[(R=5)/32=0.15625, (D=16)/32=0.5] would mean that a foreground process would never have to compete
with a background process started with the command nice -n 10. The limit of 120 CPU time slices
accumulated would mean that the maximum CPU penalty for the foreground process would be 18.
# /usr/samples/kernel/schedtune -r 6 -d 16
[(R=6)/32=0.1875, (D=16)/32=0.5] would mean that, if the background process were started with the
command nice -n 10, it would be at least one second before the background process began to receive
any CPU time. Foreground processes, however, would still be distinguishable on the basis of CPU usage.
Long-running foreground processes that should probably be in the background would ultimately
accumulate enough CPU usage to keep them from interfering with the true foreground.
# /usr/samples/kernel/schedtune -r 32 -d 32
[(R=32)/32=1, (D=32)/32=1] would mean that long-running threads would reach a C value of 120 and
remain there, contending on the basis of their nice values. New threads would have priority, regardless of
their nice value, until they had accumulated enough time slices to bring them within the priority value
range of the existing threads.
Here are some guidelines for R and D:
v Smaller values of R narrow the priority range and make the nice value have more of an impact on the
priority.
v Larger values of R widen the priority range and make the nice value have less of an impact on the
priority.
v Smaller values of D decay CPU usage at a faster rate and can cause CPU-intensive threads to be
scheduled sooner.
v Larger values of D decay CPU usage at a slower rate and penalize CPU-intensive threads more (thus
favoring interactive-type threads).
If you conclude that one or both parameters need to be modified to accommodate your workload, you can
enter the schedtune command while logged on as root user. The changed values will remain until the
next schedtune command modifies them or until the next system boot. Values can be reset to their
defaults with the command schedtune -D, but remember that all schedtune parameters are reset by that
command, including VMM memory load control parameters. To make a change to the parameters that will
persist across boots, add an appropriate line at the end of the /etc/inittab file.
Example of a Priority Calculation
The example shows R=4 and D=31 and assumes no other runnable threads:
current_effective_priority
| base process priority
|
| nice value
|
|
| count (time slices consumed)
|
|
|
| (schedtune -r)
Chapter 7. Monitoring and Tuning CPU Use
127
|
|
|
|
|
0
p = 40 + 20 + (0
* 4/32) =
60
10 ms
p = 40 + 20 + (1
* 4/32) =
60
20 ms
p = 40 + 20 + (2
* 4/32) =
60
30 ms
p = 40 + 20 + (3
* 4/32) =
60
40 ms
p = 40 + 20 + (4
* 4/32) =
60
50 ms
p = 40 + 20 + (5
* 4/32) =
60
60 ms
p = 40 + 20 + (6
* 4/32) =
60
70 ms
p = 40 + 20 + (7
* 4/32) =
60
80 ms
p = 40 + 20 + (8
* 4/32) =
61
90 ms
p = 40 + 20 + (9
* 4/32) =
61
100ms
p = 40 + 20 + (10 * 4/32) =
61
.
(skipping forward to 1000msec or 1 second)
.
time 1000ms
p = 40 + 20 + (100 * 4/32) =
72
time 1000ms
swapper recalculates the accumulated CPU usage counts of
all processes. For the above process:
new_CPU_usage = 100 * 31/32 = 96 (if d=31)
after decaying by the swapper: p = 40 + 20 + ( 96 * 4/32) = 72
(if d=16, then p = 40 + 20 + (100/2 * 4/32) = 66)
time 1010ms
p = 40 + 20 + ( 97 * 4/32) =
72
time 1020ms
p = 40 + 20 + ( 98 * 4/32) =
72
time 1030ms
p = 40 + 20 + ( 99 * 4/32) =
72
..
time 1230ms
p = 40 + 20 + (119 * 4/32) =
74
time 1240ms
p = 40 + 20 + (120 * 4/32) =
75
count <= 120
time 1250ms
p = 40 + 20 + (120 * 4/32) =
75
time 1260ms
p = 40 + 20 + (120 * 4/32) =
75
..
time 2000ms
p = 40 + 20 + (120 * 4/32) =
75
time 2000ms
swapper recalculates the counts of all processes.
For above process 120 * 31/32 = 116
time 2010ms
p = 40 + 20 + (117 * 4/32) =
74
time
time
time
time
time
time
time
time
time
time
time
Modifying the Scheduler Time Slice with the schedtune Command
The length of the scheduler time slice can be modified with the schedtune command. To change the time
slice, use the schedtune -t option.
In AIX Version 4, the value of -t is the number of ticks for the time slice and only SCHED_RR threads will
use the nondefault time slice value (see Scheduling Policy for Threads for a description of fixed priority
threads).
Changing the time slice takes effect instantly and does not require a reboot.
A thread running with SCHED_OTHER or SCHED_RR scheduling policy can use the CPU for up to a full
time slice (the default time slice being 1 clock tick), a clock tick being 10 ms.
In some situations, too much context switching is occurring and the overhead of dispatching threads can
be more costly than allowing these threads to run for a longer time slice. In these cases, increasing the
time slice might have a positive impact on the performance of fixed-priority threads. Use the vmstat and
sar commands for determining the number of context switches per second.
In an environment in which the length of the time slice has been increased, some applications might not
need or should not have the full time slice. These applications can give up the processor explicitly with the
yield() system call (as can programs in an unmodified environment). After a yield() call, the calling thread
is moved to the end of the dispatch queue for its priority level.
128
Performance Management Guide
CPU-Efficient User ID Administration (The mkpasswd Command)
To improve login response time and conserve CPU time in systems with many users, the operating system
can use a hashed version of the /etc/passwd file to look up user IDs. When this facility is used, the
/etc/passwd file still exists, but is not used in normal processing. The hashed versions of the file are built
by the mkpasswd command. If the hashed versions are not current, login processing reverts to a slow,
CPU-intensive sequential search through /etc/passwd.
Prior to AIX 4.3, the command to run is mkpasswd /etc/passwd. This command creates a
/etc/passwd.pag and /etc/passwd.dir file, and lookups of /etc/passwd will do a hashed lookup into
passwd.pag (as long as this file is newer than /etc/passwd and /etc/security/passwd. Otherwise,
lookups will fall back to the default behavior).
Starting with AIX 4.3, the command to run is mkpasswd -f. This command creates indexed versions of
/etc/passwd, /etc/security/passwd, and /etc/security/lastlog. The files created are /etc/passwd.nm.idx,
/etc/passwd.id.idx, /etc/security/passwd.idx, and /etc/security/lastlog.idx. Note that this will greatly
enhance performance of applications that also need the encrypted password (such as login and any other
program that needs to do password authentication).
Applications can also be changed to use alternative routines such as _getpwent() instead of getpwent(),
_getpwnam_shadow(name,0) instead of getpwnam(name), or _getpwuid_shadow(uid,0) instead of
getpwuid(uid) to do name/ID resolution in cases where the encrypted password is not needed. This
prevents a lookup of /etc/security/passwd.
Do not edit the password files by hand because the time stamps of the database files (.pag or .idx) will
not be in sync and the default lookup method (linear) will be used. If the passwd, mkuser, chuser,
rmuser commands (or the SMIT command equivalents, with fast paths of the same name) are used to
administer user IDs, the hashed files are kept up to date automatically. If the /etc/passwd file is changed
with an editor or with the pwdadm command, the hashed or index files must be rebuilt.
Note: The mkpasswd command does not affect NIS, DCE, or LDAP user databases.
Chapter 7. Monitoring and Tuning CPU Use
129
130
Performance Management Guide
Chapter 8. Monitoring and Tuning Memory Use
The memory of a system is almost constantly filled to capacity. Even if currently running programs do not
consume all available memory, the operating system retains in memory the text pages of programs that
ran earlier and the files that they used. There is no cost associated with this retention, because the
memory would have been unused anyway. In many cases, the programs or files will be used again, which
reduces disk I/O.
This chapter describes how memory use can be measured and modified. It contains the following major
sections:
v Determining How Much Memory Is Being Used
v Finding Memory-Leaking Programs
v Assessing Memory Requirements Through the rmss Command
v Tuning VMM Memory Load Control with the schedtune Command
v Tuning VMM Page Replacement with the vmtune Command
v Tuning Paging-Space Thresholds
v Choosing a Page Space Allocation Method
v Using Shared Memory
v Using AIX Memory Affinity Support
Readers who are not familiar with the operating system’s virtual-memory management may want to look at
Performance Overview of the Virtual Memory Manager (VMM) before continuing.
Determining How Much Memory Is Being Used
Several performance tools provide reports of memory usage. The reports of most interest are from the
vmstat, ps, and svmon commands.
The vmstat Command (Memory)
The vmstat command summarizes the total active virtual memory used by all of the processes in the
system, as well as the number of real-memory page frames on the free list. Active virtual memory is
defined as the number of virtual-memory working segment pages that have actually been touched (for a
definition see Late Page Space Allocation). This number can be larger than the number of real page
frames in the machine, because some of the active virtual-memory pages may have been written out to
paging space.
When determining if a system might be short on memory or if some memory tuning needs to be done, run
the vmstat command over a set interval and examine the pi and po columns on the resulting report.
These columns indicate the number of paging space page-ins per second and the number of paging space
page-outs per second. If the values are constantly non-zero, there might be a memory bottleneck. Having
occasional non-zero values is not be a concern because paging is the main principle of virtual memory.
# vmstat 2 10
kthr
memory
page
faults
cpu
----- ----------- ------------------------ ------------ ----------r b
avm
fre re pi po fr
sr cy in
sy cs us sy id wa
1 3 113726
124
0 14
6 151 600
0 521 5533 816 23 13 7 57
0 3 113643
346
0
2 14 208 690
0 585 2201 866 16 9 2 73
0 3 113659
135
0
2
2 108 323
0 516 1563 797 25 7 2 66
0 2 113661
122
0
3
2 120 375
0 527 1622 871 13 7 2 79
0 3 113662
128
0 10
3 134 432
0 644 1434 948 22 7 4 67
1 5 113858
238
0 35
1 146 422
0 599 5103 903 40 16 0 44
0 3 113969
127
0
5 10 153 529
0 565 2006 823 19 8 3 70
© Copyright IBM Corp. 1997, 2002
131
0
0
0
3 113983
3 113682
4 113701
125
121
124
0
0
0
33
20
3
5 153
9 154
29 228
424
470
635
0 559 2165 921 25 8 4 63
0 608 1569 1007 15 8 0 77
0 674 1730 1086 18 9 0 73
Notice the high I/O wait in the output and also the number of threads on the blocked queue. Most likely,
the I/O wait is due to the paging in/out from paging space.
To see if the system has performance problems with its VMM, examine the columns under memory and
page:
v memory
Provides information about the real and virtual memory.
– avm
The avm (Active Virtual Memory) column represents the number of active virtual memory pages
present at the time the vmstat sample was collected. Starting in AIX 4.3.2, a deferred-page-space
policy is in effect by default. Under this policy, the value for avm can be higher than the number of
paging space pages used. The avm statistics do not include file pages.
Note: Starting with AIX 4.3.2, there is a slight change in reporting this value. See Looking at Paging
Space and Virtual Memory for an explanation.
Prior to AIX 4.3.2, the number in the avm field divided by 256 will yield the approximate number of
megabytes (MB) allocated to paging space systemwide. This is also true for AIX 4.3.2 and later if
you change the default page space policy. Prior to AIX 4.3.2 the same information is reflected in the
Percent Used column of the lsps -s command output or with the svmon -G command under the pg
space inuse field.
– fre
The fre column shows the average number of free memory pages. A page is a 4 KB area of real
memory. The system maintains a buffer of memory pages, called the free list, that will be readily
accessible when the VMM needs space. The minimum number of pages that the VMM keeps on the
free list is determined by the minfree parameter of the vmtune command (for details, see Tuning
VMM Page Replacement with the vmtune Command).
When an application terminates, all of its working pages are immediately returned to the free list. Its
persistent pages (files), however, remain in RAM and are not added back to the free list until they
are stolen by the VMM for other programs. Persistent pages are also freed if the corresponding file is
deleted.
For this reason, the fre value may not indicate all the real memory that can be readily available for
use by processes. If a page frame is needed, then persistent pages related to terminated
applications are among the first to be handed over to another program.
If the fre value is substantially above the maxfree value, it is unlikely that the system is thrashing.
Thrashing means that the system is continuously paging in and out. However, if the system is
experiencing thrashing, you can be assured that the fre value will be small.
v page
Information about page faults and paging activity. These are averaged over the interval and given in
units per second.
– re
Note: In AIX Version 4, reclaiming is no longer supported because the value it delivered by
providing limited information about the performance of the system was outweighed by the
negative affect on system performance due to the algorithm that kept track of reclaims.
– pi
The pi column details the number (rate) of pages paged in from paging space. Paging space is the
part of virtual memory that resides on disk. It is used as an overflow when memory is over
committed. Paging space consists of logical volumes dedicated to the storage of working set pages
132
Performance Management Guide
that have been stolen from real memory. When a stolen page is referenced by the process, a page
fault occurs, and the page must be read into memory from paging space.
Due to the variety of configurations of hardware, software and applications, there is no absolute
number to look out for. But five page-ins per second per paging space should be the upper limit. This
guideline should not be rigidly adhered to, but used as a reference. This field is important as a key
indicator of paging-space activity. If a page-in occurs, there must have been a previous page-out for
that page. It is also likely in a memory-constrained environment that each page-in will force a
different page to be stolen and, therefore, paged out. But systems could also work fine when they
have close to 10 pi per second for 1 minute and then work without any page-ins.
– po
The po column shows the number (rate) of pages paged out to paging space. Whenever a page of
working storage is stolen, it is written to paging space, if it does not yet reside in paging space or if it
was modified. If not referenced again, it will remain on the paging device until the process terminates
or disclaims the space. Subsequent references to addresses contained within the faulted-out pages
results in page faults, and the pages are paged in individually by the system. When a process
terminates normally, any paging space allocated to that process is freed. If the system is reading in a
significant number of persistent pages (files), you might see an increase in po without corresponding
increases in pi. This does not necessarily indicate thrashing, but may warrant investigation into
data-access patterns of the applications.
– fr
Number of pages that were freed per second by the page-replacement algorithm during the interval.
As the VMM page-replacement routine scans the Page Frame Table (PFT), it uses criteria to select
which pages are to be stolen to replenish the free list of available memory frames. The criteria
include both kinds of pages, working (computational) and file (persistent) pages. Just because a
page has been freed, it does not mean that any I/O has taken place. For example, if a persistent
storage (file) page has not been modified, it will not be written back to the disk. If I/O is not
necessary, minimal system resources are required to free a page.
– sr
Number of pages that were examined per second by the page-replacement algorithm during the
interval. The VMM page-replacement code scans the PFT and steals pages until the number of
frames on the free list is at least the maxfree value. The page-replacement code might have to scan
many entries in the PFT before it can steal enough to satisfy the free list requirements. With stable,
unfragmented memory, the scan rate and free rate might be nearly equal. On systems with multiple
processes using many different pages, the pages are more volatile and disjoint. In this scenario, the
scan rate might greatly exceed the free rate.
Memory is over committed when the ratio of fr to sr (fr:sr) is high.
An fr:sr ratio of 1:4 means that for every page freed, four pages had to be examined. It is difficult to
determine a memory constraint based on this ratio alone, and what constitutes a high ratio is
workload/application dependent.
– cy
Number of cycles per second of the clock algorithm. The VMM uses a technique known as the clock
algorithm to select pages to be replaced. This technique takes advantage of a referenced bit for
each page as an indication of what pages have been recently used (referenced). When the
page-stealer routine is called, it cycles through the PFT, examining each page’s referenced bit.
The cy column shows how many times per second the page-replacement code has scanned the
PFT. Because the free list can be replenished without a complete scan of the PFT and because all of
the vmstat fields are reported as integers, this field is usually zero. If not, it indicates a complete
scan of the PFT, and the stealer has to scan the PFT again, because fre is still under the maxfree
value.
One way to determine the appropriate amount of RAM for a system is to look at the largest value for avm
as reported by the vmstat command. Multiply that by 4 K to get the number of bytes and then compare
that to the number of bytes of RAM on the system. Ideally, avm should be smaller than total RAM. If not,
Chapter 8. Monitoring and Tuning Memory Use
133
some amount of virtual memory paging will occur. How much paging occurs will depend on the difference
between the two values. Remember, the idea of virtual memory is that it gives us the capability of
addressing more memory than we have (some of the memory is in RAM and the rest is in paging space).
But if there is far more virtual memory than real memory, this could cause excessive paging which then
results in delays. If avm is lower than RAM, then paging-space paging could be caused by RAM being filled
up with file pages. In that case, tuning the minperm/maxperm values, could reduce the amount of
paging-space paging (see Tuning VMM Page Replacement with the vmtune Command).
The vmstat -I Command
In operating system versions later than AIX 4.3.3, the vmstat -I (uppercase i) command displays additional
information, such as file pages in per-second, file pages out per-second (that is, any VMM page-ins and
page-outs that are not paging space page-ins or paging space page-outs). The re and cy columns are not
reported with this flag.
The vmstat -s Command
The summary (-s) option sends a summary report to standard output starting from system initialization
expressed in absolute counts rather than on an interval basis. The recommended way of using these
statistics is to run this command before a workload, save the output, and then run it again after the
workload and save its output. The next step is to determine the difference between the two sets of output.
An awk script called vmstatit that does this automatically is provided in Determining Whether the Problem
is Related to Disk or Memory.
# vmstat -s
3231543 total address trans. faults
63623 page ins
383540 page outs
149 paging space page ins
832 paging space page outs
0 total reclaims
807729 zero filled pages faults
4450 executable filled pages faults
429258 pages examined by clock
8 revolutions of the clock hand
175846 pages freed by the clock
18975 backtracks
0 lock misses
40 free frame waits
0 extend XPT waits
16984 pending I/O waits
186443 start I/Os
186443 iodones
141695229 cpu context switches
317690215 device interrupts
0 software interrupts
0 traps
55102397 syscalls
The page-in and page-out numbers in the summary represent virtual memory activity to page in or out
pages from page space and file space. The paging space ins and outs are representative of only page
space. If you are experiencing considerable I/O wait time, but are not memory-constrained, the problem
might be due to poor distribution of I/O across drives. To determine if the problem is due to paging space
or files, take multiple samples of the vmstat -s command and see if the page ins and outs to paging
space are the majority of the total paging. If the system is paging too much, using the vmtune command
might help (see Tuning VMM Page Replacement with the vmtune Command). Creating separate paging
spaces on separate volumes might provide some benefit, but increasing the memory would definitely help.
The ps Command
The ps command can also be used to monitor memory usage of individual processes. The ps v PID
command provides the most comprehensive report on memory-related statistics for an individual process,
such as:
134
Performance Management Guide
v
v
v
v
v
Page faults
Size of working segment that has been touched
Size of working segment and code segment in memory
Size of text segment
Size of resident set
v Percentage of real memory used by this process
An example:
# ps v
PID
36626
TTY STAT
pts/3 A
TIME PGIN
0:00
0
SIZE
316
RSS
LIM
408 32768
TSIZ
51
TRS %CPU %MEM COMMAND
60 0.0 0.0 ps v
The most important columns on the resulting ps report are described as follows:
PGIN
Number of page-ins caused by page faults. Since all I/O is classified as page faults, this is
basically a measure of I/O volume.
SIZE
Virtual size (in paging space) in kilobytes of the data section of the process (displayed as SZ by
other flags). This number is equal to the number of working segment pages of the process that
have been touched times 4. If some working segment pages are currently paged out, this number
is larger than the amount of real memory being used. SIZE includes pages in the private segment
and the shared-library data segment of the process.
RSS
Real-memory (resident set) size in kilobytes of the process. This number is equal to the sum of
the number of working segment and code segment pages in memory times 4. Remember that
code segment pages are shared among all of the currently running instances of the program. If 26
ksh processes are running, only one copy of any given page of the ksh executable program
would be in memory, but the ps command would report that code segment size as part of the RSS
of each instance of the ksh program.
TSIZ
Size of text (shared-program) image. This is the size of the text section of the executable file.
Pages of the text section of the executable program are only brought into memory when they are
touched, that is, branched to or loaded from. This number represents only an upper bound on the
amount of text that could be loaded. The TSIZ value does not reflect actual memory usage. This
TSIZ value can also be seen by executing the dump -ov command against an executable
program (for example, dump -ov /usr/bin/ls).
TRS
Size of the resident set (real memory) of text. This is the number of code segment pages times 4.
This number exaggerates memory use for programs of which multiple instances are running. The
TRS value can be higher than the TSIZ value because other pages may be included in the code
segment such as the XCOFF header and the loader section.
%MEM
Calculated as the sum of the number of working segment and code segment pages in memory
times 4 (that is, the RSS value), divided by the size of the real memory of the machine in KB,
times 100, rounded to the nearest full percentage point. This value attempts to convey the
percentage of real memory being used by the process. Unfortunately, like RSS, it tends the
exaggerate the cost of a process that is sharing program text with other processes. Further, the
rounding to the nearest percentage point causes all of the processes in the system that have RSS
values under 0.005 times real memory size to have a %MEM of 0.0.
Note: The ps command does not indicate memory consumed by shared memory segments or
memory-mapped segments. Because many applications use shared memory or memory-mapped
segments, the svmon command is a better tool to view the memory usage of these segments.
Chapter 8. Monitoring and Tuning Memory Use
135
The svmon Command
The svmon command provides a more in-depth analysis of memory usage. It is more informative, but also
more intrusive, than the vmstat and ps commands. The svmon command captures a snapshot of the
current state of memory. However, it is not a true snapshot because it runs at the user level with interrupts
enabled.
To determine whether svmon is installed and available, run the following command:
# lslpp -lI fileset_name
where fileset_name is perfagent.tools in AIX 4.3 or bos.perf.tools in AIX 5 or later.
The svmon command can only be executed by the root user.
If an interval is used (-i option), statistics will be displayed until the command is killed or until the number
of intervals, which can be specified right after the interval, is reached.
You can use four different reports to analyze the displayed information:
Global (-G)
Displays statistics describing the real memory and paging space in use for the whole system.
Process (-P)
Displays memory usage statistics for active processes.
Segment (-S)
Displays memory usage for a specified number of segments or the top ten highest memory-usage
processes in descending order.
Detailed Segment (-D)
Displays detailed information on specified segments.
Additional reports are available in AIX 4.3.3 and later, as follows:
User (-U)
Displays memory usage statistics for the specified login names. If no list of login names is
supplied, memory usage statistics display all defined login names.
Command (-C)
Displays memory usage statistics for the processes specified by command name.
Workload Management Class (-W)
Displays memory usage statistics for the specified workload management classes. If no classes
are supplied, memory usage statistics display all defined classes.
To support 64-bit applications, the output format of the svmon command was modified in AIX 4.3.3 and
later.
Additional reports are available in operating system versions later than 4.3.3, as follows:
Frame (-F)
Displays information about frames. When no frame number is specified, the percentage of used
memory is reported. When a frame number is specified, information about that frame is reported.
Tier (-T)
Displays information about tiers, such as the tier number, the superclass name when the -a flag is
used, and the total number of pages in real memory from segments belonging to the tier.
How Much Memory is in Use
To print out global statistics, use the -G flag. In this example, we will repeat it five times at two-second
intervals.
136
Performance Management Guide
# svmon -G -i 2 5
m e m o r y
size inuse free
16384 16250
134
16384 16254
130
16384 16254
130
16384 16254
130
16384 16254
130
pin
2006
2006
2006
2006
2006
i n u s
work pers
10675 2939
10679 2939
10679 2939
10679 2939
10679 2939
e
clnt
2636
2636
2636
2636
2636
p i n
work pers clnt
2006
0
0
2006
0
0
2006
0
0
2006
0
0
2006
0
0
p g s p a c e
size
inuse
40960
12674
40960
12676
40960
12676
40960
12676
40960
12676
The columns on the resulting svmon report are described as follows:
memory
Statistics describing the use of real memory, shown in 4 K pages.
size
Total size of memory in 4 K pages.
inuse Number of pages in RAM that are in use by a process plus the number of persistent
pages that belonged to a terminated process and are still resident in RAM. This value is
the total size of memory minus the number of pages on the free list.
free
Number of pages on the free list.
pin
Number of pages pinned in RAM (a pinned page is a page that is always resident in RAM
and cannot be paged out).
in use Detailed statistics on the subset of real memory in use, shown in 4 K frames.
pin
work
Number of working pages in RAM.
pers
Number of persistent pages in RAM.
clnt
Number of client pages in RAM (client page is a remote file page).
Detailed statistics on the subset of real memory containing pinned pages, shown in 4 K frames.
work
Number of working pages pinned in RAM.
pers
Number of persistent pages pinned in RAM.
clnt
Number of client pages pinned in RAM.
pg space
Statistics describing the use of paging space, shown in 4 K pages. This data is reported only if the
-r flag is not used. The value reported starting with AIX 4.3.2 is the actual number of paging-space
pages used (which indicates that these pages were paged out to the paging space). This differs
from the vmstat command in that vmstat’s avm column which shows the virtual memory accessed
but not necessarily paged out.
size
Total size of paging space in 4 K pages.
inuse Total number of allocated pages.
In our example, there are 16384 pages of total size of memory. Multiply this number by 4096 to see the
total real memory size (64 MB). While 16250 pages are in use, there are 134 pages on the free list and
2006 pages are pinned in RAM. Of the total pages in use, there are 10675 working pages in RAM, 2939
persistent pages in RAM, and 2636 client pages in RAM. The sum of these three parts is equal to the
inuse column of the memory part. The pin part divides the pinned memory size into working, persistent and
client categories. The sum of them is equal to the pin column of the memory part. There are 40960 pages
(160 MB) of total paging space, and 12676 pages are in use. The inuse column of memory is usually
greater than the inuse column of pg space because memory for file pages is not freed when a program
completes, while paging-space allocation is.
In AIX 4.3.3 and later, systems the output of the same command looks similar to the following:
Chapter 8. Monitoring and Tuning Memory Use
137
# svmon -G -i 2 5
memory
pg space
pin
in use
memory
pg space
pin
in use
memory
pg space
pin
in use
memory
pg space
pin
in use
memory
pg space
pin
in use
size
65527
131072
inuse
64087
55824
free
1440
work
5918
47554
pers
0
13838
clnt
0
2695
size
65527
131072
inuse
64091
55824
free
1436
work
5918
47558
pers
0
13838
clnt
0
2695
size
65527
131072
inuse
64091
55824
free
1436
work
5918
47558
pers
0
13838
clnt
0
2695
size
65527
131072
inuse
64090
55824
free
1437
work
5918
47558
pers
0
13837
clnt
0
2695
size
65527
131072
inuse
64168
55824
free
1359
work
5921
47636
pers
0
13837
clnt
0
2695
pin
5909
virtual
81136
pin
5909
virtual
81137
pin
5909
virtual
81137
pin
5909
virtual
81137
pin
5912
virtual
81206
The additional output field is the virtual field, which shows the number of pages allocated in the system
virtual space.
Who is Using Memory?
The following command displays the memory usage statistics for the top ten processes. If you do not
specify a number, it will display all the processes currently running in this system.
# svmon -Pau 10
Pid
15012
2750
15706
17172
21150
17764
2910
19334
13664
17520
138
Command
maker4X.exe
X
dtwm
dtsession
dtterm
aixterm
dtterm
dtterm
dtterm
aixterm
Performance Management Guide
Inuse
4783
4353
3257
2986
2941
2862
2813
2813
2804
2801
Pin
1174
1178
1174
1174
1174
1174
1174
1174
1174
1174
Pgspace
4781
5544
4003
3827
3697
3644
3705
3704
3706
3619
Pid: 15012
Command: maker4X.exe
Segid
1572
142
1bde
2c1
9ab
404
1d9b
909
5a3
1096
1b9d
1af8
0
...
Type
pers
pers
pers
pers
pers
work
work
work
work
work
work
clnt
work
Description
Inuse
/dev/hd3:62
0
/dev/hd3:51
0
/dev/hd3:50
0
/dev/hd3:49
1
/dev/hd2:53289
1
kernel extension
27
lib data
39
shared library text
864
sreg[4]
9
sreg[3]
32
private
1057
961
kernel
1792
Pin
0
0
0
0
0
27
0
0
0
0
1
0
1146
Pgspace
0
0
0
0
0
0
23
7
12
32
1219
0
3488
Address Range
0..-1
0..-1
0..-1
0..7
0..0
0..24580
0..607
0..65535
0..32768
0..32783
0..1306 : 65307..65535
0..1716
0..32767 : 32768..65535
The output is divided into summary and detail sections. The summary section lists the top ten highest
memory-usage processes in descending order.
Pid 15012 is the process ID that has the highest memory usage. The Command indicates the command
name, in this case maker4X.exe. The Inuse column (total number of pages in real memory from segments
that are used by the process) shows 4783 pages (each page is 4 KB). The Pin column (total number of
pages pinned from segments that are used by the process) shows 1174 pages. The Pgspace column (total
number of paging-space pages that are used by the process) shows 4781 pages.
The detailed section displays information about each segment for each process that is shown in the
summary section. This includes the segment ID, the type of the segment, description (a textual description
of the segment, including the volume name and i-node of the file for persistent segments), number of
pages in RAM, number of pinned pages in RAM, number of pages in paging space, and address range.
The Address Range specifies one range for a persistent or client segment and two ranges for a working
segment. The range for a persistent or a client segment takes the form ’0..x,’ where x is the maximum
number of virtual pages that have been used. The range field for a working segment can be ’0..x :
y..65535’, where 0..x contains global data and grows upward, and y..65535 contains stack area and grows
downward. For the address range, in a working segment, space is allocated starting from both ends and
working towards the middle. If the working segment is non-private (kernel or shared library), space is
allocated differently. In this example, the segment ID 1b9d is a private working segment; its address range
is 0..1306 : 65307..65535. The segment ID 909 is a shared library text working segment; its address
range is 0..65535.
A segment can be used by multiple processes. Each page in real memory from such a segment is
accounted for in the Inuse field for each process using that segment. Thus, the total for Inuse may exceed
the total number of pages in real memory. The same is true for the Pgspace and Pin fields. The sum of
Inuse, Pin, and Pgspace of all segments of a process is equal to the numbers in the summary section.
You can use one of the following commands to display the file name associated with the i-node:
v ncheck -i i-node_number volume_name
v find file_system_associated_with_lv_name -xdev -inum inode_number -print
To get a similar output in AIX 4.3.3 and later, use the following command:
# svmon -Put 10
-----------------------------------------------------------------------------Pid Command
Inuse
Pin
Pgsp Virtual
64-bit
Mthrd
2164 X
15535
1461
34577
37869
N
N
Chapter 8. Monitoring and Tuning Memory Use
139
Vsid
1966
Esid Type Description
2 work process private
4411
0
d work shared library text
0 work kernel seg
396e
2ca3
43d5
2661
681f
356d
34e8
5c97
5575
4972
4170
755d
6158
1
f
3
-
pers
work
work
work
work
work
work
pers
pers
pers
pers
pers
pers
code,/dev/hd2:18950
shared library data
shmat/mmap
/dev/hd4:2
/dev/hd2:19315
/dev/hd2:19316
/dev/hd3:28
/dev/hd9var:94
/dev/hd9var:90
Inuse
9984
3165
2044
200
32
31
29
29
18
2
1
0
0
0
0
0
Pin Pgsp Virtual Addr Range
4 31892 32234
0..32272 :
65309..65535
0 1264 1315
0..65535
1455 1370 4170
0..32767 :
65475..65535
0
0..706
0
0
32
0..32783
0
6
32
0..32783
0
0
29
0..32783
0
25
29
0..32783
0
18
24
0..310
2
2
4
0..32767
0
0..0
0
0..0
0
0..5
0
0..0
0
0..0
0
0..0
-----------------------------------------------------------------------------Pid Command
Inuse
Pin
Pgsp Virtual
64-bit
Mthrd
25336 austin.ibm.
12466
1456
2797
11638
N
N
Vsid
14c3
4411
0
13c5
d21
9e6
942
2ca3
49f0
1b07
623
2de9
1541
5d15
4513
cc4
242a
...
Esid Type Description
2 work process private
d work shared library text
0 work kernel seg
1
f
5
4
-
clnt
pers
work
pers
work
clnt
pers
pers
clnt
mmap
pers
pers
mmap
pers
code
/dev/andy:563
shared library data
/dev/cache:16
/dev/andy:8568
/dev/hd2:22539
mapped to sid 761b
/dev/andy:487
/dev/andy:486
mapped to sid 803
/dev/andy:485
Inuse
5644
3165
2044
735
603
190
43
32
10
0
0
0
0
0
0
0
0
Pin Pgsp Virtual Addr Range
1 161 5993
0..6550 :
65293..65535
0 1264 1315
0..65535
1455 1370 4170
0..32767 :
65475..65535
0
0..4424
0
0..618
0
2
128
0..3303
0
0..42
0
0
32
0..32783
0
0..471
0
0..0
0
0..1
0
0..0
0
0
0..3
0
0..45
0
0
0..0
The Vsid column is the virtual segment ID, and the Esid column is the effective segment ID. The effective
segment ID reflects the segment register that is used to access the corresponding pages.
Detailed Information on a Specific Segment ID
The -D option displays detailed memory-usage statistics for segments.
# svmon -D 404
Segid: 404
Type: working
Description: kernel extension
Address Range: 0..24580
Size of page space allocation: 0 pages ( 0.0 Mb)
Inuse: 28 frames ( 0.1 Mb)
Page
Frame
Pin
Ref
Mod
12294
3320
pin
ref
mod
24580
1052
pin
ref
mod
12293
52774
pin
ref
mod
24579
20109
pin
ref
mod
12292
19494
pin
ref
mod
12291
52108
pin
ref
mod
24578
50685
pin
ref
mod
140
Performance Management Guide
12290
24577
12289
24576
12288
4112
4111
4110
4109
4108
4107
4106
4105
4104
4103
4102
4101
4100
4099
4098
4097
51024
1598
35007
204
206
53007
53006
53005
53004
53003
53002
53001
53000
52999
52998
52997
52996
52995
52994
52993
52992
pin
pin
pin
pin
pin
pin
pin
pin
pin
pin
pin
pin
pin
pin
pin
pin
pin
pin
pin
pin
pin
ref
ref
ref
ref
ref
ref
mod
mod
mod
mod
mod
mod
mod
mod
mod
mod
mod
mod
mod
mod
mod
mod
mod
mod
mod
mod
mod
The detail columns are explained as follows:
Page
Specifies the index of the page within the segment.
Frame Specifies the index of the real memory frame that the page resides in.
Pin
Specifies a flag indicating whether the page is pinned.
Ref
Specifies a flag indicating whether the page’s reference bit is on.
Mod
Specifies a flag indicating whether the page is modified.
The size of page space allocation is 0 because all the pages are pinned in real memory.
An example output from AIX 4.3.3 and later, is very similar to the following:
# svmon -D 629 -b
Segid: 629
Type: working
Address Range: 0..77
Size of page space allocation: 7 pages (
Virtual: 11 frames ( 0.0 Mb)
Inuse: 7 frames ( 0.0 Mb)
Page
0
3
7
8
5
1
77
Frame
32304
32167
32321
32320
32941
48357
47897
Pin
N
N
N
N
N
N
N
0.0 Mb)
Ref
Y
Y
Y
Y
Y
N
N
Mod
Y
Y
Y
Y
Y
Y
Y
The -b flag shows the status of the reference and modified bits of all the displayed frames. After it is
shown, the reference bit of the frame is reset. When used with the -i flag, it detects which frames are
accessed between each interval.
Note: Use this flag with caution because of its performance impacts.
Chapter 8. Monitoring and Tuning Memory Use
141
List of Top Memory Usage of Segments
The -S option is used to sort segments by memory usage and to display the memory-usage statistics for
the top memory-usage segments. If count is not specified, then a count of 10 is implicit. The following
command sorts system and non-system segments by the number of pages in real memory and prints out
the top 10 segments of the resulting list.
# svmon -Sau
Segid
0
1
1435
11f5
11f3
681
909
101
a0a
1bf9
Type
work
work
work
work
clnt
clnt
work
work
work
work
Description
kernel
private, pid=4042
private, pid=3006
private, pid=14248
Inuse
1990
1553
1391
1049
991
960
shared library text
900
vmm data
497
shared library data
247
private, pid=21094
221
Pin Pgspace Address Range
1408
3722 0..32767 : 32768..65535
1
1497 0..1907 : 65307..65535
3
1800 0..4565 : 65309..65535
1
1081 0..1104 : 65307..65535
0
0 0..1716
0
0 0..1880
0
8 0..65535
496
1 0..27115 : 43464..65535
0
718 0..65535
1
320 0..290 : 65277..65535
All output fields are described in the previous examples.
An example output from AIX 4.3.3 and later is similar to the following:
# svmon -Sut 10
Vsid
1966
Esid Type Description
- work
14c3
- work
5453
- work
4411
5a1e
- work
- work
340d
- work misc kernel tables
380e
0
- work kernel pinned heap
- work kernel seg
6afb
2faa
- pers /dev/notes:92
- clnt
Inuse
9985
Pin Pgsp Virtual Addr Range
4 31892 32234
0..32272 :
65309..65535
5644
1 161 5993
0..6550 :
65293..65535
3437
1 2971 4187
0..4141 :
65303..65535
3165
0 1264 1315
0..65535
2986
1
13 2994
0..3036 :
65295..65535
2643
0 993 2645
0..15038 :
63488..65535
2183 1055 1416 2936
0..65535
2044 1455 1370 4170
0..32767 :
65475..65535
1522
0
0..10295
1189
0
0..2324
Correlating svmon and vmstat Outputs
There are some relationships between the svmon and vmstat outputs. The svmon report of AIX 4.3.2
follows (the example is the same with AIX 4.3.3 and later, although the output format is different):
# svmon -G
m e m o r y
size inuse free
16384 16254 130
pin
2016
i n u s e
work pers clnt
11198 2537 2519
p i n
work pers clnt
2016
0
0
p g s p a c e
size inuse
40960 13392
The vmstat command was run in a separate window while the svmon command was running. The
vmstat report follows:
# vmstat 5
kthr
memory
page
faults
cpu
----- ----------- ------------------------ ------------ ----------r b
avm
fre re pi po fr
sr cy in
sy cs us sy id wa
0 0 13392
130 0
0
0
0
2
0 125 140 36 2 1 97 0
0 0 13336
199 0
0
0
0
0
0 145 14028 38 11 22 67 0
0 0 13336
199 0
0
0
0
0
0 141
49 31 1 1 98 0
0 0 13336
199 0
0
0
0
0
0 142
49 32 1 1 98 0
142
Performance Management Guide
0
0
0
0 13336
0 13336
0 13336
199
199
199
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 145
0 163
0 142
49
49
49
32
33
32
1
1
0
1 99
1 92
1 98
0
6
0
The global svmon report shows related numbers. The vmstatfre column relates to the svmon memory
free column. The number that vmstat reports as Active Virtual Memory (avm) is reported by the svmon
command as pg space inuse (13392).
The vmstat avm column provides the same figures as the pg space inuse column of the svmon command
except starting with AIX 4.3.2 where Deferred Page Space Allocation is used. In that case, the svmon
command shows the number of pages actually paged out to paging space whereas the vmstat command
shows the number of virtual pages accessed but not necessarily paged out (see Looking at Paging Space
and Virtual Memory).
Correlating svmon and ps Outputs
There are some relationships between the svmon and ps outputs. The svmon report of AIX 4.3.2 follows
(the example is the same with AIX 4.3.3 and later, although the output format is different):
# svmon -P 7226
Pid
7226
Command
telnetd
Inuse
936
Pin
1
Pgspace
69
Pid:
7226
Command: telnetd
Segid Type Description
Inuse
828 pers /dev/hd2:15333
0
1d3e work lib data
0
909 work shared library text
930
1cbb work sreg[3]
0
1694 work private
6
12f6 pers code,/dev/hd2:69914
0
Pin
0
0
0
0
1
0
Pgspace
0
28
8
1
32
0
Address Range
0..0
0..559
0..65535
0..0
0..24 : 65310..65535
0..11
TSIZ
33
TRS %CPU %MEM COMMAND
0 0.0 0.0 telnetd
Compare with the ps report, which follows:
# ps v 7226
PID
TTY STAT TIME PGIN SIZE
7226
- A
0:00
51 240
RSS LIM
24 32768
SIZE refers to the virtual size in KB of the data section of the process (in paging space). This number is
equal to the number of working segment pages of the process that have been touched (that is, the number
of paging-space pages that have been allocated) times 4. It must be multiplied by 4 because pages are in
4 K units and SIZE is in 1 K units. If some working segment pages are currently paged out, this number is
larger than the amount of real memory being used. The SIZE value (240) correlates with the Pgspace
number from the svmon command for private (32) plus lib data (28) in 1 K units.
RSS refers to the real memory (resident set) size in KB of the process. This number is equal to the sum of
the number of working segment and code segment pages in memory times 4. Remember that code
segment pages are shared among all of the currently running instances of the program. If 26 ksh
processes are running, only one copy of any given page of the ksh executable program would be in
memory, but the ps command would report that code segment size as part of the RSS of each instance of
the ksh program. The RSS value (24) correlates with the Inuse numbers from the svmon command for
private (6) working-storage segments, for code (0) segments, and for lib data (0) of the process in 1-K
units.
TRS refers to the size of the resident set (real memory) of text. This is the number of code segment pages
times four. As was noted earlier, this number exaggerates memory use for programs of which multiple
instances are running. This does not include the shared text of the process. The TRS value (0) correlates
Chapter 8. Monitoring and Tuning Memory Use
143
with the number of the svmon pages in the code segment (0) of the Inuse column in 1 K units. The TRS
value can be higher than the TSIZ value because other pages, such as the XCOFF header and the loader
section, may be included in the code segment.
The following calculations can be made for the values mentioned:
SIZE = 4 * Pgspace of (work lib data + work private)
RSS = 4 * Inuse of (work lib data + work private + pers code)
TRS = 4 * Inuse of (pers code)
Calculating the Minimum Memory Requirement of a Program
To calculate the minimum memory requirement of a program, the formula would be:
Total memory pages (4 KB units) = T + ( N * ( PD + LD ) ) + F
where:
T
= Number of pages for text (shared by all users)
N
= Number of copies of this program running simultaneously
PD
= Number of working segment pages in process private segment
LD
= Number of shared library data pages used by the process
F
= Number of file pages (shared by all users)
Multiply the result by 4 to obtain the number of kilobytes required. You may want to add in the kernel,
kernel extension, and shared library text segment values to this as well even though they are shared by all
processes on the system. For example, some applications like CATIA and databases use very large
shared library modules. Note that because we have only used statistics from a single snapshot of the
process, there is no guarantee that the value we get from the formula will be the correct value for the
minimum working set size of a process. To get working set size, one would need to run a tool such as the
rmss command or take many snapshots during the life of the process and determine the average values
from these snapshots (see Assessing Memory Requirements Through the rmss Command).
If we estimate the minimum memory requirement for the program pacman, shown in Finding
Memory-Leaking Programs, the formula would be:
T
= 2 (Inuse of code,/dev/lv01:12302 of pers)
PD
= 1632 (Inuse of private of work)
LD
= 12 (Inuse of lib data of work)
F
= 1 (Inuse of /dev/hd2:53289 of pers)
That is: 2 + (N * (1632+ 12)) + 1, equal to 1644 * N + 3 in 4 KB units.
Finding Memory-Leaking Programs
A memory leak is a program error that consists of repeatedly allocating memory, using it, and then
neglecting to free it. A memory leak in a long-running program, such as an interactive application, is a
serious problem, because it can result in memory fragmentation and the accumulation of large numbers of
mostly garbage-filled pages in real memory and page space. Systems have been known to run out of
page space because of a memory leak in a single program.
A memory leak can be detected with the svmon command, by looking for processes whose working
segment continually grows. A leak in a kernel segment can be caused by an mbuf leak or by a device
driver, kernel extension, or even the kernel. To determine if a segment is growing, use the svmon
command with the -i option to look at a process or a group of processes and see if any segment continues
to grow.
144
Performance Management Guide
Identifying the offending subroutine or line of code is more difficult, especially in AIXwindows applications,
which generate large numbers of malloc() and free() calls. C++ provides a HeapView Debugger for
analyzing/tuning memory usage and leaks. Some third-party programs exist for analyzing memory leaks,
but they require access to the program source code.
Some uses of the realloc() subroutine, while not actually programming errors, can have the same effect
as a memory leak. If a program frequently uses the realloc() subroutine to increase the size of a data
area, the working segment of the process can become increasingly fragmented if the storage released by
the realloc() subroutine cannot be reused for anything else.
Use the disclaim() system call and free() call to release memory that is no longer required. The
disclaim() system call must be called before the free() call. It wastes CPU time to free memory after the
last malloc() call, if the program will finish soon. When the program terminates, its working segment is
destroyed and the real memory page frames that contained working segment data are added to the free
list. The following example is a memory-leaking program; its Inuse, Pgspace, and Address Range of the
private working segment are continually growing.
An output example from AIX 5.2 follows:
# svmon -P 13548 -i 1 3
Pid
13548
Vsid
Esid
0
0
48412
2
6c01b
d
4c413
f
3040c
1
ginger :svmon
Command
pacman
Esid
0
2
d
f
1
Type
work
work
work
work
pers
Pid
13548
Vsid
0
48412
6c01b
4c413
3040c
Inuse
8535
Type Description
LPage Inuse
work kernel seg
4375
work process private
2357
work shared library text
1790
work shared library data
11
pers code,/dev/mattlv:4097
2
-P 13548 -i 1 3
Pid
13548
Vsid
0
48412
6c01b
4c413
3040c
Command
pacman
Description
LPage Inuse
kernel seg
4375
process private
2411
shared library text
1790
shared library data
11
code,/dev/mattlv:4097
2
Command
pacman
Esid Type
0 work
2 work
d work
f work
1 pers
Inuse
8589
Inuse
8599
Description
LPage
kernel seg
process private
shared library text
shared library data
code,/dev/mattlv:4097
-
Inuse
4375
2421
1790
11
2
Pin
2178
Pin
2176
2
0
0
0
Pgsp
847
Pgsp
847
0
0
0
-
Pin
2178
Pin
2176
2
0
0
0
Pin
2178
Pin
2176
2
0
0
0
Virtual 64-bit Mthrd LPage
8533
N
N
N
Virtual
4375
2357
1790
11
Pgsp
847
Pgsp
847
0
0
0
-
Virtual 64-bit Mthrd LPage
8587
N
N
N
Virtual
4375
2411
1790
11
Pgsp
847
Pgsp
847
0
0
0
-
Virtual 64-bit Mthrd LPage
8597
N
N
N
Virtual
4375
2421
1790
11
-
Assessing Memory Requirements Through the rmss Command
The rmss (Reduced-Memory System Simulator) command provides you with a means to simulate different
sizes of real memories that are smaller than your actual machine, without having to extract and replace
memory boards. Moreover, the rmss command provides a facility to run an application over a range of
memory sizes, displaying, for each memory size, performance statistics such as the response time of the
application and the amount of paging. The rmss command is designed to help you answer the question:
“How many megabytes of real memory does a system need to run the operating system and a given
Chapter 8. Monitoring and Tuning Memory Use
145
application with an acceptable level of performance?”. In the multiuser context, it is designed to help you
answer the question: “How many users can run this application simultaneously in a machine with X
megabytes of real memory?”
The main use for the rmss command is as a capacity planning tool, to determine how much memory a
workload needs. It can also be used as a problem determination tool, particularly for those cases where
having more memory degrades performance.
To determine whether the rmss command is installed and available, run the following command:
# lslpp -lI fileset_name
where fileset_name is perfagent.tools in AIX 4.3 or bos.perf.tools in AIX 5 or later.
It is important to keep in mind that the memory size simulated by the rmss command is the total size of
the machine’s real memory, including the memory used by the operating system and any other programs
that may be running. It is not the amount of memory used specifically by the application itself. Because of
the performance degradation it can cause, the rmss command can be used only by a root user or a
member of the system group.
Two Styles of Using rmss
The rmss command can be invoked in two ways:
1. To change the memory size and exit.
2. As a driver program, which executes a specified application multiple times over a range of memory
sizes and displays important statistics that describe the application’s performance at each memory
size.
The first invocation technique is useful when you want to get the look and feel of how your application
performs at a given system memory size, when your application is too complex to be expressed as a
single command, or when you want to run multiple instances of the application. The second invocation
technique is appropriate when you have an application that can be invoked as an executable program or
shell script file.
Note: Before using the rmss command, run the command /usr/samples/kernel/schedtune -h 0 to turn
off VMM memory-load control (see Tuning VMM Memory Load Control with the schedtune
Command). Otherwise, VMM memory-load control may interfere with your measurements at small
memory sizes. When your experiments are complete, reset the memory load control parameters to
the values that are typically in effect on your system (if you typically use the default parameters,
use /usr/samples/kernel/schedtune -D).
Using rmss to Change the Memory Size and Exit
To change the memory size and exit, use the -c flag. For example, to change the memory size to 128 MB,
use the following:
# rmss -c 128
The memory size is an integer or decimal fraction number of megabytes (for example, 128.25).
Additionally, the size must be between 8 MB and the amount of physical real memory in your machine.
Depending on the hardware and software configuration, the rmss command may not be able to change
the memory size to small sizes, because of the size of inherent system structures such as the kernel.
When the rmss command is unable to change to a given memory size, it displays an error message.
The rmss command reduces the effective memory size of a system by stealing free page frames from the
list of free frames that is maintained by the VMM. The stolen frames are kept in a pool of unusable frames
and are returned to the free frame list when the effective memory size is to be increased. Also, the rmss
command dynamically adjusts certain system variables and data structures that must be kept proportional
to the effective size of memory.
146
Performance Management Guide
It may take a short while (up to 15 to 20 seconds) to change the memory size. In general, the more you
want to reduce the memory size, the longer the rmss command takes to complete. When successful, the
rmss command responds with the following message:
Simulated memory size changed to
128.00 Mb.
To display the current memory size, use the -p flag, as follows:
# rmss -p
The rmss output is as follows:
Simulated memory size is
128.00 Mb.
Finally, if you want to reset the memory size to the actual memory size of the machine, use the -r flag, as
follows:
# rmss -r
No matter what the current simulated memory size, using the -r flag sets the memory size to be the
physical real memory size of the machine. A side effect of the rmss -r command is that the related
vmtune parameters are also reset to their corresponding defaults.
Because this example was run on a 256 MB machine, the rmss command responded as follows:
Simulated memory size changed to
256.00 Mb.
Note: The rmss command reports usable real memory. On machines that contain bad memory or
memory that is in use, the rmss command reports the amount of real memory as the amount of
physical real memory minus the memory that is bad or in use by the system. For example, the
rmss -r command might report:
Simulated memory size changed to 79.9062 Mb.
This could be a result of some pages being marked bad or a result of a device that is reserving
some pages for its own use (and thus not available to the user).
Using the -c, -p, and -r Flags: The -c, -p and -r flags of the rmss command have an advantage over
the other options in that they allow you to experiment with complex applications that cannot be expressed
as a single executable program or shell script file. On the other hand, the -c, -p, and -r options have a
disadvantage in that they force you to do your own performance measurements. Fortunately, you can use
the command vmstat -s to measure the paging-space activity that occurred while your application ran.
By running the command vmstat -s, running your application, then running the command vmstat -s again,
and subtracting the number of paging-space page-ins before from the number of paging-space page-ins
after, you can determine the number of paging-space page-ins that occurred while your program ran.
Furthermore, by timing your program, and dividing the number of paging-space page-ins by the program’s
elapsed run time, you can obtain the average paging-space page-in rate.
It is also important to run the application multiple times at each memory size, for two reasons:
v When changing memory size, the rmss command often clears out a lot of memory. Thus, the first time
you run your application after changing memory sizes it is possible that a substantial part of the run
time may be due to your application reading files into real memory. But, since the files may remain in
memory after your application terminates, subsequent executions of your application may result in
substantially shorter elapsed times.
v To get a feel for the average performance of the application at that memory size. It is impossible to
duplicate the system state each time your application runs. Because of this, the performance of your
application can vary significantly from run to run.
To summarize, consider the following set of steps as a desirable way to invoke the rmss command:
Chapter 8. Monitoring and Tuning Memory Use
147
while there are interesting memory sizes to investigate:
{
change to an interesting memory size using rmss -c;
run the application once as a warm-up;
for a couple of iterations:
{
use vmstat -s to get the "before" value of paging-space page ins;
run the application, while timing it;
use vmstat -s to get the "after" value of paging-space page ins;
subtract the "before" value from the "after" value to get the
number of page ins that occurred while the application ran;
divide the number of paging-space page ins by the response time
to get the paging-space page-in rate;
}
}
run rmss -r to restore the system to normal memory size (or reboot)
The calculation of the (after - before) paging I/O numbers can be automated by using the vmstatit script
described in Determining Whether the Problem is Related to Disk or Memory.
Using rmss to Run a Command over a Range of Memory Sizes
The -s, -f, -d, -n, and -o flags are used in combination to invoke the rmss command as a driver program.
As a driver program, the rmss command executes a specified application over a range of memory sizes
and displays statistics describing the application’s performance at each memory size. The syntax for this
invocation style of the rmss command is as follows:
rmss [ -s smemsize ] [ -f fmemsize ] [ -d memdelta ]
[ -n numiterations ] [ -o outputfile ] command
Each of the following flags is discussed in detail below. The -s, -f, and -d flags are used to specify the
range of memory sizes.
-n
This flag is used to specify the number of times to run and measure the command at each
memory size.
-o
This flag is used to specify the file into which to write the rmss report, while command is the
application that you wish to run and measure at each memory size.
-s
This flag specifies the starting size.
-f
This flag specifies the final size.
-d
This flag specifies the difference between sizes.
All values are in integer or decimal fractions of megabytes. For example, if you wanted to run and
measure a command at sizes 256, 224, 192, 160 and 128 MB, you would use the following combination:
-s 256 -f 128 -d 32
Likewise, if you wanted to run and measure a command at 128, 160, 192, 224, and 256 MB, you would
use the following combination:
-s 128 -f 256 -d 32
If the -s flag is omitted, the rmss command starts at the actual memory size of the machine. If the -f flag
is omitted, the rmss command finishes at 8 MB. If the -d flag is omitted, there is a default of 8 MB
between memory sizes.
What values should you choose for the -s, -f, and -d flags? A simple choice would be to cover the memory
sizes of systems that are being considered to run the application you are measuring. However, increments
of less than 8 MB can be useful, because you can get an estimate of how much space you will have when
you settle on a given size. For instance, if a given application thrashes at 120 MB but runs without
page-ins at 128 MB, it would be useful to know where within the 120 to 128 MB range the application
starts thrashing. If it starts at 127 MB, you may want to consider configuring the system with more than
148
Performance Management Guide
128 MB of memory, or you may want to try to modify the application so that there is more space. On the
other hand, if the thrashing starts at 121 MB, you know that you have enough space with a 128 MB
machine.
The -n flag is used to specify how many times to run and measure the command at each memory size.
After running and measuring the command the specified number of times, the rmss command displays
statistics describing the average performance of the application at that memory size. To run the command
3 times at each memory size, you would use the following:
-n 3
If the -n flag is omitted, the rmss command determines during initialization how many times your
application must be run to accumulate a total run time of 10 seconds. The rmss command does this to
ensure that the performance statistics for short-running programs will not be significantly skewed by
outside influences, such as daemons.
Note: If you are measuring a very brief program, the number of iterations required to accumulate 10
seconds of CPU time can be very large. Because each execution of the program takes a minimum
of about 2 elapsed seconds of rmss overhead, specify the -n parameter explicitly for short
programs.
What are good values to use for the -n flag? If you know that your application takes much more than 10
seconds to run, you can specify -n 1 so that the command is run twce, but measured only once at each
memory size. The advantage of using the -n flag is that the rmss command will finish sooner because it
will not have to spend time during initialization to determine how many times to run your program. This can
be particularly valuable when the command being measured is long-running and interactive.
It is important to note that the rmss command always runs the command once at each memory size as a
warm-up before running and measuring the command. The warm-up is needed to avoid the I/O that occurs
when the application is not already in memory. Although such I/O does affect performance, it is not
necessarily due to a lack of real memory. The warm-up run is not included in the number of iterations
specified by the -n flag.
The -o flag is used to specify a file into which to write the rmss report. If the -o flag is omitted, the report
is written into the file rmss.out.
Finally, command is used to specify the application to be measured. It can be an executable program or
shell script, with or without command-line arguments. There are some limitations on the form of the
command however. First, it cannot contain the redirection of input or output (for example, foo > output or
foo < input). This is because the rmss command treats everything to the right of the command name as
an argument to the command. To redirect, place the command in a shell script file.
Usually, if you want to store the rmss output in a specific file, use the -o option. If you want to redirect the
standard output of the rmss command (for example, to concatenate it to the end of an existing file) then
use the Korn shell to enclose the rmss invocation in parentheses, as follows:
# (rmss -s 24 -f 8 foo) >> output
Interpreting rmss Results
The Report Generated for the foo Program example was produced by running the rmss command on an
actual application program, although the name of the program has been changed to foo for anonymity.
The specific command that would have been used to generate the report is as follows:
# rmss -s 16 -f 8 -d 1 -n 1 -o rmss.out foo
Chapter 8. Monitoring and Tuning Memory Use
149
Report Generated for the foo Program
Hostname: widgeon.austin.ibm.com
Real memory size:
16.00 Mb
Time of day: Thu Jan 6 19:04:04 2000
Command: foo
Simulated memory size initialized to
16.00 Mb.
Number of iterations per memory size = 1 warm-up + 1 measured = 2.
Memory size Avg. Pageins Avg. Response Time
Avg. Pagein Rate
(megabytes)
(sec.)
(pageins / sec.)
----------------------------------------------------------------16.00
115.0
123.9
0.9
15.00
112.0
125.1
0.9
14.00
179.0
126.2
1.4
13.00
81.0
125.7
0.6
12.00
403.0
132.0
3.1
11.00
855.0
141.5
6.0
10.00
1161.0
146.8
7.9
9.00
1529.0
161.3
9.5
8.00
2931.0
202.5
14.5
The report consists of four columns. The leftmost column gives the memory size, while the Avg. Pageins
column gives the average number of page-ins that occurred when the application was run at that memory
size. It is important to note that the Avg. Pageins column refers to all page-in operations, including code,
data, and file reads, from all programs, that completed while the application ran. The Avg. Response Time
column gives the average amount of time it took the application to complete, while the Avg. Pagein Rate
column gives the average rate of page-ins.
Concentrate on the Avg. Pagein Rate column. From 16 MB to 13 MB, the page-in rate is relatively small
(< 1.5 page-ins per second). However, from 13 MB to 8 MB, the page-in rate grows gradually at first, and
then rapidly as 8 MB is reached. The Avg. Response Time column has a similar shape: relatively flat at
first, then increasing gradually, and finally increasing rapidly as the memory size is decreased to 8 MB.
Here, the page-in rate actually decreases when the memory size changes from 14 MB (1.4 page-ins per
second) to 13 MB (0.6 page-ins per second). This is not cause for alarm. In an actual system, it is
impossible to expect the results to be perfectly smooth. The important point is that the page-in rate is
relatively low at both 14 MB and 13 MB.
Finally, you can make a couple of deductions from the report. First, if the performance of the application is
deemed unacceptable at 8 MB (as it probably would be), then adding memory would enhance
performance significantly. Note that the response time rises from approximately 124 seconds at 16 MB to
202 seconds at 8 MB, an increase of 63 percent. On the other hand, if the performance is deemed
unacceptable at 16 MB, adding memory will not enhance performance much, because page-ins do not
slow the program appreciably at 16 MB.
Report for a 16 MB Remote Copy
The following example illustrates a report that was generated (on a client machine) by running the rmss
command on a command that copied a 16 MB file from a remote (server) machine through NFS.
Hostname: xray.austin.ibm.com
Real memory size:
48.00 Mb
Time of day: Mon Jan 10 18:16:42 2000
Command: cp /mnt/a16Mfile /dev/null
Simulated memory size initialized to
48.00 Mb.
Number of iterations per memory size = 1 warm-up + 4 measured = 5.
Memory size
(megabytes)
150
Avg. Pageins
Avg. Response Time Avg. Pagein Rate
(sec.)
(pageins / sec.)
Performance Management Guide
----------------------------------------------------------------48.00
0.0
2.7
0.0
40.00
0.0
2.7
0.0
32.00
0.0
2.7
0.0
24.00
1520.8
26.9
56.6
16.00
4104.2
67.5
60.8
8.00
4106.8
66.9
61.4
Note that the response time and page-in rate in this report start relatively low, rapidly increase at a
memory size of 24 MB, and then reach a plateau at 16 and 8 MB. This report shows the importance of
choosing a wide range of memory sizes when you use the rmss command. If this user had only looked at
memory sizes from 24 MB to 8 MB, he or she might have missed an opportunity to configure the system
with enough memory to accommodate the application without page-ins.
Hints for Using the -s, -f, -d, -n, and -o Flags
One helpful feature of the rmss command, when used in this way, is that it can be terminated (by the
interrupt key, Ctrl-C by default) without destroying the report that has been written to the output file. In
addition to writing the report to the output file, this causes the rmss command to reset the memory size to
the physical memory size of the machine.
You can run the rmss command in the background, even after you have logged out, by using the nohup
command. To do this, precede the rmss command by the nohup command, and follow the entire
command with an & (ampersand), as follows:
# nohup rmss -s 48 -f 8 -o foo.out foo &
Guidelines to Consider When Running the rmss Command
No matter which rmss invocation style you are using, it is important to re-create the end-user environment
as closely as possible. For instance, are you using the same model CPU, the same model disks, the same
network? Will the users have application files mounted from a remote node via NFS or some other
distributed file system? This last point is particularly important, because pages from remote files are
treated differently by the VMM than pages from local files.
Likewise, it is best to eliminate any system activity that is not related to the desired system configuration or
the application you are measuring. For instance, you do not want to have people working on the same
machine as the rmss command unless they are running part of the workload you are measuring.
Note: You cannot run multiple invocations of the rmss command simultaneously.
When you have completed all runs of the rmss command, it is best to shut down and reboot the system.
This will remove all changes that the rmss command has made to the system and will restore the VMM
memory load control parameters to their typical settings.
Tuning VMM Memory Load Control with the schedtune Command
The VMM memory load control facility, described in VMM Memory Load Control Facility, protects an
overloaded system from thrashing.
For early versions of the operating system, if a large number of processes hit the system at the same
time, memory became overcommitted and thrashing occurred, causing performance to degrade rapidly. A
memory-load control mechanism was developed that could detect thrashing. Certain parameters affect the
function of the load control mechanism.
With the schedtune command, the root user can affect the criteria used to determine thrashing, the
criteria used to determine which processes to suspend, the length of time to wait after thrashing ends
before reactivating processes, the minimum number of processes exempt from suspension, or reset values
to the defaults. To determine whether the schedtune command is installed and available, run the following
command:
Chapter 8. Monitoring and Tuning Memory Use
151
# lslpp -lI bos.adt.samples
Memory Load Control Tuning
Memory load control is intended to smooth out infrequent peaks in load that might otherwise cause the
system to thrash. It trades multiprogramming for throughput and is not intended to act continuously in a
configuration that has too little RAM to handle its normal workload. The design was made for batch jobs
and is not very discriminating. The AIX Workload Manager provides a better solution to protect critical
tasks.
The correct solution to a fundamental, persistent RAM shortage is to add RAM, not to experiment with
memory load control in an attempt to trade off response time for memory. The situations in which the
memory-load-control facility may really need to be tuned are those in which there is more RAM, not less
than the defaults were chosen for. An example would be configurations in which the defaults are too
conservative.
You should not change the memory load control parameter settings unless your workload is consistent and
you believe the default parameters are ill-suited to your workload.
The default parameter settings shipped with the system are always in force unless changed. The default
values of these parameters have been chosen to ″fail safe″ across a wide range of workloads. Changed
parameters last only until the next system boot. All memory load control tuning activities must be done by
the root user. The system administrator can use the schedtune command to change the parameters to
tune the algorithm to a particular workload or to disable it entirely. The source and object code of the
schedtune command are in /usr/samples/kernel.
Note: The schedtune command is in the samples directory because it is very VMM-implementation
dependent. The schedtune code that accompanies each release of the operating system is tailored
specifically to the VMM in that release. Running the schedtune command from one release on a
different release might result in an operating-system failure. It is also possible that the functions of
the schedtune command may change from release to release. Do not propagate shell scripts or
/etc/inittab entries that include the schedtune command to a new release without checking the
schedtune documentation for the new release to make sure that the scripts will still have the
desired effect.
The schedtune -? command provides a terse description of the flags and options. A schedtune invocation
with no flags displays the current parameter settings, as follows:
# /usr/samples/kernel/schedtune
THRASH
SUSP
-h
-p
-m
-w
SYS PROC MULTI
WAIT
0
4
2
1
CLOCK
-c
%usDELTA
100
FORK
-e
-f
GRACE
TICKS
2
10
SCHED_FIFO2
IDLE MIGRATION
-a
-b
AFFINITY_LIM BARRIER/16
7
4
SCHED
-d
-r
SCHED_D SCHED_R
16
16
-t
-s
TIMESLICE MAXSPIN
1
16384
FIXED_PRI
-F
GLOBAL(1)
0
The first five parameters specify the thresholds for the memory load control algorithm. These parameters
set rates and thresholds for the algorithm. If the algorithm shows that RAM is overcommitted, the PROC
(-p), MULTI (-m), WAIT (-w), and GRACE (-e) values are used. Otherwise, these values are ignored. If
memory load control is disabled, these latter values are not used.
After a tuning experiment, memory load control can be reset to its default characteristics by executing the
command schedtune -D.
152
Performance Management Guide
When you have determined the number of processes that ought to be able to run in your system during
periods of peak activity, you can add a schedtune command at the end of the /etc/inittab file, which
ensures that it will be run each time the system is booted, overriding the defaults that would otherwise
take effect with a reboot. For example, an appropriate /etc/inittab line for raising the minimum level of
multiprogramming to 4 on an AIX Version 4 system would be as follows:
schedtune:2:wait:/usr/samples/kernel/schedtune -m 4
Remember, do not propagated this line to a new release of the operating system without a check of the
operating system documentation.
While it is possible to vary other parameters that control the suspension rate of processes and the criteria
by which individual processes are selected for suspension, it is impossible to predict with any confidence
the effect of such changes on a particular configuration and workload. Great caution should be exercised if
memory load control parameter adjustments other than those discussed here are considered.
The h Parameter
The h parameter controls the threshold defining memory overcommitment. Memory load control attempts
to suspend processes when this threshold is exceeded during any one-second period. The threshold is a
relationship between two direct measures: the number of pages written to paging space in the last second
(po) and the number of page steals occurring in the last second (fr). You can see both these values in the
vmstat output. The number of page writes is usually much less than the number of page steals. Memory
is overcommitted when the following is true:
po/fr > 1/h or po*h > fr
The command schedtune -h 0 effectively disables memory load control. If a system has at least 128 MB
of memory, the default value is 0, otherwise the default value is 6. With at least 128 MB of RAM, the
normal VMM algorithms usually correct thrashing conditions on the average more efficiently than by using
memory load control.
In some specialized situations, it might be appropriate to disable memory load control from the outset. For
example, if you are using a terminal emulator with a time-out feature to simulate a multiuser workload,
memory load control intervention may result in some responses being delayed long enough for the
process to be killed by the time-out feature. Another example is, if you are using the rmss command to
investigate the effects of reduced memory sizes, disable memory load control to avoid interference with
your measurement.
If disabling memory load control results in more, rather than fewer, thrashing situations (with
correspondingly poorer responsiveness), then memory load control is playing an active and supportive role
in your system. Tuning the memory load control parameters then may result in improved performance or
you may need to add RAM.
A lower value of h raises the thrashing detection threshold; that is, the system is allowed to come closer to
thrashing before processes are suspended. Regardless of the system configuration, when the above po/fr
fraction is low, thrashing is unlikely.
To alter the threshold to 4, enter the following:
# /usr/samples/kernel/schedtune -h 4
In this way, you permit the system to come closer to thrashing before the algorithm starts suspending
processes.
The p Parameter
The p parameter determines whether a process is eligible for suspension and is used to set a threshold for
the ratio of two measures that are maintained for every process: the number of repages (r) and the
number of page faults that the process has accumulated in the last second (f). A high ratio of repages to
Chapter 8. Monitoring and Tuning Memory Use
153
page faults means the individual process is thrashing. A process is considered eligible for suspension (it is
thrashing or contributing to overall thrashing) when the following is true:
r/f > 1/p or r*p > f
The default value of p is 4, meaning that a process is considered to be thrashing (and a candidate for
suspension) when the fraction of repages to page faults over the last second is greater than 25 percent. A
low value of p results in a higher degree of individual process thrashing being allowed before a process is
eligible for suspension.
To disable processes from being suspended by the memory load control, do the following:
# /usr/samples/kernel/schedtune -p 0
Note that fixed-priority processes and kernel processes are exempt from being suspended.
The m Parameter
The m parameter determines a lower limit for the degree of multiprogramming, which is defined as the
number of active processes. Active processes are those that can be run and are waiting for page I/O.
Processes that are waiting for events and processes suspended are not considered active nor is the wait
process considered active.
Setting the minimum multiprogramming level, m, effectively keeps m processes from being suspended.
Suppose a system administrator knows that at least ten processes must always be resident and active in
RAM for successful performance, and suspects that memory load control was too vigorously suspending
processes. If the command schedtune -m 10 was issued, the system would never suspend so many
processes that fewer than ten were competing for memory. The m parameter does not count:
v The kernel processes
v Processes that have been pinned in RAM with the plock() system call
v Fixed-priority processes with priority values less than 60
v Processes awaiting events
The system default of m=2 ensures that the kernel, all pinned processes, and two user processes will
always be in the set of processes competing for RAM.
While m=2 is appropriate for a desktop, single-user configuration, it is frequently too small for larger,
multiuser, or server configurations with large amounts of RAM.
If the system you are installing is larger than 32 MB, but less than 128 MB, and is expected to support
more than five active users at one time, consider raising the minimum level of multiprogramming of the
VMM memory-load-control mechanism.
As an example, if your conservative estimate is that four of your most memory-intensive applications
should be able to run simultaneously, leaving at least 16 MB for the operating system and 25 percent of
real memory for file pages, you could increase the minimum multiprogramming level from the default of 2
to 4 with the following command:
# /usr/samples/kernel/schedtune -m 4
On these systems, setting m to 4 or 6 may result in the best performance. Lower values of m, while
allowed, mean that at times as few as one user process may be active.
When the memory requirements of the thrashing application are known, the m value can be suitably
chosen. Suppose thrashing is caused by numerous instances of one application of size M. Given the
system memory size N, the m parameter should be set to a value close to N/M. Setting m too low would
unnecessarily limit the number of processes that could be active at the same time.
154
Performance Management Guide
The w Parameter
The w parameter controls the number of one-second intervals during which the po/fr fraction (explained in
the The h Parameter) must remain below 1/h before suspended processes are reactivated. The default
value of one second is close to the minimum value allowed, which is zero. A value of one second
aggressively attempts to reactivate processes as soon as a one-second safe period has occurred. Large
values of w run the risk of unnecessarily poor response times for suspended processes while the
processor is idle for lack of active processes to run.
To alter the wait time to reactivate processes after two seconds, enter the following:
# /usr/samples/kernel/schedtune -w 2
The e Parameter
Each time a suspended process is reactivated, it is exempt from suspension for a period of e elapsed
seconds. This ensures that the high cost (in disk I/O) of paging in the pages of a suspended process
results in a reasonable opportunity for progress. The default value of e is 2 seconds.
To alter this parameter, enter the following:
# /usr/samples/kernel/schedtune -e 1
Suppose thrashing is caused occasionally by an application that uses lots of memory but runs for about T
seconds. The default system setting for e (2 seconds) would probably cause this application swapping in
and out T/2 times on a busy system. In this case, resetting e to a longer time would help this application to
progress. System performance would improve when this offending application is pushed through quickly.
Tuning VMM Page Replacement with the vmtune Command
The memory management algorithm, discussed in Real-Memory Management, tries to keep the size of the
free list and the percentage of real memory occupied by persistent segment pages within specified
bounds. These bounds can be altered with the vmtune command, which can only be run by the root user.
Changes made by this tool remain in effect until the next reboot of the system. To determine whether the
vmtune command is installed and available, run the following command:
# lslpp -lI bos.adt.samples
Note: The vmtune command is in the samples directory because it is very VMM-implementation
dependent. The vmtune code that accompanies each release of the operating system is tailored
specifically to the VMM in that release. Running the vmtune command from one release on a
different release might result in an operating-system failure. It is also possible that the functions of
vmtune may change from release to release. Do not propagate shell scripts or /etc/inittab entries
that include the vmtune command to a new release without checking the vmtune documentation
for the new release to make sure that the scripts will still have the desired effect.
Executing the vmtune command on AIX 4.3.3 with no options results in the following output:
# /usr/samples/kernel/vmtune
vmtune: current values:
-p
-P
-r
-R
minperm maxperm minpgahead maxpgahead
52190
208760
2
8
-f
minfree
120
-F
maxfree
128
-N
-W
pd_npages maxrandwrt
524288
0
-M
-w
-k
-c
-b
-B
-u
-l
-d
maxpin npswarn npskill numclust numfsbufs hd_pbuf_cnt lvm_bufcnt lrubucket defps
209581
4096
1024
-s
sync_release_ilock
0
1
-n
nokillroot
0
93
-S
v_pinshm
0
96
9
131072
1
-h
strict_maxperm
0
Chapter 8. Monitoring and Tuning Memory Use
155
number of valid memory pages = 261976
maximum pinable=80.0% of real memory
number of file memory pages = 19772
maxperm=79.7% of real memory
minperm=19.9% of real memory
numperm=7.5% of real memory
The output shows the current settings for all the parameters.
Executing the vmtune command on AIX 5.2 results in the following output:
# /usr/samples/kernel/vmtune
vmtune: current values:
-M
-p
-P
-t
maxpin% minperm% maxperm% maxclient%
80.1
20.0
80.0
80.0
-f
minfree
120
-F
maxfree
128
-l
lrubucket
131072
-r
-R
-N
-W
-w
-k
-c
minpgahead maxpgahead pd_npages maxrandwrt npswarn npskill numclust
2
8
65536
0
4096
1024
1
-b
-B
numfsbufs hd_pbuf_cnt
186
448
-u
lvm_bufcnt
9
-d
defps
1
-s
sync_release_ilock
0
-S
v_pinshm
0
-L
-g
-h
-n
-j
lgpg_regions lgpg_size strict_maxperm nokilluid j2_nPagesPerWriteBehindCluster
0
0
0
0
8
-J
-z
-Z
j2_maxRandomWrite j2_nRandomCluster j2_nBufferPerPagerDevice
0
0
512
-q
-Q
-V
-i
j2_minPageReadAhead j2_maxPageReadAhead num_spec_dataseg spec_dataseg_int
2
8
0
512
-y
mem_affinity_on
0
maxpin pages
=
memory size pages
=
remote pageouts scheduled
file pages
compressed pages
=
client pages
=
838861
1048576
0
7690
0
0
minperm pages
=
199609
maxperf pages
=
798436
maxclient pages =
798436
numperm =
0.7% of real memory
numcompress= 0.0% of real memory
numclient=
0.0% of real memory
Choosing minfree and maxfree Settings
The purpose of the free list is to keep track of real-memory page frames released by terminating
processes and to supply page frames to requestors immediately, without forcing them to wait for page
steals and the accompanying I/O to complete. The minfree limit specifies the free-list size below which
page stealing to replenish the free list is to be started. The maxfree parameter is the size above which
stealing will end.
The objectives in tuning these limits are to ensure that:
v Any activity that has critical response-time objectives can always get the page frames it needs from the
free list.
v The system does not experience unnecessarily high levels of I/O because of premature stealing of
pages to expand the free list.
The default value of minfree and maxfree depend on the memory size of the machine. The default value
of maxfree is determined by this formula:
maxfree = minimum (# of memory pages/128, 128)
156
Performance Management Guide
By default the minfree value is the value of maxfree - 8. However, the difference between minfree and
maxfree should always be equal to or greater than maxpgahead. Or in other words, the value of maxfree
should always be greater than or equal to minfree plus the size of maxpgahead. The minfree/maxfree
values will be different if there is more than one memory pool. Memory pools were introduced in AIX 4.3.3
for MP systems with large amounts of RAM. Each memory pool will have its own minfree/maxfree which
are determined by the previous formulas, but the minfree/maxfree values shown by the vmtune
command will be the sum of the minfree/maxfree for all memory pools.
Remember, that minfree pages in some sense are wasted, because they are available, but not in use. If
you have a short list of the programs you want to run fast, you can investigate their memory requirements
with the svmon command (see Determining How Much Memory Is Being Used), and set minfree to the
size of the largest. This technique risks being too conservative because not all of the pages that a process
uses are acquired in one burst. At the same time, you might be missing dynamic demands that come from
programs not on your list that may lower the average size of the free list when your critical programs run.
A less precise but more comprehensive tool for investigating an appropriate size for minfree is the vmstat
command. The following is a portion of a vmstat command output obtained while running an C
compilation on an otherwise idle system.
# vmstat 1
kthr
memory
page
faults
cpu
----- ----------- ------------------------ ------------ ----------r b
avm
fre re pi po fr
sr cy in
sy cs us sy id wa
0 0 3085
118
0
0
0
0
0
0 115
2 19 0 0 99 0
0 0 3086
117
0
0
0
0
0
0 119 134 24 1 3 96 0
2 0 3141
55
2
0
6 24
98
0 175 223 60 3 9 54 34
0 1 3254
57
0
0
6 176 814
0 205 219 110 22 14 0 64
0 1 3342
59
0
0 42 104 249
0 163 314 57 43 16 0 42
1 0 3411
78
0
0 49 104 169
0 176 306 51 30 15 0 55
1 0 3528
160
1
0 10 216 487
0 143 387 54 50 22 0 27
1 0 3627
94
0
0
0 72 160
0 148 292 79 57 9 0 34
1 0 3444
327
0
0
0 64 102
0 132 150 41 82 8 0 11
1 0 3505
251
0
0
0
0
0
0 128 189 50 79 11 0 11
1 0 3550
206
0
0
0
0
0
0 124 150 22 94 6 0 0
1 0 3576
180
0
0
0
0
0
0 121 145 30 96 4 0 0
0 1 3654
100
0
0
0
0
0
0 124 145 28 91 8 0 1
1 0 3586
208
0
0
0 40
68
0 123 139 24 91 9 0 0
Because the compiler has not been run recently, the code of the compiler itself must be read in. All told,
the compiler acquires about 2 MB in about 6 seconds. On this 32 MB system, maxfree is 64 and minfree
is 56. The compiler almost instantly drives the free list size below minfree, and several seconds of rapid
page-stealing activity take place. Some of the steals require that dirty working segment pages be written to
paging space, which shows up in the po column. If the steals cause the writing of dirty permanent segment
pages, that I/O does not appear in the vmstat report (unless you have directed the vmstat command to
report on the I/O activity of the physical volumes to which the permanent pages are being written).
This example describes a fork() and exec() environment (not an environment where a process is long
lived, such as in a database) and is not intended to suggest that you set minfree to 500 to accommodate
large compiles. It suggests how to use the vmstat command to identify situations in which the free list has
to be replenished while a program is waiting for space. In this case, about 2 seconds were added to the
compiler execution time because there were not enough page frames immediately available. If you
observe the page frame consumption of your program, either during initialization or during normal
processing, you will soon have an idea of the number page frames that need to be in the free list to keep
the program from waiting for memory.
If we concluded from the example above that minfree needed to be 128, and we had set maxpgahead to
16 to improve sequential performance, we would use the following vmtune command:
# /usr/samples/kernel/vmtune -f 128 -F 144
Chapter 8. Monitoring and Tuning Memory Use
157
Tuning Memory Pools
In operating system versions later than AIX 4.3.3, the vmtune -m number_of_memory_pools command
allows you to change the number of memory pools that are configured at system boot time. The -m flag is
therefore not a dynamic change. The change is written to the kernel file if it is an MP kernel (the change is
not allowed on a UP kernel). A value of 0 restores the default number of memory pools.
By default, the vmtune -m command writes to the /usr/lib/boot/unix_mp file (/usr/lib/boot/unix_64 if
running on a 64–bit kernel), but this can be changed with the command vmtune -U path_to_unix_file.
Before changing the kernel file, the vmtune command saves the original file as name_of_original_file.sav.
Tuning lrubucket to Reduce Memory Scanning Overhead
Tuning lrubucket can reduce scanning overhead on large memory systems. In AIX 4.3, a new parameter
lrubucket was added. The page-replacement algorithm scans memory frames looking for a free frame.
During this scan, reference bits of pages are reset, and if a free frame has not been found, a second scan
is done. In the second scan, if the reference bit is still off, the frame will be used for a new page (page
replacement).
On large memory systems, there may be too many frames to scan, so now memory is divided up into
buckets of frames. The page-replacement algorithm will scan the frames in the bucket and then start over
on that bucket for the second scan before moving on to the next bucket. The default number of frames in
this bucket is 131072 or 512 MB of RAM. The number of frames is tunable with the command vmtune -l,
and the value is in 4 K frames.
Choosing minperm and maxperm Settings
The operating system takes advantage of the varying requirements for real memory by leaving in memory
pages of files that have been read or written. If the file pages are requested again before their page
frames are reassigned, this technique saves an I/O operation. These file pages may be from local or
remote (for example, NFS) file systems.
The ratio of page frames used for files versus those used for computational (working or program text)
segments is loosely controlled by the minperm and maxperm values:
v If percentage of RAM occupied by file pages rises above maxperm, page-replacement steals only file
pages.
v If percentage of RAM occupied by file pages falls below minperm, page-replacement steals both file
and computational pages.
v If percentage of RAM occupied by file pages is between minperm and maxperm, page-replacement
steals only file pages unless the number of file repages is higher than the number of computational
repages.
In a particular workload, it might be worthwhile to emphasize the avoidance of file I/O. In another
workload, keeping computational segment pages in memory might be more important. To understand what
the ratio is in the untuned state, we use the vmtune command with no arguments.
# /usr/samples/kernel/vmtune
vmtune: current values:
-p
-P
-r
-R
minperm maxperm minpgahead maxpgahead
209508
838032
2
8
-f
minfree
120
-F
maxfree
128
-N
-W
pd_npages maxrandwrt
524288
0
-M
-w
-k
-c
-b
-B
-u
-l
-d
maxpin npswarn npskill numclust numfsbufs hd_pbuf_cnt lvm_bufcnt lrubucket defps
838852
4096
1024
1
186
160
9
131072
1
-s
sync_release_ilock
0
158
-n
nokilluid
0
-S
v_pinshm
0
Performance Management Guide
-L
lgpa_regions
0
-g
lgpg_size
0
-h
strict_maxperm
0
-t
maxclient
838032
number of valid memory pages = 1048565
maximum pinable = 80.0% of real memory
number of file memory pages = 6668
number of compressed memory pages = 0
number of client memory pages = 0
# of remote pgs sched-pageout = 0
maxperm=79.9% of real memory
minperm=20.0% of real memory
numperm=0.6% of real memory
compressed=0.0% of real memory
numclient=0.0% of real memory
maxcllient=79.9% of real memory
The default values are calculated by the following algorithm:
minperm (in pages) = ((number of memory frames) - 1024) * .2
maxperm (in pages) = ((number of memory frames) - 1024) * .8
The numperm value gives the number of file pages in memory, 19772. This is 7.5 percent of real memory.
If we know that our workload makes little use of recently read or written files, we may want to constrain
the amount of memory used for that purpose. The following command:
# /usr/samples/kernel/vmtune -p 15 -P 50
sets minperm to 15 percent and maxperm to 50 percent of real memory. This would ensure that the VMM
would steal page frames only from file pages when the ratio of file pages to total memory pages exceeded
50 percent. This should reduce the paging to page space with no detrimental effect on the persistent
storage. The maxperm value is not a strict limit, it is only considered when the VMM needs to perform
page replacement. Because of this, it is usually safe to reduce the maxperm value on most systems.
On the other hand, if our application frequently references a small set of existing files (especially if those
files are in an NFS-mounted file system), we might want to allow more space for local caching of the file
pages by using the following command:
# /usr/samples/kernel/vmtune -p 30 -P 90
NFS servers that are used mostly for reads with large amounts of RAM can benefit from increasing the
value of maxperm. This allows more pages to reside in RAM so that NFS clients can access them without
forcing the NFS server to retrieve the pages from disk again.
Another example would be a program that reads 1.5 GB of sequential file data into the working storage of
a system with 2 GB of real memory. You may want to set maxperm to 50 percent or less, because you do
not need to keep the file data in memory.
Placing a Hard Limit on Persistent File Cache with strict_maxperm
Starting with AIX 4.3.3, a new vmtune option (-h) called strict_maxperm has been added. This option,
when set to 1, places a hard limit on how much memory is used for a persistent file cache by making the
maxperm value be the upper limit for this file cache. When the upper limit is reached, the least recently
used (LRU) is performed on persistent pages.
Attention: The strict_maxperm option should only be enabled for those cases that require a hard limit
on the persistent file cache. Improper use of strict_maxperm can cause unexpected system behavior
because it changes the VMM method of page replacement.
Placing a Hard Limit on Enhanced JFS File System Cache with
maxclient
The enhanced JFS file system uses client pages for its buffer cache, which are not affected by the
maxperm and minperm threshold values. To establish hard limits on enhanced JFS file system cache,
you can tune the maxclient parameter. This parameter represents the maximum number of client pages
Chapter 8. Monitoring and Tuning Memory Use
159
that can be used for buffer cache. To change this value, you can use the vmtune command with the -t
option. The value for maxclient will be shown as a percentage of real memory.
After the maxclient threshold is reached, LRU will begin to steal client pages that have not been
referenced recently. If not enough client pages can be stolen, the LRU might replace other types of pages.
By reducing the value for maxclient, you help prevent Enhanced JFS file-page accesses from causing
LRU to replace working storage pages, minimizing paging from paging space. The maxclient parameter
also affects NFS clients and compressed pages. Also note that maxclient should generally be set to a
value that is less than or equal to maxperm, particularly in the case where strict_maxperm is enabled.
Tuning Paging-Space Thresholds
If available paging space depletes to a low level, the operating system attempts to release resources by
first warning processes to release paging space and finally by killing processes if there still is not enough
paging space available for the current processes.
Choosing npswarn and npskill Settings
The npswarn and npskill thresholds are used by the VMM to determine when to first warn processes and
eventually when to kill processes.
These two parameters can be set through the vmtune command:
npswarn (-w)
Specifies the number of free paging-space pages at which the operating system begins sending
the SIGDANGER signal to processes. If the npswarn threshold is reached and a process is
handling this signal, the process can choose to ignore it or do some other action such as exit or
release memory using the disclaim() subroutine. The default value in AIX Version 4 is determined
by the following formula:
npswarn = maximum (512, 4*npskill)
The value of npswarn must be greater than zero and less than the total number of paging-space
pages on the system. It can be changed with the command vmtune -w.
npskill (-k)
Specifies the number of free paging-space pages at which the operating system begins killing
processes. If the npskill threshold is reached, a SIGKILL signal is sent to the youngest process.
Processes that are handling SIGDANGER or processes that are using the early page-space
allocation (paging space is allocated as soon as memory is requested) are exempt from being
killed. The formula to determine the default value of npskill is as follows:
npskill = maximum (64, number_of_paging_space_pages/128)
The npskill value must be greater than zero and less than the total number of paging space
pages on the system. It can be changed with the command vmtune -k.
nokillroot and nokilluid (-n)
By setting the nokillroot option to 1 with the command vmtune -n 1, processes owned by root will
be exempt from being killed when the npskill threshold is reached. This option is only available in
AIX 4.3.3 and 4.3.3.1.
By setting the nokilluid option to a nonzero value with the command vmtune -n, user IDs lower
than this value will be exempt from being killed because of low page-space conditions. This option
is only available in operating system version 4.3.3.2 and later.
Tuning the fork() Retry Interval Parameter with schedtune
If a process cannot be forked due to a lack of paging-space pages, the scheduler will retry the fork five
times. In between each retry, the scheduler will delay for a default of 10 clock ticks.
160
Performance Management Guide
The -f parameter of the schedtune command specifies the number of clock ticks to wait before retrying a
failed fork() call. For example, if a fork() subroutine call fails because there is not enough space available
to create a new process, the system retries the call after waiting the specified number of clock ticks. The
default value is 10, and because there is one clock tick every 10 ms, the system would retry the fork() call
every 100 ms.
If the paging space is low only due to brief, sporadic workload peaks, increasing the retry interval (such as
in the following) might allow processes to delay long enough to be released:
# /usr/samples/kernel/schedtune -f 15
In this way, when the system retries the fork() call, there is a higher chance of success because some
processes might have finished their execution and, consequently, released pages from paging space.
Choosing a Page Space Allocation Method
Three page-space allocation policies are available:
v Late Page Space Allocation (LPSA)
v Early Page Space Allocation (EPSA)
v Deferred Page Space Allocation (DPSA)
Late Page Space Allocation
Prior to AIX 4.3.2, the operating system by default used a late page-space allocation policy which means
that a paging space page is only allocated when it is actually touched. Being touched means the page was
modified somehow (for example, with the bzero() subroutine or if page was requested by the calloc()
subroutine or the page was initialized to some value). With the late policy, paging space slots are allocated
if RAM pages are touched, but the pages are not assigned to a particular process until that process wants
to page out. Therefore, there is no guarantee that a process will always have sufficient paging space
available if it needed to page out because some other process can start later and consume all of the
paging space.
Early Page Space Allocation
If you want to ensure that a process will not be killed due to low paging conditions, this process can
preallocate paging space by using the Early Page Space Allocation policy. This is done by setting an
environment variable called PSALLOC to the value of early. This can be done from within the process or
at the command line (PSALLOC=early command). When the process uses the malloc() subroutine to
allocate memory, this memory will now have paging-space disk blocks reserved for this process, that is,
they are reserved for this process so that there is a guarantee that if the process needed to page out,
there will always be paging space slots available for it. If using early policy and if CPU savings is a
concern, you may want to set another environment variable called NODISCLAIM=true so that each free()
subroutine call does not also result in a disclaim() system call.
Deferred Page Space Allocation
A new page-space allocation policy (a modification of Late Page Space Allocation) was introduced in AIX
4.3.2: Deferred Page Space Allocation. Prior to AIX 4.3.2, a page-space disk block was allocated when a
page was touched. However, this paging space may never be used, especially on systems with large real
memory where paging is rare. With Deferred Page Space Allocation, the disk block allocation of paging
space is delayed until it is necessary to page out the page, which results in no wasted paging space
allocation. This does, however, result in additional overcommitment of paging space. On a system where
enough virtual memory is accessed that paging is necessary, the amount of paging space required may be
as much as was required on previously.
After a page has been paged out to paging space, the disk block is reserved for that page if that page is
paged back into RAM. Therefore, the paging space percentage-used value may not necessarily reflect the
Chapter 8. Monitoring and Tuning Memory Use
161
number of pages only in paging space because some of it may be back in RAM as well. If the page that
was paged back in is working storage of a thread, and if the thread releases the memory associated with
that page or if the thread exits, then the disk block for that page is released.
Choosing between LPSA and DPSA with the vmtune Command
Running the vmtune command with the -d option enables turning on or off the Deferred Page Space
Allocation in order to preserve the Late Page Space Allocation policy. A value of 1 indicates that DPSA
should be on, and a value of 0 indicates that DPSA should be off. If you choose to turn DPSA off, make
sure that the kernel level is at least AIX 4.3.2.6 or higher.
Looking at Paging Space and Virtual Memory
The vmstat command (avm column), ps command (SIZE, SZ), and other utilities report the amount of virtual
memory actually accessed because with DPSA, the paging space may not get touched. The svmon
command (up through AIX 4.3.2) shows the amount of paging space being used, so this value may be
much smaller than the avm value of the vmstat command.
It is safer to use the lsps -s command rather than the lsps -a command to look at available paging space
because the command lsps -a only shows paging space that is actually being used. But the command
lsps -s will include paging space being used along with paging space that was reserved using the EPSA
policy.
Using Shared Memory
By using the shmat() or mmap() subroutines, files can be explicitly mapped into memory. This avoids
buffering and avoids system-call overhead. The memory areas are known as the shared memory
segments or regions. Beginning with AIX 4.2.1 and only affecting 32-bit applications, segment 14 was
released providing 11 shared memory segments (not including the shared library data or shared library text
segments) for processes (segments 3-12 and 14). Each of these segments are 256 MB in size.
Applications can read/write the file by reading/writing in the segment. Applications can avoid overhead of
read/write system calls simply by manipulating pointers in these mapped segments.
Files or data can also be shared among multiple processes/threads. However, this requires
synchronization between these processes/threads and its handling is up to the application. Typical use is
by database applications for use as a large database buffer cache.
Paging space is allocated for shared memory regions just as it would for the process private segment
(paging space is used as soon as the pages are touched if deferred page space allocation policy is off).
Extended Shared Memory (EXTSHM)
By default, each shared memory region (whatever its size), always consumes a 256 MB region of address
space. AIX 4.2.1 and later implements Extended Shared Memory, which allows for more granular shared
memory regions that can be in size of 1 byte up to 256 MB. However, the address space consumption will
be rounded up to the next page (4096 byte) boundary. Extended Shared Memory essentially removes the
limitation of only 11 shared memory regions, but note that when using EXTSHM, the mmap services are
actually used and thus will have the same performance implications of mmap.
This feature is available to processes that have the variable EXTSHM set to ON (EXTSHM=ON) in their
process environment. There is no limit on the number of shared memory regions that a process can
attach. File mapping is supported as before, but still consumes address space that is a multiple of 256 MB
(segment size). Resizing a shared memory region is not supported in this mode. Kernel processes will still
have the same behavior. Without this environment variable set, eleven 256 MB regions are available.
Extended Shared Memory has the following restrictions:
v I/O support is restricted in the same manner as for memory-mapped regions.
162
Performance Management Guide
v Only uphysio() type of I/O is supported (no raw I/O).
v These shared memory regions cannot be used as I/O buffers where the unpinning of the buffer occurs
in an interrupt handler. For example, these regions cannot be used for async I/O buffers.
v The segments cannot be pinned using the plock() subroutine because memory-mapped segments
cannot be pinned with the plock() subroutine.
Using AIX Memory Affinity Support
Inroduction
IBM POWER-based SMP hardware systems contain several multichip modules (MCMs), each containing
multiple processors. System memory is attached to these MCMs. While any processor can access all of
the memory in the system, a processor has faster access, and higher bandwidth, when addressing
memory that is attached to its own MCM rather than memory attached to the other MCMs in the system.
AIX has optional support to recognize the division of system memory among the MCMs. If the memory
affinity support is enabled, AIX attempts to satisfy a page fault using memory attached to the MCM
containing the processor that caused the page fault. This may provide a performance benefit to the
application.
Enabling memory affinity.
Using the memory affinity support on AIX is a two step process. The following vmtune command will
enable the support:
vmtune -y 0|1
(0 disabled, 1 enabled)
Note: A bosboot and a reboot are required in order for it to take effect.
This action will only tell AIX to organize its data structures along MCM boundaries. The default memory
allocation policy rotates among the MCMs. In order to obtain preferential local MCM memory allocation, an
application must export the MEMORY_AFFINITY environment variable as follows:
MEMORY_AFFINITY=MCM
This behavior is propogated across a fork. However, for this behavior to be retained across an exec, the
variable must be contained in the environment string passed to the exec function call.
Performance Effect Of Local MCM Memory Allocation
The effect local MCM memory allocation will have on a specific application is difficult to predict. Some
applications are unaffected, some might improve, and others might degrade.
Most applications must be bound to processors to get a performance benefit from memory affinity. This is
needed to prevent the AIX dispatcher from moving the application to processors in different MCMs while
the application executes.
The most likely way to obtain a benefit from memory affinity is to limit the application to running only on
the processors contained in a single MCM. This can be done with the bindprocessor command and the
bindprocessor() function. It can also be done with the resource set affinity commands and services.
When the application requires more processors than contained in a single MCM, the performance benefit
through memory affinity depends on the memory allocation and access patterns of the various threads in
the application. Applications with threads that individually allocate and reference unique data areas may
see improved performance. Applications that share memory among all the threads are more likely to get a
degradation from memory affinity.
Chapter 8. Monitoring and Tuning Memory Use
163
Related Information
The vmtune command and Tuning VMM Page Replacement with the vmtune Command.
The bindprocessor command or sub-routine.
WLM Class Attributes and Resource Set Attributes.
164
Performance Management Guide
Chapter 9. File System, Logical Volume, and Disk I/O
Performance
This chapter focuses on file system tuning and the performance of locally attached disk drives.
If you are not familiar with the operating system concepts of volume groups, logical and physical volumes,
and logical and physical partitions, read Performance Overview of Fixed-Disk Storage Management.
File-system configuration has a large effect on overall system performance and is time-consuming to
change after installation. Deciding on the number and types of hard disks, and the sizes and placements
of paging spaces and logical volumes on those hard disks, is therefore a critical preinstallation process.
For an extensive discussion of the considerations for preinstallation disk configuration planning, see Disk
Preinstallation Guidelines.
The following sections are presented in this chapter:
v Monitoring Disk I/O
v Monitoring and Tuning File Systems
v File System Types
v Differences Between JFS and Enhanced JFS
v Potential Performance Inhibitors for JFS and Enhanced JFS
v Performance Enhancements
v
v
v
v
v
Summary of Tunable Parameters
Changing File System Attributes that Affect Performance
Changing Logical Volume Attributes That Affect Performance
Physical Volume Considerations
Volume Group Recommendations
v Reorganizing Logical Volumes
v Reorganizing File Systems
v Reorganizing JFS Log and Log Logical Volumes
v Tuning with vmtune
v Using Disk-I/O Pacing
v Tuning Logical Volume Striping
v Tuning Asynchronous Disk I/O
v Tuning Direct I/O
v Using Raw Disk I/O
v
v
v
v
v
v
Using sync/fsync Calls
Setting SCSI-Adapter and Disk-Device Queue Limits
Expanding the Configuration
Using RAID
Using SSA
Using Fast Write Cache
Monitoring Disk I/O
When you are monitoring disk I/O, use the following to determine your course of action:
v Find the most active files, file systems, and logical volumes:
© Copyright IBM Corp. 1997, 2002
165
– Can ″hot″ file systems be better located on the physical drive or be spread across multiple physical
drives? (lslv, iostat, filemon)
– Are ″hot″ files local or remote? (filemon)
– Does paging space dominate disk utilization? (vmstat, filemon)
– Is there enough memory to cache the file pages being used by running processes? (vmstat, svmon,
vmtune)
– Does the application perform a lot of synchronous (non-cached) file I/O?
v Determine file fragmentation:
– Are ″hot″ files heavily fragmented? (fileplace)
v Find the physical volume with the highest utilization:
– Is the type of drive or I/O adapter causing a bottleneck? (iostat, filemon)
Building a Pre-Tuning Baseline
Before you make significant changes in your disk configuration or tuning parameters, it is a good idea to
build a baseline of measurements that record the current configuration and performance.
Wait I/O Time Reporting
AIX 4.3.3 and later contain enhancements to the method used to compute the percentage of CPU time
spent waiting on disk I/O (wio time). The method used in AIX 4.3.2 and earlier versions of the operating
system can, under certain circumstances, give an inflated view of wio time on SMPs. The wio time is
reported by the commands sar (%wio), vmstat (wa) and iostat (% iowait).
Another change is that the wa column details the percentage of time the CPU was idle with pending disk
I/O to not only local, but also NFS-mounted disks.
Method Used in AIX 4.3.2 and Earlier
At each clock interrupt on each processor (100 times a second per processor), a determination is made as
to which of the four categories (usr/sys/wio/idle) to place the last 10 ms of time. If the CPU was busy in
usr mode at the time of the clock interrupt, then usr gets the clock tick added into its category. If the CPU
was busy in kernel mode at the time of the clock interrupt, then the sys category gets the tick. If the CPU
was not busy, a check is made to see if any I/O to disk is in progress. If any disk I/O is in progress, the
wio category is incremented. If no disk I/O is in progress and the CPU is not busy, the idle category gets
the tick.
The inflated view of wio time results from all idle CPUs being categorized as wio regardless of the number
of threads waiting on I/O. For example, systems with just one thread doing I/O could report over 90
percent wio time regardless of the number of CPUs it has.
Method Used in AIX 4.3.3 and Later
The change in AIX 4.3.3 is to only mark an idle CPU as wio if an outstanding I/O was started on that CPU.
This method can report much lower wio times when just a few threads are doing I/O and the system is
otherwise idle. For example, a system with four CPUs and one thread doing I/O will report a maximum of
25 percent wio time. A system with 12 CPUs and one thread doing I/O will report a maximum of 8.3
percent wio time.
Also, starting with AIX 4.3.3, waiting on I/O to NFS mounted file systems is reported as wait I/O time.
Assessing Disk Performance with the iostat Command
Begin the assessment by running the iostat command with an interval parameter during your system’s
peak workload period or while running a critical application for which you need to minimize I/O delays. The
following shell script runs the iostat command in the background while a copy of a large file runs in the
foreground so that there is some I/O to measure:
166
Performance Management Guide
# iostat 5 3 >io.out &
# cp big1 /dev/null
This example leaves the following three reports in the io.out file:tty: tin tout avg-cpu: % user % sys % idle
% iowait 0.0 1.3 0.2 0.6 98.9 0.3 Disks: % tm_act Kbps tps Kb_read Kb_wrtn hdisk0 0.0 0.3 0.0 29753
48076 hdisk1 0.1 0.1 0.0 11971 26460 hdisk2 0.2 0.8 0.1 91200 108355 cd0 0.0 0.0 0.0 0 0 tty: tin tout
avg-cpu: % user % sys % idle % iowait 0.8 0.8 0.6 9.7 50.2 39.5 Disks: % tm_act Kbps tps Kb_read
Kb_wrtn hdisk0 47.0 674.6 21.8 3376 24 hdisk1 1.2 2.4 0.6 0 12 hdisk2 4.0 7.9 1.8 8 32 cd0 0.0 0.0 0.0 0
0 tty: tin tout avg-cpu: % user % sys % idle % iowait 2.0 2.0 0.2 1.8 93.4 4.6 Disks: % tm_act Kbps tps
Kb_read Kb_wrtn hdisk0 0.0 0.0 0.0 0 0 hdisk1 0.0 0.0 0.0 0 0 hdisk2 4.8 12.8 3.2 64 0 cd0 0.0 0.0 0.0 0
0
The first report is the summary since the last reboot and shows the overall balance (or, in this case,
imbalance) in the I/O to each of the hard disks. hdisk1 was almost idle and hdisk2 received about 63
percent of the total I/O (from Kb_read and Kb_wrtn).
Note: The system maintains a history of disk activity. If the history is disabled (smitty chgsys ->
Continuously maintain DISK I/O history [false]), the following message displays when you run
the iostat command:
Disk history since boot not available.
The interval disk I/O statistics are unaffected by this.
The second report shows the 5-second interval during which cp ran. Examine this information carefully.
The elapsed time for this cp was about 2.6 seconds. Thus, 2.5 seconds of high I/O dependency are being
averaged with 2.5 seconds of idle time to yield the 39.5 percent % iowait reported. A shorter interval
would have given a more detailed characterization of the command itself, but this example demonstrates
what you must consider when you are looking at reports that show average activity across intervals.
TTY Report
The two columns of TTY information (tin and tout) in the iostat output show the number of characters
read and written by all TTY devices. This includes both real and pseudo TTY devices. Real TTY devices
are those connected to an asynchronous port. Some pseudo TTY devices are shells, telnet sessions, and
aixterm windows.
Because the processing of input and output characters consumes CPU resources, look for a correlation
between increased TTY activity and CPU utilization. If such a relationship exists, evaluate ways to improve
the performance of the TTY subsystem. Steps that could be taken include changing the application
program, modifying TTY port parameters during file transfer, or perhaps upgrading to a faster or more
efficient asynchronous communications adapter.
In Shell Script fastport.sh for Fast File Transfers, you can find the fastport.sh script, which is intended to
condition a TTY port for fast file transfers in raw mode; for example, when a FAX machine is to be
connected. Using the script might improve CPU performance by a factor of 3 at 38400 baud.
CPU Report
The CPU statistics columns (% user, % sys, % idle, and % iowait) provide a breakdown of CPU usage.
This information is also reported in the vmstat command output in the columns labeled us, sy, id, and wa.
For a detailed explanation for the values, see The vmstat Command. Also note the change made to %
iowait described in Wait I/O Time Reporting.
On systems running one application, high I/O wait percentage might be related to the workload. On
systems with many processes, some will be running while others wait for I/O. In this case, the % iowait
can be small or zero because running processes ″hide″ some wait time. Although % iowait is low, a
bottleneck can still limit application performance.
Chapter 9. File System, Logical Volume, and Disk I/O Performance
167
If the iostat command indicates that a CPU-bound situation does not exist, and % iowait time is greater
than 20 percent, you might have an I/O or disk-bound situation. This situation could be caused by
excessive paging due to a lack of real memory. It could also be due to unbalanced disk load, fragmented
data or usage patterns. For an unbalanced disk load, the same iostat report provides the necessary
information. But for information about file systems or logical volumes, which are logical resources, you
must use tools such as the filemon or fileplace commands.
Drive Report
When you suspect a disk I/O performance problem, use the iostat command. To avoid the information
about the TTY and CPU statistics, use the -d option. In addition, the disk statistics can be limited to the
important disks by specifying the disk names.
Remember that the first set of data represents all activity since system startup.
Disks: Shows the names of the physical volumes. They are either hdisk or cd followed by a number. If
physical volume names are specified with the iostat command, only those names specified are
displayed.
% tm_act
Indicates the percentage of time that the physical disk was active (bandwidth utilization for the
drive) or, in other words, the total time disk requests are outstanding. A drive is active during data
transfer and command processing, such as seeking to a new location. The ″disk active time″
percentage is directly proportional to resource contention and inversely proportional to
performance. As disk use increases, performance decreases and response time increases. In
general, when the utilization exceeds 70 percent, processes are waiting longer than necessary for
I/O to complete because most UNIX processes block (or sleep) while waiting for their I/O requests
to complete. Look for busy versus idle drives. Moving data from busy to idle drives can help
alleviate a disk bottleneck. Paging to and from disk will contribute to the I/O load.
Kbps
Indicates the amount of data transferred (read or written) to the drive in KB per second. This is the
sum of Kb_read plus Kb_wrtn, divided by the seconds in the reporting interval.
tps
Indicates the number of transfers per second that were issued to the physical disk. A transfer is an
I/O request through the device driver level to the physical disk. Multiple logical requests can be
combined into a single I/O request to the disk. A transfer is of indeterminate size.
Kb_read
Reports the total data (in KB) read from the physical volume during the measured interval.
Kb_wrtn
Shows the amount of data (in KB) written to the physical volume during the measured interval.
Taken alone, there is no unacceptable value for any of the above fields because statistics are too closely
related to application characteristics, system configuration, and type of physical disk drives and adapters.
Therefore, when you are evaluating data, look for patterns and relationships. The most common
relationship is between disk utilization (%tm_act) and data transfer rate (tps).
To draw any valid conclusions from this data, you have to understand the application’s disk data access
patterns such as sequential, random, or combination, as well as the type of physical disk drives and
adapters on the system. For example, if an application reads/writes sequentially, you should expect a high
disk transfer rate (Kbps) when you have a high disk busy rate (%tm_act). Columns Kb_read and Kb_wrtn
can confirm an understanding of an application’s read/write behavior. However, these columns provide no
information on the data access patterns.
Generally you do not need to be concerned about a high disk busy rate (%tm_act) as long as the disk
transfer rate (Kbps) is also high. However, if you get a high disk busy rate and a low disk transfer rate, you
may have a fragmented logical volume, file system, or individual file.
168
Performance Management Guide
Discussions of disk, logical volume and file system performance sometimes lead to the conclusion that the
more drives you have on your system, the better the disk I/O performance. This is not always true
because there is a limit to the amount of data that can be handled by a disk adapter. The disk adapter can
also become a bottleneck. If all your disk drives are on one disk adapter, and your hot file systems are on
separate physical volumes, you might benefit from using multiple disk adapters. Performance improvement
will depend on the type of access.
To see if a particular adapter is saturated, use the iostat command and add up all the Kbps amounts for
the disks attached to a particular disk adapter. For maximum aggregate performance, the total of the
transfer rates (Kbps) must be below the disk adapter throughput rating. In most cases, use 70 percent of
the throughput rate. In operating system versions later than 4.3.3 the -a or -A option will display this
information.
Assessing Disk Performance with the vmstat Command
To prove that the system is I/O bound, it is better to use the iostat command. However, the vmstat
command could point to that direction by looking at the wa column, as discussed in The vmstat Command.
Other indicators for I/O bound are:
v The disk xfer part of the vmstat output
To display a statistic about the logical disks (a maximum of four disks is allowed), use the following
command:
# vmstat hdisk0 hdisk1 1 8
kthr
memory
page
faults
---- ---------- ----------------------- -----------r b
avm
fre re pi po fr sr cy in
sy cs
0 0 3456 27743
0
0
0
0
0
0 131 149 28
0 0 3456 27743
0
0
0
0
0
0 131
77 30
1 0 3498 27152
0
0
0
0
0
0 153 1088 35
0 1 3499 26543
0
0
0
0
0
0 199 1530 38
0 1 3499 25406
0
0
0
0
0
0 187 2472 38
0 0 3456 24329
0
0
0
0
0
0 178 1301 37
0 0 3456 24329
0
0
0
0
0
0 124
58 19
0 0 3456 24329
0
0
0
0
0
0 123
58 23
cpu
----------us sy id wa
0 1 99 0
0 1 99 0
1 10 87 2
1 19 0 80
2 26 0 72
2 12 20 66
0 0 99 0
0 0 99 0
disk xfer
-----1 2 3 4
0 0
0 0
0 11
0 59
0 53
0 42
0 0
0 0
The disk xfer part provides the number of transfers per second to the specified physical volumes that
occurred in the sample interval. One to four physical volume names can be specified. Transfer statistics
are given for each specified drive in the order specified. This count represents requests to the physical
device. It does not imply an amount of data that was read or written. Several logical requests can be
combined into one physical request.
v The in column of the vmstat output
This column shows the number of hardware or device interrupts (per second) observed over the
measurement interval. Examples of interrupts are disk request completions and the 10 millisecond clock
interrupt. Since the latter occurs 100 times per second, the in field is always greater than 100. But the
vmstat command also provides a more detailed output about the system interrupts.
v The vmstat -i output
The -i parameter displays the number of interrupts taken by each device since system startup. But, by
adding the interval and, optionally, the count parameter, the statistic since startup is only displayed in
the first stanza; every trailing stanza is a statistic about the scanned interval.
# vmstat -i 1 2
priority level
0
0
0
1
0
2
0
2
1
2
3
10
3
14
5
62
type
hardware
hardware
hardware
hardware
hardware
hardware
hardware
hardware
count module(handler)
0 i_misc_pwr(a868c)
0 i_scu(a8680)
0 i_epow(954e0)
0 /etc/drivers/ascsiddpin(189acd4)
194 /etc/drivers/rsdd(1941354)
10589024 /etc/drivers/mpsdd(1977a88)
101947 /etc/drivers/ascsiddpin(189ab8c)
61336129 clock(952c4)
Chapter 9. File System, Logical Volume, and Disk I/O Performance
169
10
63
hardware 13769 i_softoff(9527c)
priority level
type count module(handler)
0
0
hardware
0 i_misc_pwr(a868c)
0
1
hardware
0 i_scu(a8680)
0
2
hardware
0 i_epow(954e0)
0
2
hardware
0 /etc/drivers/ascsiddpin(189acd4)
1
2
hardware
0 /etc/drivers/rsdd(1941354)
3
10
hardware
25 /etc/drivers/mpsdd(1977a88)
3
14
hardware
0 /etc/drivers/ascsiddpin(189ab8c)
5
62
hardware 105 clock(952c4)
10
63
hardware
0 i_softoff(9527c)
Note: The output will differ from system to system, depending on hardware and software configurations
(for example, the clock interrupts may not be displayed in the vmstat -i output although they will
be accounted for under the in column in the normal vmstat output). Check for high numbers in
the count column and investigate why this module has to execute so many interrupts.
Assessing Disk Performance with the sar Command
The sar command is a standard UNIX command used to gather statistical data about the system. With its
numerous options, the sar command provides queuing, paging, TTY, and many other statistics. With AIX
4.3.3, the sar -d option generates real-time disk I/O statistics.
# sar -d 3 3
AIX konark 3 4 0002506F4C00
12:09:50
device
%busy
12:09:53
hdisk0
hdisk1
cd0
1
0
0
12:09:56
hdisk0
hdisk1
cd0
12:09:59
Average
08/26/99
avque
r+w/s
blks/s
avwait
avserv
0.0
0.0
0.0
0
0
0
5
1
0
0.0
0.0
0.0
0.0
0.0
0.0
0
0
0
0.0
0.0
0.0
0
0
0
0
1
0
0.0
0.0
0.0
0.0
0.0
0.0
hdisk0
hdisk1
cd0
1
0
0
0.0
0.0
0.0
1
0
0
4
1
0
0.0
0.0
0.0
0.0
0.0
0.0
hdisk0
hdisk1
cd0
0
0
0
0.0
0.0
0.0
0
0
0
3
1
0
0.0
0.0
0.0
0.0
0.0
0.0
The fields listed by the sar -d command are as follows:
%busy
Portion of time device was busy servicing a transfer request. This is the same as the %tm_act
column in the iostat command report.
avque Average number of requests outstanding from the adapter to the device during that time. There
may be additonal I/O operations in the queue of the device driver. This number is a good indicator
if an I/O bottleneck exists.
Number of read/write transfers from or to device. This is the same as tps in the iostat command
report.
r+w/s
blks/s Number of bytes transferred in 512-byte units
avwait
Average number of transactions waiting for service (queue length). Average time (in milliseconds)
that transfer requests waited idly on queue for the device. This number is currently not reported
and shows 0.0 by default.
170
Performance Management Guide
avserv
Number of milliseconds per average seek. Average time (in milliseconds) to service each transfer
request (includes seek, rotational latency, and data transfer times) for the device. This number is
currently not reported and shows 0.0 by default.
Assessing Logical Volume Fragmentation with the lslv Command
The lslv command shows, among other information, the logical volume fragmentation. To check logical
volume fragmentation, use the command lslv -l lvname, as follows:
# lslv -l hd2
hd2:/usr
PV
hdisk0
COPIES
114:000:000
IN BAND
22%
DISTRIBUTION
000:042:026:000:046
The output of COPIES shows the logical volume hd2 has only one copy. The IN BAND shows how well the
intrapolicy, an attribute of logical volumes, is followed. The higher the percentage, the better the allocation
efficiency. Each logical volume has its own intrapolicy. If the operating system cannot meet this
requirement, it chooses the best way to meet the requirements. In our example, there are a total of 114
logical partitions (LP); 42 LPs are located on middle, 26 LPs on center, and 46 LPs on inner-edge. Since
the logical volume intrapolicy is center, the in-band is 22 percent (26 / (42+26+46). The DISTRIBUTION
shows how the physical partitions are placed in each part of the intrapolicy; that is:
edge : middle : center : inner-middle : inner-edge
See Position on Physical Volume for additional information about physical partitions placement.
Assessing Physical Placement of Data with the lslv Command
If the workload shows a significant degree of I/O dependency, you can investigate the physical placement
of the files on the disk to determine if reorganization at some level would yield an improvement. To see the
placement of the partitions of logical volume hd11 within physical volume hdisk0, use the following:
# lslv -p hdisk0 hd11
hdisk0:hd11:/home/op
USED USED USED USED
USED USED USED USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
1-10
11-17
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
18-27
28-34
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
35-44
45-50
USED
0052
USED
0053
USED
0054
USED
0055
USED
0056
USED
0057
USED
0058
USED
USED
USED
51-60
61-67
0059
0069
0060
0070
0061
0071
0062
0072
0063
0073
0064
0074
0065
0075
0066
0067
0068
68-77
78-84
Look for the rest of hd11 on hdisk1 with the following:
# lslv -p hdisk1 hd11
hdisk1:hd11:/home/op
0035 0036 0037 0038
0045 0046 0047 0048
0039
0049
0040
0050
0041
0051
0042
0043
0044
1-10
11-17
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
18-27
28-34
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
USED
35-44
45-50
0001
0011
0002
0012
0003
0013
0004
0014
0005
0015
0006
0016
0007
0017
0008
0009
0010
51-60
61-67
Chapter 9. File System, Logical Volume, and Disk I/O Performance
171
0018
0028
0019
0029
0020
0030
0021
0031
0022
0032
0023
0033
0024
0034
0025
0026
0027
68-77
78-84
From top to bottom, five blocks represent edge, middle, center, inner-middle, and inner-edge, respectively.
v A USED indicates that the physical partition at this location is used by a logical volume other than the one
specified. A number indicates the logical partition number of the logical volume specified with the lslv -p
command.
v A FREE indicates that this physical partition is not used by any logical volume. Logical volume
fragmentation occurs if logical partitions are not contiguous across the disk.
v A STALE physical partition is a physical partition that contains data you cannot use. You can also see the
STALE physical partitions with the lspv -m command. Physical partitions marked as STALE must be
updated to contain the same information as valid physical partitions. This process, called
resynchronization with the syncvg command, can be done at vary-on time, or can be started anytime
the system is running. Until the STALE partitions have been rewritten with valid data, they are not used to
satisfy read requests, nor are they written to on write requests.
In the previous example, logical volume hd11 is fragmented within physical volume hdisk1, with its first
logical partitions in the inner-middle and inner regions of hdisk1, while logical partitions 35-51 are in the
outer region. A workload that accessed hd11 randomly would experience unnecessary I/O wait time as
longer seeks might be needed on logical volume hd11. These reports also indicate that there are no free
physical partitions in either hdisk0 or hdisk1.
Assessing File Placement with the fileplace Command
To see how the file copied earlier, big1, is stored on the disk, we can use the fileplace command. The
fileplace command displays the placement of a file’s blocks within a logical volume or within one or more
physical volumes.
To determine whether the fileplace command is installed and available, run the following command:
# lslpp -lI perfagent.tools
Use the following command:
# fileplace -pv big1
File: big1 Size: 3554273 bytes Vol: /dev/hd10
Blk Size: 4096 Frag Size: 4096 Nfrags: 868
Compress: no
Inode: 19 Mode: -rwxr-xr-x Owner: hoetzel Group: system
Physical Addresses (mirror copy 1)
---------------------------------0001584-0001591 hdisk0
8 frags
0001624-0001671 hdisk0
48 frags
0001728-0002539 hdisk0
812 frags
32768 Bytes,
196608 Bytes,
3325952 Bytes,
0.9%
5.5%
93.5%
Logical Fragment
---------------0001040-0001047
0001080-0001127
0001184-0001995
868 frags over space of 956 frags: space efficiency = 90.8%
3 fragments out of 868 possible: sequentiality = 99.8%
This example shows that there is very little fragmentation within the file, and those are small gaps. We can
therefore infer that the disk arrangement of big1 is not significantly affecting its sequential read-time.
Further, given that a (recently created) 3.5 MB file encounters this little fragmentation, it appears that the
file system in general has not become particularly fragmented.
Occasionally, portions of a file may not be mapped to any blocks in the volume. These areas are implicitly
filled with zeroes by the file system. These areas show as unallocated logical blocks. A file that has these
holes will show the file size to be a larger number of bytes than it actually occupies (that is, the ls -l
command will show a large size, whereas the du command will show a smaller size or the number of
blocks the file really occupies on disk).
172
Performance Management Guide
The fileplace command reads the file’s list of blocks from the logical volume. If the file is new, the
information may not be on disk yet. Use the sync command to flush the information. Also, the fileplace
command will not display NFS remote files (unless the command runs on the server).
Note: If a file has been created by seeking to various locations and writing widely dispersed records, only
the pages that contain records will take up space on disk and appear on a fileplace report. The file
system does not fill in the intervening pages automatically when the file is created. However, if such
a file is read sequentially (by the cp or tar commands, for example) the space between records is
read as binary zeroes. Thus, the output of such a cp command can be much larger than the input
file, although the data is the same.
Space Efficiency and Sequentiality
Higher space efficiency means files are less fragmented and probably provide better sequential file
access. A higher sequentiality indicates that the files are more contiguously allocated, and this will
probably be better for sequential file access.
Space efficiency =
Total number of fragments used for file storage /
(Largest fragment physical address Smallest fragment physical address + 1)
Sequentiality =
(Total number of fragments Number of grouped fragments +1) /
Total number of fragments
If you find that your sequentiality or space efficiency values become low, you can use the reorgvg
command to improve logical volume utilization and efficiency (see Reorganizing Logical Volumes). To
improve file system utilization and efficiency, see Reorganizing File Systems.
In this example, the Largest fragment physical address - Smallest fragment physical address + 1 is:
0002539 - 0001584 + 1 = 956 fragments; total used fragments is: 8 + 48 + 812 = 868; the space efficiency
is 868 / 956 (90.8 percent); the sequentiality is (868 - 3 + 1) / 868 = 99.8 percent.
Because the total number of fragments used for file storage does not include the indirect blocks location,
but the physical address does, the space efficiency can never be 100 percent for files larger than 32 KB,
even if the file is located on contiguous fragments.
Assessing Paging Space I/O with the vmstat Command
I/O to and from paging spaces is random, mostly one page at a time. The vmstat reports indicate the
amount of paging-space I/O taking place. Both of the following examples show the paging activity that
occurs during a C compilation in a machine that has been artificially shrunk using the rmss command. The
pi and po (paging-space page-ins and paging-space page-outs) columns show the amount of
paging-space I/O (in terms of 4096-byte pages) during each 5-second interval. The first report (summary
since system reboot) has been removed. Notice that the paging activity occurs in bursts.
# vmstat 5 8
kthr
memory
page
faults
cpu
----- ----------- ------------------------ ------------ ----------r b
avm
fre re pi po fr
sr cy in
sy cs us sy id wa
0 1 72379
434
0 0
0
0
2
0 376 192 478 9 3 87 1
0 1 72379
391
0 8
0
0
0
0 631 2967 775 10 1 83 6
0 1 72379
391
0 0
0
0
0
0 625 2672 790 5 3 92 0
0 1 72379
175
0 7
0
0
0
0 721 3215 868 8 4 72 16
2 1 71384
877
0 12 13 44 150
0 662 3049 853 7 12 40 41
0 2 71929
127
0 35 30 182 666
0 709 2838 977 15 13 0 71
0 1 71938
122
0 0
8 32 122
0 608 3332 787 10 4 75 11
0 1 71938
122
0 0
0
3
12
0 611 2834 733 5 3 75 17
Chapter 9. File System, Logical Volume, and Disk I/O Performance
173
The following ″before and after″ vmstat -s reports show the accumulation of paging activity. Remember
that it is the paging space page ins and paging space page outs that represent true paging-space I/O. The
(unqualified) page ins and page outs report total I/O, that is both paging-space I/O and the ordinary file
I/O, performed by the paging mechanism. The reports have been edited to remove lines that are irrelevant
to this discussion.
# vmstat -s # before
# vmstat -s # after
6602 page ins
3948 page outs
544 paging space page ins
1923 paging space page outs
0 total reclaims
7022 page ins
4146 page outs
689 paging space page ins
2032 paging space page outs
0 total reclaims
The fact that more paging-space page-ins than page-outs occurred during the compilation suggests that
we had shrunk the system to the point that thrashing begins. Some pages were being repaged because
their frames were stolen before their use was complete.
Assessing Overall Disk I/O with the vmstat Command
The technique just discussed can also be used to assess the disk I/O load generated by a program. If the
system is otherwise idle, the following sequence:
#
#
#
#
#
vmstat -s >statout
testpgm
sync
vmstat -s >> statout
egrep "ins|outs" statout
yields a before and after picture of the cumulative disk activity counts, such as:
5698
5012
0
32
6671
5268
8
225
page ins
page outs
paging space
paging space
page ins
page outs
paging space
paging space
page ins
page outs
page ins
page outs
During the period when this command (a large C compile) was running, the system read a total of 981
pages (8 from paging space) and wrote a total of 449 pages (193 to paging space).
Detailed I/O Analysis with the filemon Command
The filemon command uses the trace facility to obtain a detailed picture of I/O activity during a time
interval on the various layers of file system utilization, including the logical file system, virtual memory
segments, LVM, and physical disk layers. Data can be collected on all the layers, or layers can be
specified with the -O layer option. The default is to collect data on the VM, LVM, and physical layers. Both
summary and detailed reports are generated. Since it uses the trace facility, the filemon command can be
run only by the root user or by a member of the system group.
To determine whether the filemon command is installed and available, run the following command:
# lslpp -lI perfagent.tools
Tracing is started by the filemon command, optionally suspended with the trcoff subcommand and
resumed with the trcon subcomand, and terminated with the trcstop subcommand (you may want to
issue the command nice -n -20 trcstop to stop the filemon command since the filemon command is
currently running at priority 40). As soon as tracing is terminated, the filemon command writes its report to
stdout.
174
Performance Management Guide
Note: Only data for those files opened after the filemon command was started will be collected, unless
you specify the -u flag.
The filemon command can read the I/O trace data from a specified file, instead of from the real-time trace
process. In this case, the filemon report summarizes the I/O activity for the system and period
represented by the trace file. This offline processing method is useful when it is necessary to postprocess
a trace file from a remote machine or perform the trace data collection at one time and postprocess it at
another time.
The trcrpt -r command must be executed on the trace logfile and redirected to another file, as follows:
# gennames > gennames.out
# trcrpt -r
trace.out > trace.rpt
At this point an adjusted trace logfile is fed into the filemon command to report on I/O activity captured by
a previously recorded trace session as follows:
# filemon -i trace.rpt -n gennames.out | pg
In this example, the filemon command reads file system trace events from the input file trace.rpt.
Because the trace data is already captured on a file, the filemon command does not put itself in the
background to allow application programs to be run. After the entire file is read, an I/O activity report for
the virtual memory, logical volume, and physical volume levels is displayed on standard output (which, in
this example, is piped to the pg command).
If the trace command was run with the -C all flag, then run the trcrpt command also with the -C all flag
(see Formatting a Report from trace -C Output).
The following sequence of commands gives an example of the filemon command usage:
# filemon -o fm.out -O all; cp /smit.log /dev/null ; trcstop
The report produced by this sequence, in an otherwise-idle system, is as follows:
Thu Aug 19 11:30:49 1999
System: AIX texmex Node: 4 Machine: 000691854C00
0.369 secs in measured interval
Cpu utilization: 9.0%
Most Active Files
-----------------------------------------------------------------------#MBs #opns
#rds
#wrs file
volume:inode
-----------------------------------------------------------------------0.1
1
14
0 smit.log
/dev/hd4:858
0.0
1
0
13 null
0.0
2
4
0 ksh.cat
/dev/hd2:16872
0.0
1
2
0 cmdtrace.cat
/dev/hd2:16739
Most Active Segments
-----------------------------------------------------------------------#MBs #rpgs #wpgs segid segtype
volume:inode
-----------------------------------------------------------------------0.1
13
0
5e93 ???
0.0
2
0
22ed ???
0.0
1
0
5c77 persistent
Most Active Logical Volumes
-----------------------------------------------------------------------util #rblk #wblk
KB/s volume
description
-----------------------------------------------------------------------0.06
112
0 151.9 /dev/hd4
/
0.04
16
0
21.7 /dev/hd2
/usr
Chapter 9. File System, Logical Volume, and Disk I/O Performance
175
Most Active Physical Volumes
-----------------------------------------------------------------------util #rblk #wblk
KB/s volume
description
-----------------------------------------------------------------------0.10
128
0 173.6 /dev/hdisk0
N/A
-----------------------------------------------------------------------Detailed File Stats
-----------------------------------------------------------------------FILE: /smit.log volume: /dev/hd4 (/) inode: 858
opens:
1
total bytes xfrd:
57344
reads:
14
(0 errs)
read sizes (bytes):
avg 4096.0 min
4096 max
read times (msec):
avg
1.709 min
0.002 max
FILE: /dev/null
opens:
1
total bytes xfrd:
50600
writes:
13
(0 errs)
write sizes (bytes): avg 3892.3 min
write times (msec):
avg
0.007 min
1448 max
0.003 max
4096 sdev
19.996 sdev
0.0
5.092
4096 sdev
0.022 sdev
705.6
0.006
FILE: /usr/lib/nls/msg/en_US/ksh.cat volume: /dev/hd2 (/usr) inode: 16872
opens:
2
total bytes xfrd:
16384
reads:
4
(0 errs)
read sizes (bytes):
avg 4096.0 min
4096 max
4096 sdev
0.0
read times (msec):
avg
0.042 min
0.015 max
0.070 sdev
0.025
lseeks:
10
FILE: /usr/lib/nls/msg/en_US/cmdtrace.cat volume: /dev/hd2 (/usr) inode: 16739
opens:
1
total bytes xfrd:
8192
reads:
2
(0 errs)
read sizes (bytes):
avg 4096.0 min
4096 max
4096 sdev
0.0
read times (msec):
avg
0.062 min
0.049 max
0.075 sdev
0.013
lseeks:
8
-----------------------------------------------------------------------Detailed VM Segment Stats
(4096 byte pages)
-----------------------------------------------------------------------SEGMENT: 5e93 segtype:
segment flags:
reads:
read times (msec):
read sequences:
read seq. lengths:
13
avg
1
avg
SEGMENT: 22ed segtype:
segment flags:
reads:
read times (msec):
read sequences:
read seq. lengths:
???
inode
2
(0 errs)
avg
8.102 min
2
avg
1.0 min
SEGMENT: 5c77 segtype:
segment flags:
reads:
read times (msec):
read sequences:
read seq. lengths:
persistent
pers defer
1
(0 errs)
avg 13.810 min 13.810 max
1
avg
1.0 min
1 max
176
???
(0 errs)
1.979 min
13.0 min
Performance Management Guide
0.957 max
5.970 sdev
1.310
13 max
13 sdev
0.0
7.786 max
8.418 sdev
0.316
1 max
1 sdev
0.0
13.810 sdev
0.000
1 sdev
0.0
-----------------------------------------------------------------------Detailed Logical Volume Stats
(512 byte blocks)
-----------------------------------------------------------------------VOLUME: /dev/hd4 description: /
reads:
5
(0 errs)
read sizes (blks):
avg
22.4 min
read times (msec):
avg
4.847 min
read sequences:
3
read seq. lengths:
avg
37.3 min
seeks:
3
(60.0%)
seek dist (blks):
init
6344,
avg
40.0 min
time to next req(msec): avg 70.473 min
throughput:
151.9 KB/sec
utilization:
0.06
VOLUME: /dev/hd2 description: /usr
reads:
2
(0 errs)
read sizes (blks):
avg
8.0 min
read times (msec):
avg
8.078 min
read sequences:
2
read seq. lengths:
avg
8.0 min
seeks:
2
(100.0%)
seek dist (blks):
init 608672,
avg
16.0 min
time to next req(msec): avg 162.160 min
throughput:
21.7 KB/sec
utilization:
0.04
8 max
0.938 max
40 sdev
13.792 sdev
12.8
4.819
8 max
64 sdev
22.9
8 max
72 sdev
32.0
0.224 max 331.020 sdev 130.364
8 max
7.769 max
8 sdev
8.387 sdev
0.0
0.309
8 max
8 sdev
0.0
16 max
16 sdev
0.0
8.497 max 315.823 sdev 153.663
-----------------------------------------------------------------------Detailed Physical Volume Stats
(512 byte blocks)
-----------------------------------------------------------------------VOLUME: /dev/hdisk0 description: N/A
reads:
7
(0 errs)
read sizes (blks):
avg
18.3 min
8 max
40 sdev
12.6
read times (msec):
avg
5.723 min
0.905 max 20.448 sdev
6.567
read sequences:
5
read seq. lengths:
avg
25.6 min
8 max
64 sdev
22.9
seeks:
5
(71.4%)
seek dist (blks):
init 4233888,
avg 171086.0 min
8 max 684248 sdev 296274.2
seek dist (%tot blks):init 48.03665,
avg 1.94110 min 0.00009 max 7.76331 sdev 3.36145
time to next req(msec): avg 50.340 min
0.226 max 315.865 sdev 108.483
throughput:
173.6 KB/sec
utilization:
0.10
Using the filemon command in systems with real workloads would result in much larger reports and might
require more trace buffer space. Space and CPU time consumption for the filemon command can
degrade system performance to some extent. Use a nonproduction system to experiment with the filemon
command before starting it in a production environment. Also, use offline processing and on systems with
many CPUs use the -C all flag with the trace command.
Note: Although the filemon command reports average, minimum, maximum, and standard deviation in its
detailed-statistics sections, the results should not be used to develop confidence intervals or other
formal statistical inferences. In general, the distribution of data points is neither random nor
symmetrical.
Global Reports of the filemon Command
The global reports list the most active files, segments, logical volumes, and physical volumes during the
measured interval. They are shown at the beginning of the filemon report. By default, the logical file and
Chapter 9. File System, Logical Volume, and Disk I/O Performance
177
virtual memory reports are limited to the 20 most active files and segments, respectively, as measured by
the total amount of data transferred. If the -v flag has been specified, activity for all files and segments is
reported. All information in the reports is listed from top to bottom as most active to least active.
Most Active Files:
#MBs Total number of MBs transferred over measured interval for this file. The rows are sorted by this
field in decreasing order.
#opns Number of opens for files during measurement period.
#rds
Number of read calls to file.
#wrs
Number of write calls to file.
file
File name (full path name is in detailed report).
volume:inode
The logical volume that the file resides in and the i-node number of the file in the associated file
system. This field can be used to associate a file with its corresponding persistent segment shown
in the detailed VM segment reports. This field may be blank for temporary files created and
deleted during execution.
The most active files are smit.log on logical volume hd4 and file null. The application utilizes the terminfo
database for screen management; so the ksh.cat and cmdtrace.cat are also busy. Anytime the shell
needs to post a message to the screen, it uses the catalogs for the source of the data.
To identify unknown files, you can translate the logical volume name, /dev/hd1, to the mount point of the
file system, /home, and use the find or the ncheck command:
# find / -inum 858 -print
/smit.log
or
# ncheck -i 858 /
/:
858
/smit.log
Most Active Segments:
#MBs Total number of MBs transferred over measured interval for this segment. The rows are sorted by
this field in decreasing order.
#rpgs Number of 4-KB pages read into segment from disk.
#wpgs
Number of 4-KB pages written from segment to disk (page out).
#segid
VMM ID of memory segment.
segtype
Type of segment: working segment, persistent segment (local file), client segment (remote file),
page table segment, system segment, or special persistent segments containing file system data
(log, root directory, .inode, .inodemap, .inodex, .inodexmap, .indirect, .diskmap).
volume:inode
For persistent segments, name of logical volume that contains the associated file and the file’s
i-node number. This field can be used to associate a persistent segment with its corresponding
file, shown in the Detailed File Stats reports. This field is blank for nonpersistent segments.
178
Performance Management Guide
If the command is still active, the virtual memory analysis tool svmon can be used to display more
information about a segment, given its segment ID (segid), as follows: svmon -D segid. See The svmon
Command for a detailed discussion.
In our example, the segtype ??? means that the system cannot identify the segment type, and you must
use the svmon command to get more information.
Most Active Logical Volumes:
util
Utilization of logical volume.
#rblk
Number of 512-byte blocks read from logical volume.
#wblk Number of 512-byte blocks written to logical volume.
KB/s
Average transfer data rate in KB per second.
volume
Logical volume name.
description
Either the file system mount point or the logical volume type (paging, jfslog, boot, or sysdump). For
example, the logical volume /dev/hd2 is /usr; /dev/hd6 is paging, and /dev/hd8 is jfslog. There
may also be the word compressed. This means all data is compressed automatically using
Lempel-Zev (LZ) compression before being written to disk, and all data is uncompressed
automatically when read from disk (see Compression for details).
The utilization is presented in percentage, 0.06 indicates 6 percent busy during measured interval.
Most Active Physical Volumes:
util
Utilization of physical volume.
Note: Logical volume I/O requests start before and end after physical volume I/O requests. Total
logical volume utilization will appear therefore to be higher than total physical volume
utilization.
#rblk
Number of 512-byte blocks read from physical volume.
#wblk Number of 512-byte blocks written to physical volume.
KB/s
Average transfer data rate in KB per second.
volume
Physical volume name.
description
Simple description of the physical volume type, for example, SCSI Multimedia CD-ROM Drive or
16 Bit SCSI Disk Drive.
The utilization is presented in percentage, 0.10 indicates 10 percent busy during measured interval.
Detailed Reports of the filemon Command
The detailed reports give additional information for the global reports. There is one entry for each reported
file, segment, or volume in the detailed reports. The fields in each entry are described below for the four
detailed reports. Some of the fields report a single value; others report statistics that characterize a
distribution of many values. For example, response-time statistics are kept for all read or write requests
that were monitored. The average, minimum, and maximum response times are reported, as well as the
standard deviation of the response times. The standard deviation is used to show how much the individual
response times deviated from the average. Approximately two-thirds of the sampled response times are
Chapter 9. File System, Logical Volume, and Disk I/O Performance
179
between average minus standard deviation (avg - sdev) and average plus standard deviation (avg + sdev).
If the distribution of response times is scattered over a large range, the standard deviation will be large
compared to the average response time.
Detailed File Stats: Detailed file statistics are provided for each file listed in the Most Active Files
report. These stanzas can be used to determine what access has been made to the file. In addition to the
number of total bytes transferred, opens, reads, writes, and lseeks, the user can also determine the
read/write size and times.
FILE
Name of the file. The full path name is given, if possible.
volume
Name of the logical volume/file system containing the file.
inode I-node number for the file within its file system.
opens Number of times the file was opened while monitored.
total bytes xfrd
Total number of bytes read/written from/to the file.
reads Number of read calls against the file.
read sizes (bytes)
Read transfer-size statistics (avg/min/max/sdev), in bytes.
read times (msec)
Read response-time statistics (avg/min/max/sdev), in milliseconds.
writes Number of write calls against the file.
write sizes (bytes)
Write transfer-size statistics.
write times (msec)
Write response-time statistics.
lseeks
Number of lseek() subroutine calls.
The read sizes and write sizes will give you an idea of how efficiently your application is reading and
writing information. Use a multiple of 4 KB pages for best results.
Detailed VM Segment Stats: Each element listed in the Most Active Segments report has a
corresponding stanza that shows detailed information about real I/O to and from memory.
SEGMENT
Internal operating system’s segment ID.
segtype
Type of segment contents.
segment flags
Various segment attributes.
volume
For persistent segments, the name of the logical volume containing the corresponding file.
inode For persistent segments, the i-node number for the corresponding file.
reads Number of 4096-byte pages read into the segment (that is, paged in).
read times (msec)
Read response-time statistics (avg/min/max/sdev), in milliseconds.
180
Performance Management Guide
read sequences
Number of read sequences. A sequence is a string of pages that are read (paged in)
consecutively. The number of read sequences is an indicator of the amount of sequential access.
read seq. lengths
Statistics describing the lengths of the read sequences, in pages.
writes Number of pages written from the segment to disk (that is, paged out).
write times (msec)
Write response-time statistics.
write sequences
Number of write sequences. A sequence is a string of pages that are written (paged out)
consecutively.
write seq. lengths
Statistics describing the lengths of the write sequences, in pages.
By examining the reads and read-sequence counts, you can determine if the access is sequential or
random. For example, if the read-sequence count approaches the reads count, the file access is more
random. On the other hand, if the read-sequence count is significantly smaller than the read count and the
read-sequence length is a high value, the file access is more sequential. The same logic applies for the
writes and write sequence.
Detailed Logical/Physical Volume Stats: Each element listed in the Most Active Logical Volumes /
Most Active Physical Volumes reports will have a corresponding stanza that shows detailed information
about the logical/physical volume. In addition to the number of reads and writes, the user can also
determine read and write times and sizes, as well as the initial and average seek distances for the logical /
physical volume.
VOLUME
Name of the volume.
description
Description of the volume. (Describes contents, if dealing with a logical volume; describes type, if
dealing with a physical volume.)
reads Number of read requests made against the volume.
read sizes (blks)
Read transfer-size statistics (avg/min/max/sdev), in units of 512-byte blocks.
read times (msec)
Read response-time statistics (avg/min/max/sdev), in milliseconds.
read sequences
Number of read sequences. A sequence is a string of 512-byte blocks that are read consecutively.
It indicates the amount of sequential access.
read seq. lengths
Statistics describing the lengths of the read sequences, in blocks.
writes Number of write requests made against the volume.
write sizes (blks)
Write transfer-size statistics.
write times (msec)
Write-response time statistics.
write sequences
Number of write sequences. A sequence is a string of 512-byte blocks that are written
consecutively.
Chapter 9. File System, Logical Volume, and Disk I/O Performance
181
write seq. lengths
Statistics describing the lengths of the write sequences, in blocks.
seeks Number of seeks that preceded a read or write request; also expressed as a percentage of the
total reads and writes that required seeks.
seek dist (blks)
Seek-distance statistics in units of 512-byte blocks. In addition to the usual statistics
(avg/min/max/sdev), the distance of the initial seek operation (assuming block 0 was the starting
position) is reported separately. This seek distance is sometimes very large; it is reported
separately to avoid skewing the other statistics.
seek dist (cyls)
(Physical volume only) Seek-distance statistics in units of disk cylinders.
time to next req
Statistics (avg/min/max/sdev) describing the length of time, in milliseconds, between consecutive
read or write requests to the volume. This column indicates the rate at which the volume is being
accessed.
throughput
Total volume throughput in KB per second.
utilization
Fraction of time the volume was busy. The entries in this report are sorted by this field in
decreasing order.
A long seek time can increase I/O response time and result in decreased application performance. By
examining the reads and read sequence counts, you can determine if the access is sequential or random.
The same logic applies to the writes and write sequence.
Guidelines for Using the filemon Command
Following are some guidelines for using the filemon command:
v The /etc/inittab file is always very active. Daemons specified in /etc/inittab are checked regularly to
determine whether they are required to be respawned.
v The /etc/passwd file is also always very active. Because files and directories access permissions are
checked.
v A long seek time increases I/O response time and decreases performance.
v If the majority of the reads and writes require seeks, you might have fragmented files and overly active
file systems on the same physical disk. However, for online transaction processing (OLTP) or database
systems this behavior might be normal.
v If the number of reads and writes approaches the number of sequences, physical disk access is more
random than sequential. Sequences are strings of pages that are read (paged in) or written (paged out)
consecutively. The seq. lengths is the length, in pages, of the sequences. A random file access can
also involve many seeks. In this case, you cannot distinguish from the filemon output if the file access
is random or if the file is fragmented. Use the fileplace command to investigate further.
v Remote files are listed in the volume:inode column with the remote system name.
Because the filemon command can potentially consume some CPU power, use this tool with discretion,
and analyze the system performance while taking into consideration the overhead involved in running the
tool. Tests have shown that in a CPU-saturated environment:
v With little I/O, the filemon command slowed a large compile by about one percent.
v With a high disk-output rate, the filemon command slowed the writing program by about five percent.
Summary for Monitoring Disk I/O
In general, a high % iowait indicates that the system has an application problem, a memory shortage, or
an inefficient I/O subsystem configuration. For example, the application problem might be due to
182
Performance Management Guide
requesting a lot of I/O, but not doing much with the data. Understanding the I/O bottleneck and improving
the efficiency of the I/O subsystem is the key in solving this bottleneck. Disk sensitivity can come in a
number of forms, with different resolutions. Some typical solutions might include:
v Limiting number of active logical volumes and file systems placed on a particular physical disk. The idea
is to balance file I/O evenly across all physical disk drives.
v Spreading a logical volume across multiple physical disks. This is particularly useful when a number of
different files are being accessed.
v Creating multiple Journaled File Systems (JFS) logs for a volume group and assigning them to specific
file systems (preferably on fast write cache devices). This is beneficial for applications that create,
delete, or modify a large number of files, particularly temporary files.
v If the iostat output indicates that your workload I/O activity is not evenly distributed among the system
disk drives, and the utilization of one or more disk drives is often 70-80 percent or more, consider
reorganizing file systems, such as backing up and restoring file systems to reduce fragmentation.
Fragmentation causes the drive to seek excessively and can be a large portion of overall response time.
v If large, I/O-intensive background jobs are interfering with interactive response time, you may want to
activate I/O pacing.
v If it appears that a small number of files are being read over and over again, consider whether
additional real memory would allow those files to be buffered more effectively.
v If the workload’s access pattern is predominantly random, you might consider adding disks and
distributing the randomly accessed files across more drives.
v If the workload’s access pattern is predominantly sequential and involves multiple disk drives, you might
consider adding one or more disk adapters. It may also be appropriate to consider building a striped
logical volume to accommodate large, performance-critical sequential files.
v Using fast write cache devices.
v Using asynchronous I/O.
Each technique is discussed later in this chapter.
Guidelines for Tuning File Systems
This section discusses file system tuning:
v File System Types
v
v
v
v
Differences Between JFS and Enhanced JFS
Potential Performance Inhibitors for JFS and Enhanced JFS
Performance Enhancements
Summary of Tunable Parameters
To review basic information about file systems, see AIX 5L Version 5.1 System Management Concepts:
Operating Systems and Devices.
File System Types
This section discusses the multiple file systems supported on AIX 5.1.
v Journaled File System
v Enhanced Journaled File System
v General Parallel File System
v Network File System
v RAM File System
Journaled File System
Journaled File System (JFS) is the default file system for AIX 4.3.3 and earlier, as well as for AIX 5.1
running under a 32-bit kernel.
Chapter 9. File System, Logical Volume, and Disk I/O Performance
183
A journaling file system allows for quick file system recovery after a crash has occurred by logging the
metadata of files. By enabling file-system logging, the system records every change in the metadata of the
file into a reserved area of the file system. The actual write operations are performed after the logging of
changes to the metadata has been completed.
Enhanced Journal File System
Enhanced JFS (also known as JFS2) is another native AIX journaling file system that was introduced in
AIX 5.1. Enhanced JFS is the default file system for 64-bit kernel environments. Due to address space
limitations of the 32–bit kernel, Enhanced JFS is not recommended for use in 32-bit kernel environments.
General Parallel File System (GPFS)
The General Parallel File System (GPFS) is a high-performance, shared-disk file system that can provide
fast data access to all nodes in a server cluster. Parallel and serial applications access files using standard
UNIX file system interfaces, such as those in AIX.
GPFS provides high performance by striping I/O across multiple disks, high availability through logging,
replication, and both server and disk failover.
For more information, see the IBM Redbook entitled GPFS on AIX Clusters: High Performance File
System Administration Simplified at http://www.redbooks.ibm.com/redbooks/SG246035.html.
Network File System
The Network File System (NFS) is a distributed file system that allows users to access files and directories
located on remote computers and treat those files and directories as if they were local. For example, users
can use operating system commands to create, remove, read, write, and set file attributes for remote files
and directories.
Performance tuning and other issues regarding NFS are found in the chapter Monitoring and Tuning NFS
Use.
RAM File System
A RAM disk is a simulated disk drive that resides in memory. RAM disks are designed to have significantly
higher I/O performance over physical drives, and are typically used to overcome I/O bottlenecks with
nonpersistent files. Do not use RAM disks for persistent data, as all data is lost if the system crashes or
reboots.
To set up a RAM disk, use the mkramdisk command. The following example illustrates how to set up a
RAM disk that is approximately 20 MB in size and how to create a file system on that RAM disk:
mkramdisk 40000
ls -l /dev | grep ram
mkfs -V jfs /dev/ramdiskx
mkdir /ramdiskx
mount -V jfs -o nointegrity /dev/ramdiskx /ramdiskx
where x is the logical RAM disk number. To remove a RAM disk, use the rmramdisk command. RAM
disks are also removed when the machine is rebooted.
Beginning with AIX 5.2, the previous maximum size limit of 2GB has been removed and the maximum size
of a RAM disk is now just limited by the amount of available system memory.
Differences Between JFS and Enhanced JFS
This section discusses some of the main differences between JFS and Enhanced JFS.
Kernel Address Space
AIX 5.1 offers two different types of kernels, a 32-bit kernel and a 64-bit kernel. The 32-bit and 64-bit
kernels have common libraries, commands and utilities, and header files. However, the 64-bit kernel offers
a degree of scaling for 64-bit hardware that the 32-bit kernel cannot.
184
Performance Management Guide
JFS is optimized for 32-bit kernel, and will run on 32-bit and 64-bit hardware. Enhanced JFS is optimized
to run on the 64-bit kernel, allowing it to take advantage of 64-bit functionality.
For a description of kernel address space issues and differences, see POWER-based-Architecture-Unique
Timer Access.
Journaling
Before writing actual data, a journaling file system logs the metadata, thus incurring an overhead penalty
that slows write throughput. One way of improving performance under JFS is to disable metadata logging
by using the nointegrity mount option. Note that the enhanced performance is achieved at the expense of
metadata integrity. Therefore, use this option with extreme caution because a system crash can make a
file system mounted with this option unrecoverable.
In contrast to JFS, Enhanced JFS does not allow you to disable metadata logging. However, the
implementation of journaling on Enhanced JFS makes it more suitable to handle metadata-intensive
applications. Thus, the performance penalty is not as high under Enhanced JFS as it is under JFS.
Directory Organization
An index node, or i-node, is a data structure that stores all file and directory properties. When a program
looks up a file, it searches for the appropriate i-node by looking up a file name in a directory. Because
these operations are performed very often, the mechanism used for searching is of particular importance.
JFS employs a linear organization for its directories, thus making searches linear as well. In contrast,
Enhanced JFS employs a binary tree representation for directories, thus greatly accelerating access to
files.
Scaling
The main advantage of using Enhanced JFS over JFS is scaling. Enhanced JFS provides the capability to
store much larger files than the existing JFS. The maximum size of a file under JFS is 64 gigabytes.
Under Enhanced JFS, AIX currently supports files up to 16 terabytes in size, although the file system
architecture is set up to eventually handle file sizes of up to 4 petabytes.
Another scaling issue relates to accessing a large number of files. The following illustration shows how
Enhanced JFS can improve performance for this type of access.
Chapter 9. File System, Logical Volume, and Disk I/O Performance
185
Figure 18.
For this example, we are creating, deleting, and searching directories with unique, 10-byte file names. The
results show that creating and deleting files is much faster under Enhanced JFS than under JFS.
Performance for searches was approximately the same for both file system types.
A second example shows how results for create, delete, and search operations are generally much faster
on Enhanced JFS than on JFS when using non-unique file names. In this second example, file names
were chosen to have the same first 64-bytes, appended by 10-byte unique names. The following
illustration shows the result for this test.
186
Performance Management Guide
Figure 19.
On a related note, beginning with AIX 5.2, caching of long (greater than 32 characters) file names is
supported in both the JFS and Enhanced JFS name caches. This improves the performance of directory
operations, such as ls and find on directories with numerous long file name entries.
Miscellaneous Differences Between JFS and Enhanced JFS
Cloning with a system backup (mksysb) from a 64-bit enabled JFS2 system to a 32-bit system will not be
succesful.
Unlike the JFS file system, the JFS2 file system will not allow the link() API to be used on its binary type
directory. This limitation may cause some applications that operate correctly on a JFS file system to fail on
a JFS2 file system.
Summary of Differences Between JFS and Enhanced JFS
The following table summarizes the differences between JFS and Enhanced JFS.
Table 2. Functional Differences between JFS and Enhanced JFS
Function
JFS
Enhanced JFS
Optimization
32-bit kernel
64-bit kernel
Maximum file system size
1 terabyte
4 petabytes1
Maximum file size
64 gigabytes
4 petabytes1
Number of I-nodes
Fixed at file system creation
Dynamic, limited by disk space
Large file support
As mount option
Default
Online defragmentation
Yes
Yes
namefs
Yes
Yes
DMAPI
Yes
No
Chapter 9. File System, Logical Volume, and Disk I/O Performance
187
Table 2. Functional Differences between JFS and Enhanced JFS (continued)
Function
JFS
Enhanced JFS
Compression
Yes
No
Quotas
Yes
No
Deferred update
Yes
No
Direct I/O support
Yes
Yes
Note: 1. This is an architectural limit. AIX currently only supports up to 16 terabytes.
Potential Performance Inhibitors for JFS and Enhanced JFS
This section discusses situations that can potentially inhibit JFS and Enhanced JFS performance.
Effects of File System Logging on File System Throughput
Because write operations are performed after logging of metadata has been completed, write throughput
can be affected. For a description of how to avoid performance penalties associated with logging, see
Monitoring and Tuning Disk I/O.
Compression and fragmentation
The Journaled File System supports fragmented and compressed file systems as a means of saving disk
space. On average, data compression saves disk space by about a factor of two. However, the
fragmentation and compression might incur performance loss associated with increased allocation activity.
For a description of how compression and fragmentation might affect performance, see Monitoring and
Tuning Disk I/O.
To enhance performance, both JFS and Enhanced JFS allow for online defragmentation. The file system
can be mounted and is accessible while the defragmentation process is underway.
Performance Enhancements
This section discusses policies and mechanisms that you can use to enhance file system performance
under AIX.
Release-behind Mechanism for JFS and Enhanced JFS
Release-behind is a mechanism under which pages are freed as soon as they are either committed to
permanent storage (by writes) or delivered to an application (by reads). This solution addresses a scaling
problem when performing sequential I/O on very large files whose pages may not need to be re-accessed
in the near future.
When writing a large file without using release-behind, writes will go very fast whenever there are available
pages on the free list. When the number of pages drops to minfree, VMM uses its Least Recently Used
(LRU) algorithm to find candidate pages for eviction. As part of this process, VMM needs to acquire a lock
that is also being used for writing. This lock contention can cause a sharp performance degradation.
You enable this mechanism by specifying either the rbr flag (release-behind, sequential-read), the rbw flag
(release-behind, sequential-write), or the rbrw flag (release-behind, sequential-read and -write) when
issuing the mount command.
A side-effect of using the release-behind mechanism is that you will notice an increase in CPU utilization
for the same read or write throughput rate without release-behind. This is due to the work of freeing
pages, which would normally be handled at a later time by the LRU daemon. Also note that all file page
accesses result in disk I/O since file data is not cached by VMM.
188
Performance Management Guide
Delayed Write Operations
JFS allows you to defer updates of data into permanent storage. Delayed write operations save extra disk
operations for files that are often rewritten. You can enable this feature by opening the file with the
deferred update flag, O_DEFER. This feature caches the data, allowing faster read and write operations
from other processes.
When writing to files that have been opened with this flag, data is not committed to permanent storage
until a process issues the fsync command, forcing all updated data to be committed to disk. Also, if a
process issues a synchronous write operation on a file, that is, the process has opened the file with the
O_SYNC flag, the operation is not deferred even if the file was created with the O_DEFER flag.
Note: This feature is not available for Enhanced JFS.
Direct I/O support
Both JFS and Enhanced JFS offer support for Direct I/O access to files. This access method bypasses the
file cache and transfers data directly from disk into the user space buffer, as opposed to using the normal
cache policy of placing pages in kernel memory. For a description on how to tune Direct I/O, see
Monitoring and Tuning Disk I/O.
Summary of Tunable Parameters
The following table summarizes tunable parameters for JFS and Enhanced JFS file systems.
Table 3. JFS and Enhanced JFS Tunable Parameters
Function
JFS Tuning Parameter
Enhanced JFS Tuning Parameter
Sets the maximum amount of
memory for caching files
vmtune -P maxperm
vmtune -t maxclient (less than or
equal to maxperm)
Sets the minimum amount of memory vmtune -p minperm
for caching files
No equivalent
Sets a hard limit on memory for
caching files
vmtune -h strict_maxperm
vmtune -t maxclient (always a hard
limit)
Sets the maximum pages used for
sequential read ahead
vmtune -R maxpgahead
vmtune -Q j2_maxPageReadAhead
Sets the mimimum pages used for
sequential read ahead
vmtune -r minpgahead
vmtune -q j2_minPageReadAhead
Sets the maximum number of
pending I/Os to a file
chdev -l sys0 -a maxpout maxpout
chdev -l sys0 -a maxpout maxpout
Sets the minimum number of peding
I/Os to a file at which programs
blocked by maxpout may proceed
chdev -l sys0 -a minpout minpout
chdev -l sys0 -a minpout minpout
Sets the amount of modified data
cache for a file with random writes
vmtune -W maxrandwrt
vmtune -J j2_maxRandomWrite
vmtune -z j2_nRandomCluster
Controls the gathering of I/Os for
sequential write behind
vmtune -C numclust
vmtune -j
j2_nPagesPerWriteBehindCluster
Sets the number of file system
bufstructs
vmtune -b numfsbufs
vmtune -Z j2_nBufferPerPagerDevice
Note: The amount of memory for caching files for Enhanced JFS, tuning parameter maxclient, is a
subset of the amount of memory for file caching for JFS. The value of maxclient cannot exceed the
value of maxperm.
Chapter 9. File System, Logical Volume, and Disk I/O Performance
189
Related Information
v
v
v
v
v
The mount command.
POWER-based-Architecture-Unique Timer Access
Monitoring and Tuning Disk I/O
Kernel Extensions and Device Support Programming Concepts
Monitoring and Tuning NFS Use
Changing File System Attributes that Affect Performance
The longer a file system is used, the more fragmented it becomes. With the dynamic allocation of
resources, file blocks become more and more scattered, logically contiguous files become fragmented, and
logically contiguous logical volumes (LV) become fragmented.
When files are accessed from disk, the following take effect:
v Sequential access is no longer sequential
v Random access is slower
v Access time is dominated by longer seek time
However, once the file is in memory, these effects diminish. File system performance is also affected by
physical considerations, such as:
v Types of disks and number of adapters
v Amount of memory for file buffering
v Amount of local versus remote file access
v Pattern and amount of file access by application
AIX Version 4 enables the ability to have JFS file systems larger than 2 GB in size. In addition file system
fragmentation allows for better space utilization by subdividing 4 K blocks. The number of bytes per i-node
(NBPI) is used to control how many i-nodes are created for a file system. Compression can be used for
file systems with a fragment size less than 4 KB. Fragment size and compression affect performance and
are discussed in the following.
JFS File-System Fragment Size
The fragments feature in AIX Version 4 allows the space in a file system to be allocated in less than 4 KB
chunks. When a file system is created, the system administrator can specify the size of the fragments in
the file system. The allowable sizes are 512, 1024, 2048, and 4096 bytes (the default). Files smaller than
a fragment are stored in a single fragment, conserving disk space, which is the primary objective.
Files smaller than 4096 bytes are stored in the minimum necessary number of contiguous fragments. Files
whose size is between 4096 bytes and 32 KB (inclusive) are stored in one or more (4 KB) full blocks and
in as many fragments as are required to hold the remainder. For example, a 5632-byte file would be
allocated one 4 KB block, pointed to by the first pointer in the i-node. If the fragment size is 512, then
eight fragments would be used for the first 4 KB block. The last 1.5 KB would use three fragments, pointed
to by the second pointer in the i-node. For files greater than 32 KB, allocation is done in 4 KB blocks, and
the i-node pointers point to these 4 KB blocks.
Whatever the fragment size, a full block is considered to be 4096 bytes. In a file system with a fragment
size less than 4096 bytes, however, a need for a full block can be satisfied by any contiguous sequence of
fragments totalling 4096 bytes. It need not begin on a multiple-of-4096-byte boundary.
The file system tries to allocate space for files in contiguous fragments by spreading the files themselves
across the logical volume to minimize interfile allocation interference and fragmentation.
190
Performance Management Guide
The primary performance hazard for file systems with small fragment sizes is space fragmentation. The
existence of small files scattered across the logical volume can make it impossible to allocate contiguous
or closely spaced blocks for a large file. Performance can suffer when accessing large files. Carried to an
extreme, space fragmentation can make it impossible to allocate space for a file, even though there are
many individual free fragments.
Another adverse effect on disk I/O activity is the number of I/O operations. For a file with a size of 4 KB
stored in a single fragment of 4 KB, only one disk I/O operation would be required to either read or write
the file. If the choice of the fragment size was 512 bytes, eight fragments would be allocated to this file,
and for a read or write to complete, several additional disk I/O operations (disk seeks, data transfers, and
allocation activity) would be required. Therefore, for file systems which use a fragment size of 4 KB, the
number of disk I/O operations might be far less than for file systems which employ a smaller fragment
size.
Part of a decision to create a small-fragment file system should be a policy for defragmenting the space in
that file system with the defragfs command. This policy must also take into account the performance cost
of running the defragfs command (see Defragmenting a File System).
JFS Compression
If a file system is compressed, all data is compressed automatically using Lempel-Zev (LZ) compression
before being written to disk, and all data is uncompressed automatically when read from disk. The LZ
algorithm replaces subsequent occurrences of a given string with a pointer to the first occurrence. On an
average, a 50 percent savings in disk space is realized.
In AIX Version 4, file system data is compressed at the level of an individual logical blocks. To compress
data in large units (all the logical blocks of a file together, for example) would result in the loss of more
available disk space. By individually compressing a file’s logical blocks, random seeks and updates are
carried out much more rapidly.
When a file is written into a file system for which compression is specified, the compression algorithm
compresses the data 4096 bytes (a page) at a time, and the compressed data is then written in the
minimum necessary number of contiguous fragments. Obviously, if the fragment size of the file system is 4
KB, there is no disk-space payback for the effort of compressing the data. Therefore, compression
requires fragmentation to be used, with a fragment size smaller than 4096.
Although compression should result in conserving space overall, there are valid reasons for leaving some
unused space in the file system:
v Because the degree to which each 4096-byte block of data will compress is not known in advance, the
file system initially reserves a full block of space. The unneeded fragments are released after
compression, but the conservative initial allocation policy may lead to premature ″out of space″
indications.
v Some free space is necessary to allow the defragfs command to operate.
In addition to increased disk I/O activity and free-space fragmentation problems, file systems using data
compression have the following performance considerations:
v Degradation in file system usability arising as a direct result of the data compression/decompression
activity. If the time to compress and decompress data is quite lengthy, it might not always be possible to
use a compressed file system, particularly in a busy commercial environment where data needs to be
available immediately.
v All logical blocks in a compressed file system, when modified for the first time, will be allocated 4096
bytes of disk space, and this space will subsequently be reallocated when the logical block is written to
disk. Performance costs are associated with reallocation, which does not occur in noncompressed file
systems.
Chapter 9. File System, Logical Volume, and Disk I/O Performance
191
v To perform data compression, approximately 50 CPU cycles per byte are required, and about 10 CPU
cycles per byte for decompression. Data compression therefore places a load on the processor by
increasing the number of processor cycles.
v The JFS compression kproc (jfsc) runs at a fixed priority of 30 so that while
compression/decompression is occurring, the CPU that this kproc is running on may not be available to
other processes unless they run at a better priority.
Changing Logical Volume Attributes That Affect Performance
Various factors have performance implications and can be controlled when creating a logical volume.
These options appear as prompts for values on the smitty mklv screen.
Position on Physical Volume
The Intra-Physical Volume Allocation Policy specifies what strategy should be used for choosing physical
partitions on a physical volume. The five general strategies are edge, inner-edge, middle, inner-middle,
and center.
Figure 20. Intra-Physical Volume Allocation Policy. This figure illustrates storage position on a physical volume or disk.
The disk is formatted into hundreds of tracks beginning at the outer edge of the disk and moving toward the center of
the disk. Because of the way a disk is read (the tracks spinning under a movable read/write head), data that is written
toward the center of the disk will have faster seek times than data that is written on the outer edge. In part, this is due
to the mechanical action of the read/write head and the sectors of each track having to pass under the head. Data is
more dense as it moves toward the center, resulting in less physical movement of the head. This results in faster
overall throughput.
Physical partitions are numbered consecutively, starting with number one, from the outer-most edge to the
inner-most edge.
The edge and inner-edge strategies specify allocation of partitions to the edges of the physical volume.
These partitions have the slowest average seek times, which generally result in longer response times for
any application that uses them. Edge on disks produced since the mid-1990s can hold more sectors per
track so that the edge is faster for sequential I/O.
The middle and inner-middle strategies specify to avoid the edges of the physical volume and out of the
center when allocating partitions. These strategies allocate reasonably good locations for partitions with
reasonably good average seek times. Most of the partitions on a physical volume are available for
allocation using this strategy.
The center strategy specifies allocation of partitions to the center section of each physical volume. These
partitions have the fastest average seek times, which generally result in the best response time for any
application that uses them. Fewer partitions on a physical volume satisfy the center strategy than any
other general strategy.
192
Performance Management Guide
The paging space logical volume is a good candidate for allocation at the center of a physical volume if
there is lot of paging activity. At the other extreme, the dump and boot logical volumes are used
infrequently and, therefore, should be allocated at the beginning or end of the physical volume.
The general rule, then, is that the more I/Os, either absolutely or in the course of running an important
application, the closer to the center of the physical volumes the physical partitions of the logical volume
should be allocated.
Range of Physical Volumes
The Inter-Physical Volume Allocation Policy specifies which strategy should be used for choosing physical
devices to allocate the physical partitions of a logical volume. The choices are the minimum and maximum
options.
Figure 21. Inter-Physical Volume Allocation Policy. This illustration shows 2 physical volumes. One contains partition 1
and a copy of partition 2. The other contains partition 2 with a copy of partition 1. The formula for allocation is
Maximum Inter-Disk Policy (Range=maximum) with a Single Logical Volume Copy per Disk (Strict=y).
The minimum option indicates the number of physical volumes used to allocate the required physical
partitions. This is generally the policy to use to provide the greatest reliability and availability, without
having copies, to a logical volume. Two choices are available when using the minimum option, with copies
and without, as follows:
v Without Copies: The minimum option indicates one physical volume should contain all the physical
partitions of this logical volume. If the allocation program must use two or more physical volumes, it
uses the minimum number possible, remaining consistent with the other parameters.
v With Copies: The minimum option indicates that as many physical volumes as there are copies should
be used. If the allocation program must use two or more physical volumes, the minimum number of
physical volumes possible are used to hold all the physical partitions. At all times, the constraints
imposed by other parameters such as the strict option are observed.
These definitions are applicable when extending or copying an existing logical volume. The existing
allocation is counted to determine the number of physical volumes to use in the minimum with copies
case, for example.
The maximum option indicates the number of physical volumes used to allocate the required physical
partitions. The maximum option intends, considering other constraints, to spread the physical partitions of
this logical volume over as many physical volumes as possible. This is a performance-oriented option and
should be used with copies to improve availability. If an uncopied logical volume is spread across multiple
physical volumes, the loss of any physical volume containing a physical partition from that logical volume
is enough to cause the logical volume to be incomplete.
Chapter 9. File System, Logical Volume, and Disk I/O Performance
193
Maximum Number of Physical Volumes to Use for Allocation
Sets the maximum number of physical volumes for new allocation. The value should be between one and
the total number of physical volumes in the volume group. This option relates to Range of Physical
Volumes.
Mirror Write Consistency
The LVM always ensures data consistency among mirrored copies of a logical volume during normal I/O
processing. For every write to a logical volume, the LVM generates a write request for every mirror copy. A
problem arises if the system crashes in the middle of processing a mirrored write (before all copies are
written). If mirror write consistency recovery is requested for a logical volume, the LVM keeps additional
information to allow recovery of these inconsistent mirrors. Mirror write consistency recovery should be
performed for most mirrored logical volumes. Logical volumes, such as the page space that do not use the
existing data when the volume group is re-varied on, do not need this protection.
The Mirror Write Consistency (MWC) record consists of one sector. It identifies which logical partitions may
be inconsistent if the system is not shut down correctly. When the volume group is varied back on-line, this
information is used to make the logical partitions consistent again.
Note: With Mirror Write Consistency LVs, because the MWC control sector is on the edge of the disk,
performance may be improved if the mirrored logical volume is also on the edge.
Beginning in AIX 5, a mirror write consistency option called Passive Mirror Write Consistency (MWC) is
available. The default mechanism for ensuring mirror write consistency is Active MWC. Active MWC
provides fast recovery at reboot time after a crash has occurred. However, this benefit comes at the
expense of write performance degradation, particularly in the case of random writes. Disabling Active
MWC eliminates this write-performance penalty, but upon reboot after a crash you must use the syncvg -f
command to manually synchronize the entire volume group before users can access the volume group. To
achieve this, automatic vary-on of volume groups must be disabled.
Enabling Passive MWC not only eliminates the write-performance penalty associated with Active MWC, but
logical volumes will be automatically resynchronized as the partitions are being accessed. This means that
the administrator does not have to synchronize logical volumes manually or disable automatic vary-on. The
disadvantage of Passive MWC is that slower read operations may occur until all the partitions have been
resynchronized.
You can select either mirror write consistency option within SMIT when creating or changing a logical
volume. The selection option takes effect only when the logical volume is mirrored (copies > 1).
Allocate Each Logical Partition Copy on a Separate PV
Specifies whether to follow the strict allocation policy. Strict allocation policy allocates each copy of a
logical partition on a separate physical volume. This option relates to Range of Physical Volumes.
Relocate the Logical Volume During Reorganization?
Specifies whether to allow the relocation of the logical volume during reorganization. For striped logical
volumes, the relocate parameter must be set to no (the default for striped logical volumes). Depending on
your installation you may want to relocate your logical volume.
Scheduling Policy for Reading/Writing Logical Partition Copies
Different scheduling policies can be set for the logical volume. Different types of scheduling policies are
used for logical volumes with multiple copies, as follows:
v The parallel policy balances reads between the disks. On each read, the system checks whether the
primary is busy. If it is not busy, the read is initiated on the primary. If the primary is busy, the system
194
Performance Management Guide
checks the secondary. If it is not busy, the read is initiated on the secondary. If the secondary is busy,
the read is initiated on the copy with the least number of outstanding I/Os. Writes are initiated
concurrently.
v The parallel/sequential policy always initiates reads on the primary copy. Writes are initiated
concurrently.
v The parallel/round robin policy is similar to the parallel policy except that instead of always checking the
primary copy first, it alternates between the copies. This results in equal utilization for reads even when
there is never more than one I/O outstanding at a time. Writes are initiated concurrently.
v The sequential policy results in all reads being issued to the primary copy. Writes happen serially, first to
the primary disk; only when that is completed is the second write initiated to the secondary disk.
For data that has only one physical copy, the logical volume device driver translates a logical read or write
request address into a physical address and calls the appropriate physical device driver to service the
request. This single-copy policy handles Bad Block Relocation for write requests and returns all read
errors to the calling process.
Mirroring-scheduling policies, such as parallel and parallel/round-robin, can allow performance on
read-intensive mirrored configurations to be equivalent to non-mirrored ones. Typically, performance on
write-intensive mirrored configurations is less than non-mirrored, unless more disks are used.
Enable Write Verify
Specifies whether to verify all writes to the logical volume with a follow-up read. Setting this to Yes has an
impact on performance.
Striping Size
When defining a striped logical volume, at least two physical drives are required. Stripe size can be any
power of 2 from 4 KB to 128 KB. The size of the logical volume in partitions must be an integral multiple of
the number of disk drives used. See Tuning Logical Volume Striping for a detailed discussion.
Physical Volume Considerations
The major performance issue for disk drives is application-related; that is, whether large numbers of small
accesses will be made (random), or smaller numbers of large accesses (sequential). For random access,
performance will generally be better using larger numbers of smaller capacity drives. The opposite
situation exists for sequential access (use faster drives or use striping with larger number of drives).
Volume Group Recommendations
If possible, for easier system management and better performance, the default volume group, rootvg,
should consist of only the physical volume on which the operating system is initially installed. Maintaining
only operating systems in the rootvg is a good decision because operating system updates, reinstallations,
and crash recovery can be accomplished without endangering user data. Updates and reinstallations can
be done more quickly because only the operating system is included in the changes.
One or more other volume groups should be defined for the other physical volumes that hold the user
data. Having user data in alternate volume groups allows easier exporting of that data to other systems.
Place a highly active file system on one disk and the log for that file system on another if the activity would
generate a lot of log transactions (see Reorganizing JFS Log and Log Logical Volumes). Cached devices
(such as solid-state disks, SSA with Fast Write Cache, or disk arrays with write-cache) can provide for
much better performance for log logical volumes (JFS log or database logs).
Chapter 9. File System, Logical Volume, and Disk I/O Performance
195
Performance Impacts of Mirroring rootvg
Although mirroring is common for customer data, particularly in database environments, it is used less
frequently for system volumes.
In mirroring, when a write occurs, it must occur to all logical volume copies. Typically, this write will take
longer than if the logical volume was not mirrored. Mirroring can also cause additional CPU overhead,
because two disk I/Os take more instructions to complete than one. It is important to understand the layout
of the rootvg logical volumes so one can guess where problems might exist when mirroring the rootvg
logical volumes.
Looking at logical volumes typically found in rootvg, we expect most of the files in /, including the heavily
used /usr/bin where many executable programs reside, to be read-mostly data. The paging space should
have writes only if the amount of physical memory in the system is insufficient to hold the current level of
activity. It is common for systems to page from time to time, but sustained heavy paging usually leads to
poor response time. The addition of physical memory generally resolves this issue.
The /tmp and /var file systems do see file-write activity for many types of applications. Applications, such
as the compiler, often create and write temporary files in the /tmp directory. The /var directory receives
files destined for mail and printer queues. The jfslog is essentially write-only during normal operation. Of
the remaining file systems, perhaps only the /home directory is active during normal operation. Frequently,
user home directories are placed in other file systems, which simplifies rootvg management.
The rootvg can be mirrored by mirroring each logical volume in rootvg with the mklvcopy command; or in
AIX 4.2.1 and later, rootvg can be mirrored using the mirrorvg command.
By default, the mirrorvg command uses the parallel scheduling policy and leaves write-verify off for all
logical volumes. It does not enable mirror-write consistency for page space. It does enable mirror-write
consistency for all other logical volumes. Place logical volumes that are written to frequently close to the
outer edge of the disk to minimize seek distance between the logical volume and the mirror-write
consistency cache.
The caveat to mirroring rootvg not significantly affecting performance is that, if paging space is mirrored,
the slowdown is directly related to paging rate. So systems that are configured to support very high paging
rates, with paging spaces on rootvg, might not want to implement rootvg mirroring.
In summary, mirrored rootvg might be worth considering if your workload does not have high sustained
paging rates.
Reorganizing Logical Volumes
If you find that a volume was sufficiently fragmented to require reorganization, you can use the reorgvg
command (or smitty reorgvg) to reorganize a logical volume and make it adhere to the stated policies.
This command will reorganize the placement of physical partitions within the volume group according to
the logical volume characteristics. If logical volume names are specified with the command, highest priority
is given to the first logical volume in the list. To use this command, the volume group must be varied on
and have free partitions. The relocatable flag of each logical volume must be set to yes for the
reorganization to take place, otherwise the logical volume is ignored.
By knowing the usage pattern of logical volumes, you can make better decisions governing the policies to
set for each volume. Guidelines are:
v Allocate hot LVs to different PVs.
v Spread hot LV across multiple PVs.
v Place hottest LVs in center of PVs, except for LVs that have Mirror Write Consistency Check turned on.
v Place coldest LVs on Edges of PVs (except when accessed sequentially).
196
Performance Management Guide
v
v
v
v
Make LV contiguous.
Define LV to maximum size that you will need.
Place frequently used logical volumes close together.
Place sequential files on the edge.
Recommendations for Best Performance
Whenever logical volumes are configured for better performance, the availability might be impacted.
Decide whether performance or availability is more critical to your environment.
Use these guidelines when configuring for highest performance with the SMIT command:
v If the system does mostly reads, then mirroring with scheduling policy set to parallel can provide for
better performance since the read I/Os will be directed to the copies that are least busy. If doing writes,
then mirroring will cause a performance penalty because there will be multiple copies to write as well as
the Mirror Write Consistency record to update. You may also want to set the allocation policy to Strict to
have each copy on a separate physical volume.
v Set the write verify policy to No and, if the number of copies is greater than one, set the Mirror Write
Consistency to Off.
v In general, the most frequently accessed logical volumes should be in the center in order to minimize
seek distances; however, there are some exceptions:
– Disks hold more data per track on the edges of the disk. Logical volumes being accessed in
sequential manner could be placed on the edge for better performance.
– Another exception is for logical volumes that have Mirror Write Consistency Check (MWCC) turned
on. Because the MWCC sector is on the edge of the disk, performance may be improved if the
mirrored logical volume is also on the edge.
v Logical volumes that will be accessed frequently or concurrently should be placed close together on the
disk. Locality of reference is more important than placing them in the center.
v Put moderately used logical volumes in the middle, and put seldom-used logical volumes on the edge.
v By setting the Inter-Physical Volume Allocation Policy to maximum, you also ensure that the reads and
writes are shared among PVs.
Recommendations for Highest Availability
To configure the system for highest availability (with the SMIT command), follow these guidelines:
v Use three LP copies (mirroring twice)
v Set write verify to Yes
v Set the inter policy to Minimum (mirroring copies = # of PVs)
v Set the scheduling policy to Sequential
v Set the allocation policy to Strict (no mirroring on the same PV)
v Include at least three physical volumes in a volume group
v Mirror the copies on physical volumes attached to separate buses, adapters, and power supplies
Having at least three physical volumes allows a quorum to be maintained in the event one physical volume
becomes unavailable. Using separate busses, adapters, and power allows the use of copies not attached
to the failing device.
Reorganizing File Systems
You can reduce file system fragmentation as follows:
v Copying the files to a backup media
v Recreating the file system using the mkfs fsname command or deleting the contents of the file system
v Reloading the files into the new file system
Chapter 9. File System, Logical Volume, and Disk I/O Performance
197
This procedure loads the file sequentially and reduces fragmentation.
Reorganizing a File System
For example, a system has a separate logical volume and file system hd11 (mount point: /home/op).
Because we decide that file system hd11 needs to be reorganized, we perform the following steps:
1. Back up the file system by file name. If you back up the file system by i-node instead of by name, the
restore command puts the files back in their original places, which would not solve the problem. Run
the following commands:
# cd /home/op
# find . -print | backup -ivf/tmp/op.backup
This command creates a backup file (in a different file system), containing all of the files in the file
system that is to be reorganized. If disk space on the system is limited, you can use tape to back up
the file system.
2. Run the following commands:
# cd /
# unmount /home/op
If any processes are using /home/op or any of its subdirectories, you must terminate those processes
before the unmount command can complete successfully.
3. Re-create the file system on the /home/op logical volume, as follows:
# mkfs /dev/hd11
You are prompted for confirmation before the old file system is destroyed. The name of the file system
does not change.
4. To restore the original situation (except that /home/op is empty), run the following:
# mount /dev/hd11 /home/op
# cd /home/op
5. Restore the data, as follows:
# restore -xvf/tmp/op.backup >/dev/null
Standard output is redirected to /dev/null to avoid displaying the name of each of the files that were
restored, which is time-consuming.
6. Review the large file inspected earlier (see Assessing File Placement with the fileplace Command), as
follows:
# fileplace -piv big1
We see that it is now (nearly) contiguous:
File: big1 Size: 3554273 bytes Vol: /dev/hd11
Blk Size: 4096 Frag Size: 4096 Nfrags: 868
Compress: no
Inode: 8290 Mode: -rwxr-xr-x Owner: hoetzel Group: system
INDIRECT BLOCK: 60307
Physical Addresses (mirror copy 1)
Logical Fragment
----------------------------------
----------------
0060299-0060306
0060308-0061167
hdisk1
hdisk1
8 frags
860 frags
32768 Bytes,
3522560 Bytes,
0.9%
99.1%
868 frags over space of 869 frags: space efficiency = 99.9%
2 fragments out of 868 possible: sequentiality = 99.9%
198
Performance Management Guide
0008555-0008562
0008564-0009423
The -i option that we added to the fileplace command indicates that the one-block gap between the first
eight blocks of the file and the remainder contains the indirect block, which is required to supplement the
i-node information when the length of the file exceeds eight blocks.
Some file systems or logical volumes should not be reorganized because the data is either transitory (for
example, /tmp) or not in a file system format (log). The root file system is normally not very volatile and
seldom needs reorganizing. It can only be done in install/maintenance mode. The same is true for /usr
because many of these files are required for normal system operation.
Defragmenting a File System
If a file system has been created with a fragment size smaller than 4 KB, it becomes necessary after a
period of time to query the amount of scattered unusable fragments. If many small fragments are
scattered, it is difficult to find available contiguous space.
To recover these small, scattered spaces, use either the smitty dejfs or smitty dejfs2 command or the
defragfs command. Some free space must be available for the defragmentation procedure to be used.
The file system must be mounted for read-write.
Reorganizing File System Log and Log Logical Volumes
The Journaled File System (JFS) and the Enhanced Journaled File System (JFS2) use a database
journaling technique to maintain a consistent file system structure. This involves duplicating transactions
that are made to file system metadata to the circular file system log. File system metadata includes the
superblock, i-nodes, indirect data pointers, and directories.
When pages in memory are actually written to disk by a sync() or fsync() call, commit records are written
to the log to indicate that the data is now on disk. Log transactions occur in the following situations:
v When a file is created or deleted.
v When a write() call occurs for a file opened with O_SYNC and the write causes a new disk block
allocation.
v When the fsync() or sync() subroutines are called.
v When a write causes an indirect or double-indirect block to be allocated.
File system logs enable rapid and clean recovery of file systems if a system goes down. If an application
is doing synchronous I/O or is creating and removing many files in a short amount of time, there might be
a lot of I/O going to the log logical volume. If both the log logical volume and the file system logical volume
are on the same physical disk, this could cause an I/O bottleneck. The recommendation would be to
migrate the log device to another physical disk (this is especially useful for NFS servers).
Fast-write cached devices can provide for much better performance for log logical volumes (file system log
or database logs).
AIX 4.3.2 provides a mount option called nointegrity for JFS file systems which bypasses the use of a
JFS log for the file system mounted with this option. This can provide better performance as long as the
administrator knows that the fsck command might have to be run on the file system if the system goes
down without a clean shutdown.
Use the filemon command to record information about I/Os to the file system log. If you notice that a file
system and its log device are both heavily utilized, it might be better to put each one on a separate
physical disk (assuming that there is more than one disk in that volume group).
You can have multiple log devices in a volume group. However, a log for a file system must be in the
same volume group as that of the file system. A log logical volume or file system logical volume can be
moved to another disk using the migratepv command, even while the system is running and in use.
Chapter 9. File System, Logical Volume, and Disk I/O Performance
199
Creating Log Logical Volumes
Placing the log logical volume on a physical volume different from your most active file system logical
volume will increase parallel resource usage. You can use a separate log for each file system.
When you create your logical volumes, the performance of drives differs. Try to create a logical volume for
a hot file system on a fast drive (possibly one with fast write cache), as follows:
1. Create new file system log logical volume, as follows:
# mklv -t jfslog -y LVname VGname 1 PVname
or
# mklv -t jfs2log -y LVname VGname 1 PVname
or
# smitty mklv
2. Format the log as follows:
# /usr/sbin/logform -V vfstype /dev/LVname
3. Modify /etc/filesystems and the logical volume control block (LVCB) as follows:
# chfs -a log=/dev/LVname /filesystemname
4. Unmount and then mount file system.
Another way to create the log on a separate volume is to:
v Initially define the volume group with a single physical volume.
v Define a logical volume within the new volume group (this causes the allocation of the volume group
JFS log to be on the first physical volume).
v Add the remaining physical volumes to the volume group.
v Define the high-utilization file systems (logical volumes) on the newly added physical volumes.
Tuning with vmtune
The vmtune command can be used to modify the VMM parameters that control the behavior of the
memory-management subsystem. Some options are available to alter the defaults for LVM and file
systems; the options dealing with disk I/O are discussed in the following sections.
To determine whether the vmtune command is installed and available, run the following command:
# lslpp -lI bos.adt.samples
The executable program for the vmtune command is found in the /usr/samples/kernel directory. The
vmtune command can only be executed by the root user. Changes made by this tool remain in place until
the next reboot of the system. If a permanent change is needed, place an appropriate entry in the
/etc/inittab file. For example:
vmtune:2:wait:/usr/samples/kernel/vmtune -P 50
Note: The vmtune command is in the samples directory because it is VMM-implementation dependent.
The vmtune code that accompanies each release of the operating system is tailored specifically to
the VMM in that release. Running the vmtune command from one release on a system with a
different VMM release might result in an operating system failure. It is also possible that the
functions of the vmtune command may change from release to release. Be sure to review the
appropriate tuning information before using the vmtune command to change system parameters.
Sequential Read-Ahead
The VMM sequential read-ahead feature, described in Sequential-Access Read Ahead can enhance the
performance of programs that access large files sequentially.
200
Performance Management Guide
The following illustrates a typical read-ahead situation.
Figure 22. Sequential Read-Ahead Example. This illustration shows a row of blocks similating a segmented track of file
page numbers. These block segments are numbered 0, 1 through 3, 4 through 7, 8 through 15 and 16 through 23.
The steps of a sequential read-ahead are found in the text immediately following the illustration.
In this example, minpgahead is 2 and maxpgahead is 8 (the defaults). The program is processing the file
sequentially. Only the data references that have significance to the read-ahead mechanism are shown,
designated by A through F. The sequence of steps is:
A
The first access to the file causes the first page (page 0) of the file to be read. At this point, the
VMM makes no assumptions about random or sequential access.
B
When the program accesses the first byte of the next page (page 1), with no intervening accesses
to other pages of the file, the VMM concludes that the program is accessing sequentially. It
schedules minpgahead (2) additional pages (pages 2 and 3) to be read. Thus access B causes a
total of 3 pages to be read.
C
When the program accesses the first byte of the first page that has been read ahead (page 2), the
VMM doubles the page-ahead value to 4 and schedules pages 4 through 7 to be read.
D
When the program accesses the first byte of the first page that has been read ahead (page 4), the
VMM doubles the page-ahead value to 8 and schedules pages 8 through 15 to be read.
E
When the program accesses the first byte of the first page that has been read ahead (page 8), the
VMM determines that the page-ahead value is equal to maxpgahead and schedules pages 16
through 23 to be read.
F
The VMM continues reading maxpgahead pages when the program accesses the first byte of the
previous group of read-ahead pages until the file ends.
If the program were to deviate from the sequential-access pattern and access a page of the file out of
order, sequential read-ahead would be terminated. It would be resumed with minpgahead pages if the
VMM detected that the program resumed sequential access.
The minpgahead and maxpgahead values can be changed by using options -r and -R in the vmtune
command. If you are contemplating changing these values, keep in mind:
v The values should be from the set: 0, 1, 2, 4, 8, 16, and so on. The use of other values may have
adverse performance or functional effects.
– Values should be powers of 2 because of the doubling algorithm of the VMM.
– Values of maxpgahead greater than 16 (reads ahead of more than 64 KB) exceed the capabilities of
some disk device drivers. In such a case, the read size stays at 64 KB.
– Higher values of maxpgahead can be used in systems where the sequential performance of striped
logical volumes is of paramount importance.
v A minpgahead value of 0 effectively defeats the mechanism. This can adversely affect performance.
However, it can be useful in some cases where I/O is random, but the size of the I/Os cause the VMM’s
read-ahead algorithm to take effect. Another case where turning off page-ahead is useful is the case of
NFS reads on files that are locked. On these types of files, read-ahead pages are typically flushed by
NFS so that reading ahead is not helpful. NFS and the VMM have been changed, starting with AIX
4.3.3, to automatically turn off VMM read-ahead if it is operating on a locked file.
Chapter 9. File System, Logical Volume, and Disk I/O Performance
201
v The maxpgahead values of 8 or 16 yield the maximum possible sequential I/O performance for
non-striped file systems.
v The buildup of the read-ahead value from minpgahead to maxpgahead is quick enough that for most
file sizes there is no advantage to increasing minpgahead.
v The Sequential Read-Ahead can be tuned separately for JFS and Enhanced JFS. JFS Page
Read-Ahead can be tuned with minpgahead and maxpgahead whereas j2_minPageReadAhead and
j2_maxPageReadAhead are used for Enhanced JFS.
VMM Write-Behind
Write-behind involves asynchronously writing modified pages in memory to disk after reaching a threshold
rather than waiting for the syncd daemon to flush the pages to disk. This is done to limit the number of
dirty pages in memory, reduce system overhead, and minimize disk fragmentation. There are two types of
write-behind: sequential and random.
Sequential Write-Behind
By default, a file is partitioned into 16 K partitions or 4 pages. Each of these partitions is called a cluster. If
all 4 pages of this cluster are dirty, then as soon as a page in the next partition is modified, the 4 dirty
pages of this cluster are scheduled to go to disk. Without this feature, pages would remain in memory until
the syncd daemon runs, which could cause I/O bottlenecks and fragmentation of the file.
The number of clusters that the VMM uses as a threshold is tunable. The default is one cluster. You can
delay write-behind by increasing the numclust parameter using the vmtune -c command.
For enhanced JFS, the vmtune -j command is used to specify the number of pages to be scheduled at
one time, rather than the number of clusters. The default number of pages for vmtune -j is 8.
Random Write-Behind
There may be applications that perform a lot of random I/O, that is, the I/O pattern does not meet the
requirements of the write-behind algorithm and thus all the pages stay resident in memory until the syncd
daemon runs. If the application has modified many pages in memory, this could cause a very large
number of pages to be written to disk when the syncd daemon issues a sync() call.
The write-behind feature provides a mechanism such that when the number of dirty pages in memory for a
given file exceeds a defined threshold, these pages are then scheduled to be written to disk.
The administrator can tune the threshold by using the -W option of the vmtune command. The parameter
to tune is maxrandwrt; the default value is 0 indicating that random write-behind is disabled. Increasing
this value to 128 indicates that once 128 memory-resident pages of a file are dirty, any subsequent dirty
pages are scheduled to be written to the disk. The first set of pages will be flushed after a sync() call.
For enhanced JFS vmtune options j2_nRandomCluster (-z flag) and j2_maxRandomWrite (-J flag) are
used to tune random write-behind. Both options have a default of 0. The j2_maxRandomWrite option has
the same function for enhanced JFS as maxrandwrt does for JFS. That is, it specifies a limit for the
number of dirty pages per file that can remain in memory. The j2_nRandomCluster option specifies the
number of clusters apart two consecutive writes must be in order to be considered random.
Tuning File Syncs
JFS file I/Os that are not sequential will accumulate in memory until certain conditions are met:
v The free list shrinks to minfree, and page replacement has to occur.
v The syncd daemon flushes pages at regularly scheduled intervals.
v The sync command is issued.
v Random-write-behind flushes the dirty pages after random-write-behind threshold is reached.
202
Performance Management Guide
If too many pages accumulate before one of these conditions occur, then when pages do get flushed by
the syncd daemon, the i-node lock is obtained and held until all dirty pages have been written to disk.
During this time, threads trying to access that file will get blocked because the i-node lock is not available.
Remember that the syncd daemon currently flushes all dirty pages of a file, but one file at a time. On
systems with large amount of memory and large numbers of pages getting modified, high peaks of I/Os
can occur when the syncd daemon flushes the pages.
A tunable option called sync_release_ilock has been added in AIX 4.3.2. The vmtune command with the
-s option (value of 1 means release the i-node lock while flushing the modified pages, 0 means old
behavior) allows the i-node lock to be released while dirty pages of that file are being flushed. This can
result in better response time when accessing this file during a sync() call.
This blocking effect can also be minimized by increasing the frequency of syncs in the syncd daemon.
Change /sbin/rc.boot where it invokes the syncd daemon. Then reboot the system for it to take effect.
For the current system, kill the syncd daemon and restart it with the new seconds value.
A third way to tune this behavior is by turning on random write-behind using the vmtune command (see
VMM Write-Behind).
Miscellaneous I/O Tuning Parameters
The following vmtune parameters can be useful in tuning disk I/O:
numfsbufs
If there are many simultaneous or large I/Os to a filesystem or if there are large sequential I/Os to a file
system, it is possible that the I/Os might bottleneck at the file system level while waiting for bufstructs. The
number of bufstructs per file system (known as numfsbufs) can be increased using the vmtune -b
command. The value takes effect only when a file system is mounted; so if you change the value, you
must first unmount and mount the file system again. The default value for numfsbufs is currently 93
bufstructs per file system.
j2_nBufferPerPagerDevice
In Enhanced JFS, the number of bufstructs is specified with the j2_nBufferPerPagerDevice parameter.
The default number of bufstructs for an Enhanced JFS filesystem is currently 512. The number of
bufstructs per Enhanced JFS filesystem (j2_nBufferPerPagerDevice) can be increased using the vmtune
-Z command. The value takes effect only when a file system is mounted.
lvm_bufcnt
If an application is issuing very large raw I/Os rather than writing through the file system, the same type of
bottleneck as for file systems could occur at the LVM layer. Very large I/Os combined with very fast I/O
devices would be required to cause the bottleneck to be at the LVM layer. But if it does happen, a
parameter called lvm_bufcnt can be increased by the vmtune -u command to provide for a larger number
of ″uphysio″ buffers. The value takes effect immediately. The current default value is 9 ″uphysio″ buffers.
Because the LVM currently splits I/Os into 128 K each, and because the default value of lvm_bufcnt is 9,
the 9*128 K can be written at one time. If you are doing I/Os that are larger than 9*128 K, then increasing
lvm_bufcnt may be advantageous.
hd_pbuf_cnt
The hd_pbuf_cnt parameter (-B) controls the number of pbufs available to the LVM device driver. Pbufs
are pinned memory buffers used to hold I/O requests that are pending at the LVM layer.
In AIX Version 4, coalescing of sequential I/Os is done so that a single pbuf is used for each sequential
I/O request regardless of the number of pages in that I/O. It is difficult to encounter this type of bottleneck.
With random I/O, the I/Os tend to get flushed sporadically with the exception of the case when the syncd
daemon runs.
Chapter 9. File System, Logical Volume, and Disk I/O Performance
203
The best way to determine if a pbuf bottleneck is occurring is to examine a LVM variable called
hd_pendkblked. The following script can provide the value of this variable:
#!/bin/ksh
# requires root authority to run
# determines number of times LVM had to wait on pbufs since system boot
addr=`echo "knlist hd_pendqblked" | /usr/sbin/crash 2>/dev/null |tail -1| cut -f2 -d:`
value=`echo "od $addr 1 D" | /usr/sbin/crash 2>/dev/null | tail -1| cut -f2 -d:`
echo "Number of waits on LVM pbufs are: $value"
exit 0
Starting with AIX 4.3.3, the command vmtune -a also displays the hd_pendqblked value (see fsbufwaitcnt
and psbufwaitcnt).
Note: Do not set the hd_pbuf_cnt value too high, because the value cannot be lowered without a system
reboot.
pd_npages
Specifies the number of pages that should be deleted in one chunk from RAM when a file is deleted.
Changing this value may only be beneficial to real-time applications that delete files. By reducing the value
of pd_npages, a real-time application can get better response time because few number of pages will be
deleted before a process/thread is dispatched. The default value is the largest possible file size divided by
the page size (currently 4096); if the largest possible file size is 2 GB, then pd_npages is 524288 by
default. It can be changed with option -N.
v_pinshm
When you set the v_pinshm parameter to 1 (-S 1), it causes pages in shared memory segments to be
pinned by VMM, if the application, which does the shmget(), specifies SHM_PIN as part of the flags. The
default value is 0. This option is only available with operating system 4.3.3 and later.
Applications can choose to have a tunable which specifies whether the application should use the
SHM_PIN flag (for example, the lock_sga parameter in Oracle 8.1.5 and later). Avoid pinning too much
memory, because in that case no page replacement can occur. Pinning is useful because it saves
overhead in async I/O from these shared memory segments (the async I/O kernel extension are not
required to pin the buffers).
fsbufwaitcnt and psbufwaitcnt
Two counters were added in AIX 5.1 which are incremented whenever a bufstruct was not available and
the VMM puts a thread on the VMM wait list. Use the crash command or new options (fsbufwaitcnt and
psbufwaitcnt), for the vmtune -a command to examine the values of these counters. The following is an
example of the output:
# vmtune -a
hd_pendqblked = 305
psbufwaitcnt = 0
fsbufwaitcnt = 337
xpagerbufwaitcnt
One new counter was added in AIX 5.1, that is incremented whenever a bufstruct on an Enhanced JFS
filesystem is not available. Use the vmtune -a command to examine the value of this counter
(xpagerbufwaitcnt). The following is an example of the output:
# vmtune -a
xpagerbufwaitcnt = 815
Using Disk-I/O Pacing
Disk-I/O pacing is intended to prevent programs that generate very large amounts of output from
saturating the system’s I/O facilities and causing the response times of less-demanding programs to
deteriorate. Disk-I/O pacing enforces per-segment (which effectively means per-file) high- and low-water
marks on the sum of all pending I/Os. When a process tries to write to a file that already has high-water
204
Performance Management Guide
mark pending writes, the process is put to sleep until enough I/Os have completed to make the number of
pending writes less than or equal to the low-water mark. The logic of I/O-request handling does not
change. The output from high-volume processes is slowed down somewhat.
The high- and low-water marks are set with the SMIT command by selecting System Environments ->
Change / Show Characteristics of Operating System(smitty chgsys) and then entering the number of
pages for the high- and low-water marks. The default value for the high- and low-water marks is 0, which
disables pacing. Changes to the maxpout and minpout values take effect immediately and remain in
place until they are explicitly changed.
Example
The effect of pacing on performance can be demonstrated with an experiment that consists of starting a vi
editor session on a new file while another process is copying a 64 MB file with the cp command. The file
is copied from hdisk1 to hdisk0 and the vi executable program is located on hdisk0. For the vi session to
start, it must page itself in as well as perform a few other I/Os, which it does randomly one page at a time.
This takes about 50 physical I/Os, which can be completed in 0.71 seconds on a slow machine when
there is no contention for the disk. With the high-water mark set to the default of 0, the logical writes from
the cp command run ahead of the physical writes, and a large queue builds up.
Each I/O started by the vi session must wait its turn in the queue before the next I/O can be issued, and
thus the vi command is not able to complete its needed I/O until after the cp command finishes. The
following table shows the elapsed seconds for cp execution and vi initialization with different pacing
parameters.
High-Water Mark
Low-Water Mark
cp (sec)
vi (sec)
0
0
50
vi not done
0
0
50
vi finished after cp
9
6
77
2.7
17
8
64
3.4
17
12
58
3.6
33
16
55
4.9
33
24
52
9.0
It is important to notice that the cp duration is always longer when pacing is set. Pacing sacrifices some
throughput on I/O-intensive programs to improve the response time of other programs. The challenge for a
system administrator is to choose settings that result in a throughput/response-time trade-off that is
consistent with the organization’s priorities.
The high- and low-water marks were chosen by trial and error, based on our knowledge of the I/O path.
Choosing them is not straightforward because of the combination of write-behind and asynchronous writes.
High-water marks of (4 * n) + 1, where n is a positive integer, work particularly well, because of the
following interaction:
v The write-behind feature sends the previous four pages to disk when a logical write occurs to the first
byte of the fifth page.
v If the pacing high-water mark were a multiple of 4 (for example, 8), a process would hit the high-water
mark when it requested a write that extended into the 9th page. It would then be put to sleep, before
the write-behind algorithm had an opportunity to detect that the fourth dirty page is complete and the
four pages were ready to be written.
v The process would then sleep with four full pages of output until its outstanding writes fell below the
pacing low-water mark.
Chapter 9. File System, Logical Volume, and Disk I/O Performance
205
v If, on the other hand, the high-water mark had been set to 9, write-behind would get to schedule the
four pages for output before the process was suspended.
One limitation of pacing is that it does not offer as much control when a process writes buffers larger than
4 KB. When a write is sent to the VMM and the high-water mark has not been met, the VMM performs
start I/Os on all pages in the buffer, even if that results in exceeding the high-water mark. Pacing works
well on the cp command because the cp command writes 4 KB at a time; but if the cp command wrote
larger buffers, the times shown in the previous table for starting the vi session would increase.
Disk-I/O pacing is a tuning parameter that can improve interactive response time in some situations where
foreground or background programs that write large volumes of data are interfering with foreground
requests. If not used properly, however, it can reduce throughput excessively. The settings in the previous
table are a good place to start, but some experimenting will be needed to find the best settings for your
workload. For the newer systems, view these numbers as the minimum starting point.
Programs whose presence in a workload may make imposition of disk-I/O pacing necessary include:
v Programs that generate large amounts of output algorithmically, and thus are not constrained by the
time required to read input. Some such programs may need pacing on comparatively fast processors
and not need it on comparatively slow processors.
v Programs that write large, possibly somewhat modified, files that have been read in their entirety shortly
before writing begins (by a previous command, for example).
v Filters, such as the tar command, that read a file and write it out again with little processing. The need
for pacing can be exacerbated if the input is being read from a faster disk drive than the output is being
written to.
Tuning Logical Volume Striping
Striping is a technique for spreading the data in a logical volume across several disk drives in such a way
that the I/O capacity of the disk drives can be used in parallel to access data on the logical volume. The
primary objective of striping is very high-performance reading and writing of large sequential files, but there
are also benefits for random access. The following illustration gives a simple example.
206
Performance Management Guide
Figure 23. Striped Logical Volume /dev/lvs0. This illustration shows three physical volumes or drives. Each physical
drive is partitioned into two logical volumes. The first partition, or logical volume, of drive 1 contains stripe unit 1 and 4.
Partition 1 of drive 2 contains stripe unit 2 and 5 and partition 1 of drive 3 contains stripe unit 3 and 6. The second
partition of drive one contains stripe unit n and n+3. The second partition of drive 2 contains stripe unit n+1 and n+4.
Partition 2 of drive 3 contains stripe unit n+2 and n+5.
In an ordinary logical volume, the data addresses correspond to the sequence of blocks in the underlying
physical partitions. In a striped logical volume, the data addresses follow the sequence of stripe units. A
complete stripe consists of one stripe unit on each of the physical devices that contains part of the striped
logical volume. The LVM determines which physical blocks on which physical drives correspond to a block
being read or written. If more than one drive is involved, the necessary I/O operations are scheduled for all
drives simultaneously.
As an example, a hypothetical lvs0 has a stripe-unit size of 64 KB, consists of six 2 MB partitions, and
contains a journaled file system (JFS). If an application is reading a large sequential file and read-ahead
has reached a steady state, each read will result in two or three I/Os being scheduled to each of the disk
drives to read a total of eight pages (assuming that the file is on consecutive blocks in the logical volume).
The read operations are performed in the order determined by the disk device driver. The requested data
is assembled from the various pieces of input and returned to the application.
Although each disk device will have a different initial latency, depending on where its accessor was at the
beginning of the operation, after the process reaches a steady state, all three disks should be reading at
close to their maximum speed.
Designing a Striped Logical Volume
When a striped logical volume is defined, you specify:
drives At least two physical drives are required. The drives used should have little other activity when the
performance-critical sequential I/O is taking place.
Some combinations of disk adapter and disk drive will require dividing the workload of a striped
logical volume between two or more adapters.
stripe unit size
Although this can be any power of 2 from 4 KB through 128 KB, take sequential read-ahead into
Chapter 9. File System, Logical Volume, and Disk I/O Performance
207
account, because that will be the mechanism that issues most of the reads. The objective is to
have each read-ahead operation result in at least one I/O, ideally an equal number, to each disk
drive (see previous figure).
size
The number of physical partitions allocated to the logical volume must be an integral multiple of
the number of disk drives used.
attributes
Beginning with AIX 4.3.3, striped logical volumes can be mirrored; that is, copies can now be set
to more than 1.
Tuning for Striped Logical Volume I/O
Sequential and random disk I/Os benefit from disk striping. The following techniques have yielded the
highest levels of sequential I/O throughput:
v Spread the logical volume across as many physical volumes as possible.
v Use as many adapters as possible for the physical volumes.
v Create a separate volume group for striped logical volumes.
v Set a stripe-unit size of 64 KB.
v Set minpgahead to 2 (vmtune -r). See Sequential Read-Ahead.
v Set maxpgahead to 16 times the number of disk drives (vmtune -R). This causes page-ahead to be
done in units of the stripe-unit size (64 KB) times the number of disk drives, resulting in the reading of
one stripe unit from each disk drive for each read-ahead operation.
v Request I/Os for 64 KB times the number of disk drives. This is equal to the maxpgahead value.
v Modify maxfree (vmtune -F) to accommodate the change in maxpgahead (maxfree = minfree +
maxpgahead). See Choosing minfree and maxfree Settings.
v Use 64-byte aligned I/O buffers. If the logical volume will occupy physical drives that are connected to
two or more disk adapters, the I/O buffers used should be allocated on 64-byte boundaries. This avoids
having the LVM serialize the I/Os to the different disks. The following code would yield a 64-byte-aligned
buffer pointer:
char *buffer;
buffer = malloc(MAXBLKSIZE+64);
buffer = ((int)buffer + 64) & ~0x3f;
If the striped logical volumes are on raw logical volumes and writes larger than 1.125 MB are being done
to these striped raw logical volumes, increasing the lvm_bufcnt parameter with the command vmtune -u
might increase throughput of the write activity. See Miscellaneous I/O Tuning Parameters.
The example above is for a JFS striped logical volume. The same techniques apply to enhanced JFS,
except the vmtune parameters used will be the enhanced JFS equivalents.
Also, it is not a good idea to mix striped and non-striped logical volumes in the same physical volume. All
physical volumes should be the same size within a set of striped logical volumes.
Mirrored Striped Logical Volume Performance Implications
AIX 4.3.3 allows striping and mirroring together on the same logical volume. This feature provides a
convenient mechanism for high-performance redundant storage. Measurements indicate that read and file
system write performance of striping and mirroring is approximately equal to the unmirrored case,
assuming you have twice as many disks.
File system writes benefit from caching in the file system, which allows considerable overlap of writes to
disk with the program initiating the writes. The raw write performance suffers. Because it is synchronous,
both writes must complete before control is returned to the initiating program. Performing larger writes
increases raw write throughput. Also, Mirror Write Consistency (MWC) affects the performance of this
case.
208
Performance Management Guide
In summary, striping and mirroring allow redundant storage for very high-performance access.
Tuning Asynchronous Disk I/O
If an application does a synchronous I/O operation, it must wait for the I/O to complete. In contrast,
asynchronous I/O operations run in the background and do not block user applications. This improves
performance, because I/O operations and applications processing can run simultaneously. Many
applications, such as databases and file servers, take advantage of the ability to overlap processing and
I/O.
Applications can use the aio_read(), aio_write(), or lio_listio() subroutines (or their 64-bit counterparts) to
perform asynchronous disk I/O. Control returns to the application from the subroutine as soon as the
request has been queued. The application can then continue processing while the disk operation is being
performed.
To manage asynchronous I/O, each asynchronous I/O request has a corresponding control block in the
application’s address space. This control block contains the control and status information for the request.
It can be used again when the I/O operation is completed.
After issuing an asynchronous I/O request, the user application can determine when and how the I/O
operation is completed. This information is provided in any of three ways:
v The application can poll the status of the I/O operation.
v The system can asynchronously notify the application when the I/O operation is done.
v The application can block until the I/O operation is complete.
In AIX Version 4, async I/O on JFS file systems is handled by kprocs. Async I/O on raw logical volume
partitions is handled directly by the kernel. Starting with AIX 4.3.2 (and with a PTF for 4.3.1), Virtual
Shared Disk (VSD) devices do not use kprocs.
Each I/O is handled by a single kproc, and typically the kproc cannot process any more requests from the
queue until that I/O has completed. The default minimum number of servers configured when async I/O is
enabled is 1. This is the minservers attribute. There is also a maximum number of async I/O servers that
can get created which is controlled by the maxservers attribute; this has a default value of 10. The
number of servers limits the number of asynchronous disk I/O operations that can be in progress in the
system simultaneously. The number of servers can be set with the SMIT command (smitty->Devices>Asynchronous I/O->Change/Show Characteristics of Asynchronous I/O->{MINIMUM | MAXIMUM}
number of servers or smitty aio) or with the chdev command.
In systems that seldom run applications that use asynchronous I/O, the defaults are usually adequate.
If the number of async I/O requests is high, then the recommendation is to increase maxservers to
approximately the number of simultaneous I/Os there might be. In most cases, it is better to leave the
minservers parameter at the default value because the AIO kernel extension will generate additional
servers if needed.
Note: AIO actions performed against a raw Logical Volume do not use kproc server processes. The
setting of maxservers and minservers have no effect in this case.
By looking at the CPU utilization of the AIO servers, if the utilization is evenly divided among all of them,
that means that they’re all being used; you may want to try increasing them in this case. To see the AIO
servers by name, run the pstat -a command. Run the ps -k command to see the AIO servers as the name
kproc.
Chapter 9. File System, Logical Volume, and Disk I/O Performance
209
For environments in which the performance of asynchronous disk I/O is critical and the volume of requests
is high, but you do not have an approximate number of simultaneous I/Os, it is recommended that
maxservers be set to at least 10*(number of disks accessed asynchronously).
This could be achieved for a system with three asynchronously accessed disks as follows:
# chdev -l aio0 -a maxservers=’30’
In addition, you can set the maximum number of asynchronous I/O REQUESTS outstanding, and the
server PRIORITY. If you have a system with a high volume of asynchronous I/O applications, it might be
appropriate to increase the REQUESTS number and lower the PRIORITY number.
Tuning Direct I/O
When you are processing normal I/O to files, the I/O goes from the application buffer to the VMM and from
there back to the application buffer. The contents of the buffer are cached in RAM through the VMM’s use
of real memory as a file buffer cache. If the file cache hit rate is high, then this type of cached I/O is very
effective in improving overall I/O performance. But applications that have poor cache hit rates or
applications that do very large I/Os may not get much benefit from the use of normal cached I/O. In
operating system version 4.3, direct I/O was introduced as an alternative I/O method for files.
Direct I/O is only supported for program working storage (local persistent files). The main benefit of direct
I/O is to reduce CPU utilization for file reads and writes by eliminating the copy from the VMM file cache to
the user buffer. If the cache hit rate is low, then most read requests have to go to the disk. Writes are
faster with normal cached I/O in most cases. But if a file is opened with O_SYNC or O_DSYNC (see
Using sync/fsync Calls), then the writes have to go to disk. In these cases, direct I/O can benefit
applications because the data copy is eliminated.
Another benefit is that direct I/O allows applications to avoid diluting the effectiveness of caching of other
files. When a file is read or written, that file competes for space in memory which could cause other file
data to get pushed out of memory. If an application developer knows that certain files have poor
cache-utilization characteristics, then only those files could be opened with O_DIRECT.
For direct I/O to work efficiently, the I/O request should be appropriate for the type of file system being
used. The finfo() and ffinfo() subroutines can be used to query the offset, length, and address alignment
requirements for fixed block size file systems, fragmented file systems, and bigfile file systems (direct I/O
is not supported on compressed file systems). The information queried are contained in the structure
diocapbuf as described in /usr/include/sys/finfo.h.
To avoid consistency issues, if there are multiple calls to open a file and one or more of the calls did not
specify O_DIRECT and another open specified O_DIRECT, the file stays in the normal cached I/O mode.
Similarly, if the file is mapped into memory through the shmat() or mmap() system calls, it stays in normal
cached mode. If the last conflicting, non-direct access is eliminated, then the file system will move the file
into direct I/O mode (either by using the close(), munmap(), or shmdt() subroutines). Changing from
normal mode to direct I/O mode can be expensive because all modified pages in memory will have to be
flushed to disk at that point.
Performance of Direct I/O Reads
Even though the use of direct I/O can reduce CPU usage, it typically results in longer elapsed times,
especially for small I/O requests, because the requests would not be cached in memory.
Direct I/O reads cause synchronous reads from the disk, whereas with normal cached policy, the reads
may be satisfied from the cache. This can result in poor performance if the data was likely to be in
memory under the normal caching policy. Direct I/O also bypasses the VMM read-ahead algorithm
because the I/Os do not go through the VMM. The read-ahead algorithm is very useful for sequential
210
Performance Management Guide
access to files because the VMM can initiate disk requests and have the pages already be resident in
memory before the application has requested the pages. Applications can compensate for the loss of this
read-ahead by using one of the following methods:
v Issuing larger read requests (minimum of 128 K)
v Issuing asynchronous direct I/O read-ahead by the use of multiple threads
v Using the asynchronous I/O facilities such as aio_read() or lio_listio()
Performance of Direct I/O Writes
Direct I/O writes bypass the VMM and go directly to the disk, so that there can be a significant
performance penalty; in normal cached I/O, the writes can go to memory and then be flushed to disk later
by a sync or write-behind operation. Because direct I/O writes do not get copied into memory, when a sync
operation is performed, it will not have to flush these pages to disk, thus reducing the amount of work the
syncd daemon has to perform.
Performance Example
In the following example, performance is measured on an RS/6000 E30, 233 MHz, with AIX 4.3.1 and JFS
file systems (your performance will vary depending on system configuration). KBPS is the throughput in
kilobytes per second, and %CPU is CPU usage in percent.
# of 2.2 GB SSA Disks
# of PCI SSA Adapters
1
1
2
1
4
1
6
1
8
1
KBPS
7108
14170
18725
18519
17892
%CPU
23.9
56.1
92.1
97.0
98.3
KBPS
7098
14150
22035
27588
30062
%CPU
4.4
9.1
22.0
39.2
54.4
KBPS
7258
14499
28504
30946
32165
%CPU
1.6
3.2
10.0
20.9
24.5
Sequential read throughput, using JFS I/O
Sequential read throughput, using JFS direct I/O
Sequential read throughput, using raw I/O
Summary
Direct I/O requires substantially fewer CPU cycles than regular I/O. I/O-intensive applications that do not
get much benefit from the caching provided by regular I/O can enhance performance by using direct I/O.
The benefits of direct I/O will grow in the future as increases in CPU speeds continue to outpace increases
in memory speeds.
What types of programs are good candidates for direct I/O? Programs that are typically CPU-limited and
perform lots of disk I/O. “Technical” codes that have large sequential I/Os are good candidates.
Applications that do numerous small I/Os will typically see less performance benefit, because direct I/O
cannot do read ahead or write behind. Applications that have benefited from striping are also good
candidates.
Using Raw Disk I/O
Some applications, such as databases, do not require a file system because they perform functions such
as logging, keeping track of data, and caching. Performance of these applications is usually better when
using raw I/O rather than using file I/O because it avoids the additonal work of memory copies, logging,
and inode locks.
Chapter 9. File System, Logical Volume, and Disk I/O Performance
211
When using raw I/O, applications should use the /dev/rlv* character special files. The /dev/lv* block
special files should not be used, as they break large I/Os into multiple 4K I/Os. The /dev/rhdisk* and
/dev/hdisk* raw disk interfaces should not be used because they degrade performance and can also
cause data consistency problems.
Using sync/fsync Calls
If a file is opened with O_SYNC or O_DSYNC, then each write will cause the data for that write to be
flushed to disk before the write returns. If the write causes a new disk allocation (the file is being extended
instead of overwriting an existing page), then that write will also cause a corresponding JFS log write.
Forced synchronization of the contents of real memory and disk takes place in several ways:
v An application program makes an fsync() call for a specified file. This causes all of the pages that
contain modified data for that file to be written to disk. The writing is complete when the fsync() call
returns to the program.
v An application program makes a sync() call. This causes all of the file pages in memory that contain
modified data to be scheduled for writing to disk. The writing is not necessarily complete when the
sync() call returns to the program.
v A user can enter the sync command, which in turn issues a sync() call. Again, some of the writes may
not be complete when the user is prompted for input (or the next command in a shell script is
processed).
v The /usr/sbin/syncd daemon issues a sync() call at regular intervals, usually every 60 seconds. This
ensures that the system does not accumulate large amounts of data that exists only in volatile RAM.
A sync operation has several effects, aside from its small CPU consumption:
v It causes writes to be clumped, rather than spread out.
v It causes at least 28 KB of system data to be written, even if there has been no I/O activity since the
previous sync operation.
v It accelerates the writing of data to disk, defeating the write-behind algorithm. This effect is significant
mainly in programs that issue an fsync() call after every write.
v When sync() or fsync() calls occur, log records are written to the JFS log device to indicate that the
modified data has been committed to disk.
Setting SCSI-Adapter and Disk-Device Queue Limits
The operating system has the ability to enforce limits on the number of I/O requests that can be
outstanding from the SCSI adapter to a given SCSI bus or disk drive. These limits are intended to exploit
the hardware’s ability to handle multiple requests while ensuring that the seek-optimization algorithms in
the device drivers are able to operate effectively.
For non-IBM devices, it is sometimes appropriate to modify the default queue-limit values that have been
chosen to handle the worst possible case. The following sections describe situations in which the defaults
should be changed and the recommended new values.
Non-IBM Disk Drive
For IBM disk drives, the default number of requests that can be outstanding at any given time is 3 (8 for
SSA). This value is based on complex performance considerations, and no direct interface is provided for
changing it. The default hardware queue depth for non-IBM disk drives is 1. If a specific non-IBM disk
drive does have the ability to buffer multiple requests, the system’s description of that device should be
changed accordingly.
As an example, the default characteristics of a non-IBM disk drive are displayed with the lsattr command,
as follows:
212
Performance Management Guide
# lsattr -D -c disk -s scsi -t osdisk
pvid
none Physical volume identifier
False
clr_q
no
Device CLEARS its Queue on error
q_err
yes Use QERR bit
q_type
none Queuing TYPE
queue_depth
1
Queue DEPTH
reassign_to
120 REASSIGN time out value
rw_timeout
30
READ/WRITE time out value
start_timeout 60
START unit time out value
You can use SMIT (the fast path is smitty chgdsk) or the chdev command to change these parameters.
For example, if your system contained a non-IBM SCSI disk drive hdisk5, the following command enables
queuing for that device and sets its queue depth to 3:
# chdev -l hdisk5 -a q_type=simple -a queue_depth=3
Non-IBM Disk Array
A disk array appears to the operating system as a single, rather large, disk drive. A non-IBM disk array,
like a non-IBM disk drive, is of class disk, subclass SCSI, type osdisk (which stands for ″Other SCSI Disk
Drive″). Because a disk array actually contains a number of physical disk drives, each of which can handle
multiple requests, the queue depth for the disk array device has to be set to a value high enough to allow
efficient use of all of the physical devices. For example, if hdisk7 were an eight-disk non-IBM disk array,
an appropriate change would be as follows:
# chdev -l hdisk7 -a q_type=simple -a queue_depth=24
If the disk array is attached through a SCSI-2 Fast/Wide SCSI adapter bus, it may also be necessary to
change the outstanding-request limit for that bus.
Expanding the Configuration
Unfortunately, every performance-tuning effort ultimately does reach a point of diminishing returns. The
question then becomes, ″What hardware do I need, how much of it, and how do I make the best use of
it?″ That question is especially tricky with disk-limited workloads because of the large number of variables.
Changes that might improve the performance of a disk-limited workload include:
v Adding disk drives and spreading the existing data across them. This divides the I/O load among more
accessors.
v Acquiring faster disk drives to supplement or replace existing ones for high-usage data.
v Adding one or more disk adapters to attach the current and new disk drives.
v Adding RAM to the system and increasing the VMM’s minperm and maxperm parameters to improve
the in-memory caching of high-usage data.
For guidance more closely focused on your configuration and workload, you can use a
measurement-driven simulator, such as BEST/1.
Using RAID
Redundant Array of Independent Disks (RAID) is a term used to describe the technique of improving data
availability through the use of arrays of disks and various data-striping methodologies. Disk arrays are
groups of disk drives that work together to achieve higher data-transfer and I/O rates than those provided
by single large drives. An array is a set of multiple disk drives plus a specialized controller (an array
controller) that keeps track of how data is distributed across the drives. Data for a particular file is written
in segments to the different drives in the array rather than being written to a single drive.
Arrays can also provide data redundancy so that no data is lost if a single drive (physical disk) in the array
should fail. Depending on the RAID level, data is either mirrored or striped.
Chapter 9. File System, Logical Volume, and Disk I/O Performance
213
Subarrays are contained within an array subsystem. Depending on how you configure it, an array
subsystem can contain one or more sub-arrays, also referred to as Logical Units (LUN). Each LUN has its
own characteristics (RAID level, logical block size and logical unit size, for example). From the operating
system, each subarray is seen as a single hdisk with its own unique name.
RAID algorithms can be implemented as part of the operating system’s file system software, or as part of
a disk device driver (common for RAID 0 and RAID 1). These algorithms can be performed by a locally
embedded processor on a hardware RAID adapter. Hardware RAID adapters generally provide better
performance than software RAID because embedded processors offload the main system processor by
performing the complex algorithms, sometimes employing specialized circuitry for data transfer and
manipulation.
LVM
AIX LVM supports the following RAID options:
RAID 0
Striping
RAID 1
Mirroring
RAID 10 or 0+1
Mirroring and striping
RAID Levels and Their Performance Implications
Each of the RAID levels supported by disk arrays uses a different method of writing data and hence
provides different benefits.
RAID 0 - For Performance
RAID 0 is also known as data striping. It is well-suited for program libraries requiring rapid loading of large
tables, or more generally, applications requiring fast access to read-only data, or fast writing. RAID 0 is
only designed to increase performance; there is no redundancy, so any disk failures require reloading from
backups. Select RAID Level 0 for applications that would benefit from the increased performance
capabilities of this RAID Level. Never use this level for critical applications that require high availability.
RAID 1 - For Availability/Good Read Response Time
RAID 1 is also known as disk mirroring. It is most suited to applications that require high data availability,
good read response times, and where cost is a secondary issue. The response time for writes can be
somewhat slower than for a single disk, depending on the write policy; the writes can either be executed in
parallel for speed or serially for safety. Select RAID Level 1 for applications with a high percentage of read
operations and where the cost is not the major concern.
RAID 2 - Rarely Used
RAID 2 is rarely used. It implements the same process as RAID 3, but can utilize multiple disk drives for
parity, while RAID 3 can use only one.
RAID 3 - For CAD/CAM, Sequential Access to Large Files
RAID 3 and RAID 2 are parallel process array mechanisms, where all drives in the array operate in
unison. Similar to data striping, information to be written to disk is split into chunks (a fixed amount of
data), and each chunk is written out to the same physical position on separate disks (in parallel). More
advanced versions of RAID 2 and 3 synchronize the disk spindles so that the reads and writes can truly
occur simultaneously (minimizing rotational latency buildups between disks). This architecture requires
parity information to be written for each stripe of data; the difference between RAID 2 and RAID 3 is that
RAID 2 can utilize multiple disk drives for parity, while RAID 3 can use only one. The LVM does not
support Raid 3; therefore, a RAID 3 array must be used as a raw device from the host system.
Performance is very good for large amounts of data but poor for small requests because every drive is
always involved, and there can be no overlapped or independent operation. It is well-suited for large data
214
Performance Management Guide
objects such as CAD/CAM or image files, or applications requiring sequential access to large data files.
Select RAID 3 for applications that process large blocks of data. RAID 3 provides redundancy without the
high overhead incurred by mirroring in RAID 1.
RAID 4 - Less Used (Parity Volume Bottleneck)
RAID 4 addresses some of the disadvantages of RAID 3 by using larger chunks of data and striping the
data across all of the drives except the one reserved for parity. Write requests require a
read/modify/update cycle that creates a bottleneck at the single parity drive. Therefore, RAID 4 is not used
as often as RAID 5, which implements the same process, but without the parity volume bottleneck.
RAID 5 - High Availability and Fewer Writes Than Reads
RAID 5, as has been mentioned, is very similar to RAID 4. The difference is that the parity information is
distributed across the same disks used for the data, thereby eliminating the bottleneck. Parity data is never
stored on the same drive as the chunks that it protects. This means that concurrent read and write
operations can now be performed, and there are performance increases due to the availability of an extra
disk (the disk previously used for parity). There are other enhancements possible to further increase data
transfer rates, such as caching simultaneous reads from the disks and transferring that information while
reading the next blocks. This can generate data transfer rates at up to the adapter speed.
RAID 5 is best used in environments requiring high availability and fewer writes than reads. Select RAID
level 5 for applications that manipulate small amounts of data, such as transaction processing applications.
RAID 6 - Seldom Used
RAID 6 is similar to RAID 5, but with additional parity information written that permits data recovery if two
disk drives fail. Extra parity disk drives are required, and write performance is slower than a similar
implementation of RAID 5.
RAID 7 - A Definition of 3rd Parties
The RAID 7 architecture gives data and parity the same privileges. The level 7 implementation allows each
individual drive to access data as fast as possible. This is achieved by three features:
v Independent control and data paths for each I/O device/interface.
v Each device/interface is connected to a high-speed data bus that has a central cache capable of
supporting multiple host I/O paths.
v A real time, process-oriented operating system is embedded into the disk drive array architecture. The
embedded operating system ″frees″ the drives by allowing each drive head to move independently of
the other disk drives. Also, the RAID 7 embedded operating system is enabled to handle a
heterogeneous mix of disk drive types and sizes.
RAID 10 - RAID-0+1
RAID-0+1, also known in the industry as RAID 10, implements block interleave data striping and mirroring.
RAID 10 is not formally recognized by the RAID Advisory Board (RAB), but, it is an industry standard term.
In RAID 10, data is striped across multiple disk drives, and then those drives are mirrored to another set of
drives.
The performance of RAID 10 is approximately the same as RAID 0 for sequential I/Os. RAID 10 provides
an enhanced feature for disk mirroring that stripes data and copies the data across all the drives of the
array. The first stripe is the data stripe; the second stripe is the mirror (copy) of the first data stripe, but it
is shifted over one drive. Because the data is mirrored, the capacity of the logical drive is 50 percent of the
physical capacity of the hard disk drives in the array.
Summary of RAID Levels
The advantages and disadvantages of the different RAID levels are summarized in the following table:
RAID Level
Availability
Capacity
Performance
Cost
0
none
100 percent
high
low
Chapter 9. File System, Logical Volume, and Disk I/O Performance
215
RAID Level
Availability
Capacity
Performance
Cost
1
mirroring
50 percent
medium/high
high
2/3
parity
varies between 50 100%
medium
medium
4/5/6/7
parity
varies between 50 100%
medium
medium
10
mirroring
50 percent
high
high
RAID Performance Summary
The most common RAID implementations are: 0, 1, 3 and 5. Levels 2, 4 and 6 have problems with
performance and are functionally not better than the other ones. In most cases, RAID 5 is used instead of
RAID 3 because of the bottleneck when using only one disk for parity.
RAID 0 and RAID 1 can be implemented with software support only. RAID 3, 5 and 7 require both
hardware and software support (special RAID adapters or RAID array controllers).
For further information, see Configuring and Implementing the IBM Fibre Channel RAID Storage Server.
Using SSA
Serial Storage Architecture (SSA) is a high performance, serial interconnect technology used to connect
disk devices and host adapters. SSA subsystems are built up of loops of adapters and disks. SSA is a
serial link designed especially for low-cost, high-performance connection to disk drives. It is a two-signal
connection (transmit and receive) providing full duplex communication. It uses a self-clocking code, which
means that the data clock is recovered from the data signal itself rather than being transmitted as a
separate signal.
Guidelines for Improving SSA Performance
Examine these guidelines in terms of your situation:
v Limit the number of disks per adapter so that the adapter is not flooded. With high throughputs using
large block sizes, five to six disks can flood the adapter.
v Mirror across different adapters.
v The performance will be affected by the location of the logical volume on the disk. A contiguous,
unfragmented partition in a logical volume will enhance performance.
v You can turn off mirror write consistency cache for logical volume mirroring, but doing so removes the
guarantee of consistency of data in case of a crash. In this case, you would have to recopy all logical
volumes to make sure they are consistent. However, the removal does provide a 20 percent plus
enhancement in performance.
v For mirrored environments, make sure you are using the parallel scheduling policy.
v If any of the logical volume exists on more than one disk, stagger the partitions over the disks. This is
automatically accomplished when the logical volume is created with the inter policy set to Maximum.
v Balance the I/Os across the loop; do not place all the workload on one half of the loop.
v In a multi-initiator environment, place the disks adjacent to the adapter that is using them.
For further information, see Monitoring and Managing IBM SSA Disk Subsystems and A Practical Guide to
Serial Storage Architecture for AIX.
Additional information about IBM storage solutions can be found in IBM Storage Solutions for e-business
and Introduction to Storage Area Network.
216
Performance Management Guide
Using Fast Write Cache
Fast write cache (FWC) is an optional nonvolatile cache that provides redundancy with the standard
adapter cache. The FWC tracks writes that have not been committed to disk.
FWC can significantly improve the response time for write operations. However, care must be taken not to
flood the cache with write requests faster than the rate at which the cache can destage its data. It can also
adversely affect the maximum I/O rate because additional processing is required in the adapter card to
determine if the data that is being transferred is in the cache.
Fast write cache typically provides significant advantages in specialized workloads, for example, copying a
database onto a new set of disks. If the fast write cache is spread over multiple adapters, this can multiply
the benefit.
The FWC can also reduce JFS log bottlenecks because of three properties of the JFS log, as follows:
1. The JFS log is write-intensive. The FWC does not cache unmodified data.
2. The writes are small and frequent. Because the cache capacity is not enormous, it works best for
high-rate small I/Os that are gathered together in the adapter into larger physical I/Os. Larger I/Os tend
to have better performance because less disk rotations are normally needed to write the data.
3. Logs are not typically very large relative to the cache size, so the log does not tend to ″wash″ the
cache frequently. Therefore, the log loses the benefit of rewriting over existing cached data. Although
other array controllers with write caches have proven effective with logs, this article only discusses log
performance with the FWC.
When single disk bandwidth becomes the limiting performance factor, one solution is to stripe several
RAID 5 devices into a logical volume in conjunction with the FWC option of the SSA adapter. When the
adapter is configured for RAID 5, writes equal to or larger than the stripe size bypass the cache. That is
why 128 KB writes to a 2+p array with FWC are slower than 127 KB writes, and are equal to 128 KB
writes to 2+p without the FWC. This is intended to keep bulk sequential I/O from ″washing″ the cache. The
stripe size is 64 KB times the number of data disks in the RAID 5.
Chapter 9. File System, Logical Volume, and Disk I/O Performance
217
218
Performance Management Guide
Chapter 10. Monitoring and Tuning Communications I/O Use
This chapter discusses several different communications protocols and ways to monitor and tune them. It
contains the following major sections:
v UDP and TCP/IP Performance Overview
v Analyzing Network Performance
v Tuning TCP and UDP Performance
v Tuning mbuf Pool Performance
v Tuning Asynchronous Connections for High-Speed Transfers
v Tuning Name Resolution
v Improving telnetd/rlogind Performance
v Tuning the SP Network
UDP and TCP/IP Performance Overview
To understand the performance characteristics of UDP (user datagram protocol) and TCP/IP, you must first
understand some of the underlying architecture. The following figure illustrates the structure that will be
discussed in this chapter.
© Copyright IBM Corp. 1997, 2002
219
Figure 24. UDP and TCP/IP Data Flow. The figure shows the path of data from an application in one system to
another application in a remote system. The steps of the data flow are described in the text immediately following the
illustration.
Note: (ARP) address resolution protocol, see “Send Flow” on page 228.
The figure shows the path of data from an application in one system to another application in a remote
system. The processing at each of the layers is discussed in this chapter, but key points are as follows:
220
Performance Management Guide
v The application’s write request causes the data to be copied from the application’s working segment to
the socket send buffer.
v The socket layer or subsystem calls UDP or TCP.
v The operating system has variable size clusters, so an optimum size is used when:
– UDP copies and computes the checksum of the data into a socket buffer.
– TCP copies the data to socket buffer.
v If the size of the data is larger than the maximum transfer unit (MTU) of the LAN, then:
– TCP breaks the output into segments that comply with the MTU limit.
– UDP leaves the breaking up of the output to the IP layer.
v If necessary, IP fragments the output into pieces that comply with the MTU, so that no outgoing packet
exceeds the MTU limit.
v The packets are put on the device output queue and transmitted by the LAN adapter to the receiving
system. If the output queue for the device overflows, the packet is discarded.
v Arriving packets are placed on the device driver’s receive queue, and pass through the Interface layer
to IP.
v If IP in the receiving system determines that IP in the sending system had fragmented a block of data, it
coalesces the fragments into their original form and passes the data to TCP or UDP.
– TCP reassembles the original segments and places the input in the socket receive buffer.
– UDP passes the input on to the socket receive buffer. If the input socket (udp_recvspace) limit is
reached, the packet is discarded.
v When the application makes a read request, the appropriate data is copied from the socket receive
buffer in kernel memory into the buffer in the application’s buffer.
Communication Subsystem Memory (mbuf) Management
To avoid fragmentation of kernel memory and the overhead of numerous calls to the xmalloc() subroutine,
the various layers of the communication subsystem share common buffer pools. The mbuf management
facility controls different buffer sizes. The pools consist of pinned pieces of kernel virtual memory; this
means that they always reside in physical memory and are never paged out. The result is that the real
memory available for paging in application programs and data has been decreased by the amount that the
mbuf pools have been increased.
In addition to avoiding duplication, sharing the mbuf and cluster pools allows the various layers to pass
pointers to one another, reducing mbuf management calls and copying of data.
For additional details, see Tuning mbuf Pool Performance.
Socket Layer
Sockets provide the application program interface (API) to the communication subsystem. Several types of
sockets provide various levels of service by using different communication protocols. Sockets of type
SOCK_DGRAM use the UDP protocol. Sockets of type SOCK_STREAM use the TCP protocol.
The processes of opening, reading, and writing to sockets are similar to those for manipulating files.
The sizes of the buffers in system virtual memory (that is, the total number of bytes from the mbuf pools)
that are used by the input and output sides of each socket are limited by system-wide default values
(which can be overridden for a given socket by a call to the setsockopt() subroutine):
udp_sendspace and udp_recvspace
Buffer sizes for datagram sockets in bytes. The defaults are 9216 and 42080, respectively.
Chapter 10. Monitoring and Tuning Communications I/O Use
221
tcp_sendspace and tcp_recvspace
Buffer sizes for stream sockets in bytes. The defaults for both values are 16384. With AIX 4.3.3
and later, these two parameters can also be set using ISNO (see Interface-Specific Network
Options (ISNO)).
Use the following to display these values:
# no -a
A root user can set these values as follows:
# no -o udp_sendspace=NewValue
The NewValue parameter must be less than or equal to the sb_max parameter, which controls the
maximum amount of space that can be used by a socket’s send or receive buffer. The default value of the
sb_max parameter depends on the operating system version and amount of real memory. The sb_max
value is displayed with the command no -a and set with the no command, as follows:
# no -o sb_max=NewLimit
Note: Socket send or receive buffer sizes are limited to no more than sb_max bytes, because sb_max is
a ceiling on buffer space consumption. The two quantities are not measured in the same way,
however. The socket buffer size limits the amount of data that can be held in the socket buffers.
The sb_max value limits the number of bytes of mbufs that can be in the socket buffer at any given
time. In an Ethernet environment, for example, each 2048-byte mbuf cluster might hold just 1500
bytes of data. In that case, sb_max would have to be 1.37 times larger than the specified socket
buffer size to allow the buffer to reach its specified capacity. The guideline is to set sb_max to at
least twice the size of the largest socket buffer.
Send Flow
As an application writes to a socket, the socket layer calls the transport layer (either TCP or UDP), which
copies the data from user space into the socket send buffer in kernel space. Depending on the amount of
data being copied into the socket send buffer, the code puts the data into either mbufs or clusters.
Receive Flow
On the receive side, an application opens a socket and attempts to read data from it. If there is no data in
the socket receive buffer, the socket layer causes the application thread to go to the sleep state (blocking)
until data arrives. When data arrives, it is put on the receive socket buffer queue and the application
thread is made dispatchable. The data is then copied into the application’s buffer in user space, the mbuf
chain is freed, and control is returned to the application.
Socket Creation
In AIX 4.3.1 and later, the sockthresh value determines how much of the system’s network memory can
be used before socket creation is disallowed. The value of sockthresh is given as a percentage of
thewall. It has a default of 85 percent and can be set to any value from 1 to 100. However, sockthresh
cannot be set to a value lower than the amount of memory currently in use.
The sockthresh option is intended to prevent situations where many connections are opened until all the
network memory on the machine is used. This leaves no memory for other operations, and the machine
hangs and must be rebooted to recover. Use sockthresh to set the point at which new sockets should not
be allowed. Calls to the socket() and socketpair() subroutines will fail with an error of ENOBUFS, and
incoming connection requests will be silently discarded. This allows the remaining network memory to be
used by existing connections and prevents the machine from hanging.
The netstat -m statistic sockets not created because sockthresh was reached is incremented each time
a socket creation fails because the amount of network memory already in use is over the sockthresh limit.
Use the following to display the sockthresh value:
# no -o sockthresh
222
Performance Management Guide
A root user can set the value as follows:
# no -o sockthresh=NewValue
The default value can be set as follows:
# no -d sockthresh
Ephemeral Ports
When an application requests that the system assign the port (application is not requesting a specific port
number), this is called an ephemeral port. Prior to AIX 4.3.1, the ephemeral port range was from 1024 to
5000. Starting with AIX 4.3.1, the default starting ephemeral port number is 32768, and the default largest
ephemeral port number is 65535.
Using the no command, these values can be tuned with the tcp_ephemeral_low and
tcp_ephemeral_high parameters. The maximum range would be to set tcp_ephemeral_low to 1024 and
tcp_ephemeral_high to 65535. UDP ports have the same tunable parameters available through
udp_ephemeral_low and udp_ephemeral_high (defaults are identical).
Relative Level of Function in UDP and TCP
The following two sections contain descriptions of the function of UDP and TCP. To facilitate comparison of
UDP and TCP, both descriptions are divided into subsections on connection, error detection, error
recovery, flow control, data size, and MTU handling.
UDP Layer
UDP provides a low-cost protocol for applications that have the facilities to deal with communication
failures. UDP is most suitable for request-response applications. Because such an application has to
handle a failure to respond anyway, it is little additional effort to handle communication error as one of the
causes of failure to respond. For this reason, and because of its low overhead, subsystems such as NFS,
ONC RPC, DCE RPC, and DFS use UDP.
Features of the UDP layer are as follows:
Connection
None. UDP is essentially a stateless protocol. Each request received from the caller is handled
independently of those that precede or follow it. (If the connect() subroutine is called for a
datagram socket, the information about the destination is considered a hint to cache the resolved
address for future use. It does not actually bind the socket to that address or affect UDP on the
receiving system.)
Error detection
Checksum creation and verification. The sending UDP builds the checksum and the receiving UDP
checks it. If the check fails, the packet is dropped.
Error recovery
None. UDP does not acknowledge receipt of packets, nor does it detect their loss in transmission
or through buffer-pool overflow. Consequently, UDP never retransmits a packet. Recovery must be
performed by the application.
Flow control
None. When UDP is asked to send, it sends the packet to IP. When a packet arrives from IP, it is
placed in the socket-receive buffer. If either the device driver/adapter buffer queue or the
socket-receive buffer is full when the packet arrives, the packet is dropped without an error
indication. The application or subsystem that sent the packet must detect the failure by timeout or
sequence and retry the transmission. Various statistics show counts of discarded packets (see the
netstat -s and netstat -D commands in The netstat Command).
Data size
Must fit in one buffer. This means that the buffer pools on both sides of UDP must have buffer
sizes that are adequate for the applications’ requirements. The maximum size of a UDP packet is
Chapter 10. Monitoring and Tuning Communications I/O Use
223
64 KB. Of course, an application that builds large blocks can break them into multiple datagrams
itself (for example, DCE), but it is simpler to use TCP.
MTU handling
None. Dealing with data larger than the maximum transfer unit (MTU) size for the interface is left
to IP. If IP has to fragment the data to make it fit the MTU, loss of one of the fragments becomes
an error that the application or subsystem must deal with timeout and retransmit logic.
Send Flow: If udp_sendspace is large enough to hold the datagram, the application’s data is copied
into mbufs in kernel memory. If the datagram is larger than udp_sendspace, an error is returned to the
application.
The operating system chooses optimum size buffers from a power of 2 size buffer. For example, a write of
8704 bytes is copied into two clusters, a 8192-byte and a 512-byte cluster. UDP adds the UDP header (in
the same mbuf, if possible), checksums the data, and calls the IP ip_output() routine.
Receive Flow: UDP verifies the checksum and queues the data onto the proper socket. If the
udp_recvspace limit is exceeded, the packet is discarded. A count of these discards is reported by the
netstat -s command under udp: as socket buffer overflows. If the application is waiting for a receive or
read on the socket, it is put on the run queue. This causes the receive to copy the datagram into the
user’s address space and release the mbufs, and the receive is complete. Usually, the receiver responds
to the sender to acknowledge the receipt and also return a response message.
In AIX 4.1.1 and later, UDP checksums the data ″on the fly″ when it copies it into the kernel mbuf. When
receiving, this same optimization can be done, but the application must enable it with the
SO_CKSUMRECV option on a setsockopt() call. Applications that receive large UDP buffers should
program to use this option for better performance.
TCP Layer
TCP provides a reliable transmission protocol. TCP is most suitable for applications that, at least for
periods of time, are mostly output or mostly input. With TCP ensuring that packets reach their destination,
the application is freed from error detection and recovery responsibilities. Applications that use TCP
transport include ftp, rcp, and telnet. DCE can use TCP if it is configured to use a connection-oriented
protocol.
Features of the TCP layer are as follows:
Connection
Explicit. The instance of TCP that receives the connection request from an application (call it the
initiator, sender,or transmitter) establishes a session with its counterpart on the other system,
which you will call the listener, or receiver. All exchanges of data and control packets are within
the context of that session.
Error detection
Checksum creation and verification. The sending TCP builds the checksum and the receiving TCP
checks it. If checksum verification fails, the receiver does not acknowledge receipt of the packet.
Some PCI adapters now have TCP checksum offload. For example, the Gigabit Ethernet adapter
for transmits and receives, and the ATM 155 adapter for transmits. The default is set to on. The
transmit can be disabled with the ifconfig command and the checksum_offload parameter, while
the receive requires a chdev command to set cx_checksum=no.
Error recovery
Full. TCP detects checksum failures and loss of a packet or fragment through timeout. In error
situations TCP retransmits the data until it is received correctly (or notifies the application of an
unrecoverable error).
Flow control
Enforced. TCP uses a discipline called a sliding window to ensure delivery to the receiving
application. The sliding window concept is illustrated in the following figure. (The records shown in
224
Performance Management Guide
the figure are for clarity only. TCP processes data as a stream of bytes and does not keep track of
record boundaries, which are application-defined.)
Figure 25. TCP Sliding Window. This illustration depicts the TCP Sliding Window. A full description is in the text
immediately following the figure.
In this figure, the sending application is sleeping because it has attempted to write data that would
cause TCP to exceed the send socket buffer space (that is, tcp_sendspace). The sending TCP
still has the last part of rec5, all of rec6 and rec7, and the beginning of rec8. The receiving TCP
has not yet received the last part of rec7 or any of rec8. The receiving application got rec4 and the
beginning of rec5 when it last read the socket, and it is now processing that data. When the
receiving application next reads the socket, it will receive (assuming a large enough read), the rest
of rec5, rec6, and as much of rec7 and rec8 as has arrived by that time.
After the next read, the following occur:
v The receiving TCP will be able to acknowledge that data
v The sending TCP will be able to discard the data
v The pending write will complete
The sending application will wake up. To avoid excessive LAN traffic when the application is
reading in tiny amounts, TCP delays acknowledgment until the receiving application has read a
total amount of data that is at least half the receive window size or twice the maximum segment
size.
If there is no data to send back, the receiver will delay up to 200 ms and then send the ACK. The
delay time can be tuned by a new no parameter called fasttimeo. The default value is 200 ms,
and the range of values can be between 50 ms and 200 ms. Reducing this value may enhance
performance of request/response type of applications.
Note: When using TCP to exchange request/response messages, the application must use the
setsockopt() subroutine to turn on the TCP_NODELAY option. This causes TCP to send
Chapter 10. Monitoring and Tuning Communications I/O Use
225
the message immediately (within the constraints of the sliding window), even though it is
less than MTU-size. Otherwise, TCP would wait for up to 200 milliseconds for more data to
send before transmitting the message. Starting with AIX 4.3.3 the tcp_nodelay parameter
can be set with the ifconfig or chdev command to set TCP_NODELAY on TCP sockets
(see Interface-Specific Network Options (ISNO)).
In the course of establishing a session, the initiator and the listener converse to determine the
receive space for each end point. The size defines the size of the receive window. As data is
written to the socket, it is moved into the sender’s buffer. When the receiver indicates that it has
space available, the sender transmits enough data to fill that space (assuming that it contains that
much data). When the receiving application reads from the socket, the receiving socket returns as
much data as it has in its receive socket buffer. TCP then informs the sender that the data has
been successfully delivered by sending a packet to advance the receiver window. Only then does
the sending TCP discard the data from its own buffer, effectively moving the window to the right by
the amount of data delivered. If the window is full because the receiving application has fallen
behind, the sending thread will be blocked (or receive a specific errno) when it tries to write to the
socket.
The value of tcp_recvspace and tcp_sendspace are independent. The tcp_sendspace controls
the buffering in the kernel of the sender. The tcp_recvspace controls the receiver space and
translates into TCP’s receive window.
If the rfc1323 parameter is 1, the maximum TCP window size is 4 GB (instead of 64 KB).
Data size
Indefinite. TCP does not process records or blocks; it processes a stream of bytes. If a
send buffer is larger than the receiver can handle, it is segmented into MTU-size packets.
Because it handles shortages of buffer space under the covers, TCP does not guarantee
that the number and size of data receives will be the same as the number and size of
sends. It is the responsibility of the two sides of the application to identify record or block
boundaries, if any, within the stream of data.
MTU handling
Handled by segmentation in TCP. When the connection is established, the initiator and the
listener negotiate a maximum segment size (MSS) to be used. The MSS is typically
smaller than the MTU (see Tuning TCP Maximum Segment Size). If the output packet size
exceeds the MSS, TCP does the segmentation, thus making fragmentation in IP
unnecessary. The receiving TCP typically puts the segments on the socket receive queue
as they arrive. If the receiving TCP detects the loss of a segment, it withholds
acknowledgment and holds back the succeeding segments until the missing segment has
been received successfully.
The additional operations performed by TCP to ensure a reliable connection result in about 5 to 10 percent
higher processor cost than in UDP.
Send Flow: When the TCP layer receives a write request from the socket layer, it allocates a new mbuf
for its header information and copies the data in the socket-send buffer either into the TCP-header mbuf, if
there is room, or into a newly allocated mbuf chain. If the data being copied is in clusters, the data is not
actually copied into new clusters. Instead, a pointer field in the new mbuf header (this header is part of the
mbuf structure and is unrelated to the TCP header) is set to point to the clusters containing the data,
thereby avoiding the overhead of one or more 4 KB copies. TCP then checksums the data (unless it is
offloaded by certain PCI adapters), updates its various state variables, which are used for flow control and
other services, and finally calls the IP layer with the header mbuf now linked to the new mbuf chain.
Receive Flow: When the TCP input routine receives input data from IP, the following occur:
v It checksums the TCP header and data for corruption detection (unless it is offloaded by certain PCI
adapters)
226
Performance Management Guide
v
v
v
v
Determines which connection this data is for
Removes its header information
Links the mbuf chain onto the socket-receive buffer associated with this connection
Uses a socket service to wake up the application (if it is sleeping as described earlier)
IP Layer
The Internet Protocol provides a basic datagram service to the higher layers. If it is given a packet larger
than the MTU of the interface, it fragments the packet and sends the fragments to the receiving system,
which reassembles them into the original packet. If one of the fragments is lost in transmission, the
incomplete packet is ultimately discarded by the receiver. MTU path discovery can be enabled as
described in Tuning TCP Maximum Segment Size.
The length of time IP waits for a missing fragment is controlled by the ipfragttl parameter, which is set
and displayed with the no command.
Following are some default values and value ranges for different network types:
Network Type
Default (bytes)
Range (bytes)
X.25
576
60-2058
SLIP
1006
60-4096
Standard Ethernet
1500
60 - 1500
IEEE 802.3 Ethernet
1492
60 - 1492
Gigabit Ethernet
9000 (Jumbo Frames)
N/A
Token-Ring 4 Mbps
1492
60 - 4096
Token-Ring 16 Mbps
1492
60 - 17800
FDDI
4352
1 - 4352
SLA (socc)
61428
1 - 61428
ATM
9180
1 - 65527
HIPPI
65536
60 - 65536
SP Switch
65520
1 - 65520
Note: In general, you can increase the transmit and receive queues. This requires some memory, but
avoids some problems. See Adapter Transmit and Receive Queue Tuning.
Send Flow
When the IP output routine receives a packet from UDP or TCP, it identifies the interface to which the
mbuf chain should be sent, updates and checksums the IP part of the header, and passes the packet to
the interface (IF) layer.
IP determines the proper device driver and adapter to use, based on the network number. The driver
interface table defines the maximum MTU for this network. If the datagram is less than the MTU size, IP
adds the IP header in the existing mbuf, checksums the IP header, and calls the driver to send the frame.
If the driver send queue is full, an EAGAIN error is returned to IP, which returns it to UDP, which returns it
to the sending application. The sender should delay and try again.
If the datagram is larger than the MTU size (which only occurs in UDP), IP fragments the datagram into
MTU-size fragments, appends a IP header (in an mbuf) to each, and calls the driver once for each
fragment frame. If the driver’s send queue is full, an EAGAIN error is returned to IP. IP discards all
Chapter 10. Monitoring and Tuning Communications I/O Use
227
remaining unsent fragments associated with this datagram and returns EAGAIN to UDP. UDP returns
EAGAIN the sending application. Since IP and UDP do not queue messages, it is up to the application to
delay and try the send again.
Receive Flow
In AIX Version 4, in general, interfaces do not perform queuing and directly call the IP input queue routine
to process the packet; the loopback interface will still perform queuing. In the case of queuing, the demux
layer places incoming packets on this queue. If the queue is full, packets are dropped and never reach the
application. If packets are dropped at the IP layer, a statistic called ipintrq overflows in the output of the
netstat -s command is incremented. If this statistic increases in value, then use the no command to tune
the ipqmaxlen tunable.
In AIX Version 4, the demux layer (formerly called the IF layer) calls IP on the interrupt thread. IP checks
the IP header checksum to make sure the header was not corrupted and determines if the packet is for
this system. If so, and the frame is not a fragment, IP passes the mbuf chain to the TCP or UDP input
routine.
If the received frame is a fragment of a larger datagram (which only occurs in UDP), IP retains the frame.
When the other fragments arrive, they are merged into a logical datagram and given to UDP when the
datagram is complete. IP holds the fragments of an incomplete datagram until the ipfragttl time (as
specified by the no command) expires. The default ipfragttl time is 30 seconds (an ipfragttl value of 60).
If any fragments are lost due to problems such as network errors, lack of mbufs, or transmit queue
overruns, IP never receives them. When ipfragttl expires, IP discards the fragments it did receive. This is
reported as a result from the netstat -s command. Under ip:, see fragments dropped after timeout.
Demux Layer
The interface layer (IF) is used on output and is the same level as the demux layer (used for input) in AIX
Version 4. It places transmit requests on to a transmit queue, where the requests are then serviced by the
network interface device driver. The size of the transmit queue is tunable, as described in Adapter
Transmit and Receive Queue Tuning.
Send Flow
When the demux layer receives a packet from IP, it attaches the link-layer header information to the
beginning of the packet, checks the format of the mbufs to make sure they conform to the device driver’s
input specifications, and then calls the device driver write routine.
The address resolution protocol (ARP) is also handled in this layer. ARP translates a 32-bit Internet
Protocol (IP) address into a 48-bit hardware address.
Receive Flow
In AIX Version 4, when the demux layer receives a packet from the device driver, it calls IP on the interrupt
thread to perform IP input processing.
If the dog threads are enabled (see Enabling Thread Usage on LAN Adapters (dog threads)), the incoming
packet will be queued to the thread and the thread will handle calling IP, TCP, and the socket code.
LAN Adapters and Device Drivers
The operating system environment supports many different kinds of LAN adapters. You can choose from a
wide variety of network interfaces. As the following table shows, as the speed of these networks varies, so
does the performance.
Name
Speed
Ethernet (en)
10 Mbit/sec - Gigabits/sec
IEEE 802.3 (et)
10 Mbit/sec - Gigabits/sec
Token-Ring (tr)
4 or 16 Mbit/sec
228
Performance Management Guide
Name
Speed
X.25 protocol (xt)
64 Kb/sec
Serial Line Internet Protocol, SLIP (sl)
64 Kb/sec
loopback (lo)
N/A
FDDI (fi)
100 Mbit/sec
SOCC (so)
220 Mbit/sec
ATM (at)
100s Mbit/sec (many Gb/sec)
Refer to the PCI Adapter Placement Reference and RS/6000 Systems Handbook for slot placement
guidelines and limitations that may exist on the number of adapters that can be supported for connectivity
and the number that can be supported for maximum performance.
Several PCI machines have secondary PCI buses bridged onto a primary PCI bus. Some medium- to
high-speed adapters perform slower on these secondary bus slots and some adapters are not
recommended to be used in these slots. Machines with some secondary PCI slots include E30, F40, and
SP 332 MHz SMP-wide nodes.
The adapters differ, not only in the communications protocol and transmission medium they support, but
also in their interface to the I/O bus and the processor. Similarly, the device drivers vary in the technique
used to convey the data between memory and the adapter. The following description of send and receive
flow applies to most adapters and device drivers, but details vary.
Send Flow
At the device-driver layer, the mbuf chain containing the packet is enqueued on the transmit queue. The
maximum total number of output buffers that can be queued is controlled by the system parameter
xmt_que_size. In some cases, the data is copied into driver-owned DMA buffers. The adapter is then
signaled to start DMA operations.
At this point, control returns up the path to the TCP or UDP output routine, which continues sending as
long as it has data to send. When all data has been sent, control returns to the application, which then
runs asynchronously while the adapter transmits data. Device driver dependent, when the adapter has
completed transmission, it sends an interrupt to the system. When the interrupt is handled, the
device-interrupt routines are called to adjust the transmit queues and free the mbufs that held the
transmitted data.
Receive Flow
When frames are received by an adapter, they are transferred from the adapter into a driver-managed
receive queue. The receive queue can consist of mbufs or the device driver can manage a separate pool
of buffers for the device. In either case, the data is in an mbuf chain when it is passed from the device
driver to the demux layer.
Some drivers receive frames through Direct Memory Access (DMA) into a pinned area of memory and
then allocate mbufs and copy the data into them. Drivers/adapters that receive large-MTU frames may
have the frames accessed directly into cluster mbufs. The driver transfers the frame to the correct network
protocol (IP in this example) by calling a demultiplexing function that identifies the packet type and puts
the mbuf containing the buffer on the input queue for that network protocol. If no mbufs are available or if
the higher-level input queue is full, the incoming frames are discarded.
Chapter 10. Monitoring and Tuning Communications I/O Use
229
Analyzing Network Performance
When performance problems arise, your system might be totally innocent, while the real culprit is buildings
away. An easy way to tell if the network is affecting overall performance is to compare those operations
that involve the network with those that do not. If you are running a program that does a considerable
amount of remote reads and writes and it is running slowly, but everything else seems to be running as
usual, then it is probably a network problem. Some of the potential network bottlenecks can be caused by
the following:
v Client-network interface
v Network bandwidth
v Network topology
v Server network interface
v Server CPU load
v Server memory usage
v Server bandwidth
v Inefficient configuration
Several tools can measure network statistics and give a variety of information, but only part of this
information is related to performance tuning.
To enhance performance, you can use the no (network options) command and the nfso command for
tuning NFS options. You can also use the chdev and ifconfig commands to change system and network
parameters.
The ping Command
The ping command is useful for the following:
v Determining the status of the network and various foreign hosts.
v Tracking and isolating hardware and software problems.
v Testing, measuring, and managing networks.
Some ping command options relevant to performance tuning are as follows:
-c
Specifies the number of packets. This option is useful when you get an IP trace log. You can
capture a minimum of ping packets.
-s
Specifies the length of packets. You can use this option to check fragmentation and reassembly.
-f
Sends the packets at 10 ms intervals or immediately after each response. Only the root user can
use this option.
If you need to load your network or systems, the -f option is convenient. For example, if you suspect that
your problem is caused by a heavy load, load your environment intentionally to confirm your suspicion.
Open several aixterm windows and run the ping -f command in each window. Your Ethernet utilization
quickly gets to around 100 percent. The following is an example:
# date ; ping -c 1000 -f wave ; date
Fri Jul 23 11:52:39 CDT 1999
PING wave.austin.ibm.com: (9.53.153.120): 56 data bytes
.
----wave.austin.ibm.com PING Statistics---1000 packets transmitted, 1000 packets received, 0% packet loss
round-trip min/avg/max = 1/1/23 ms
Fri Jul 23 11:52:42 CDT 1999
Note: This command can be very hard on a network and should be used with caution. Flood-pinging can
only be performed by the root user.
230
Performance Management Guide
In this example, 1000 packets were sent for 3 seconds. Be aware that this command uses IP and Internet
Control Message Protocol (ICMP) protocol and therefore, no transport protocol (UDP/TCP) and application
activities are involved. The measured data, such as round-trip time, does not reflect the total performance
characteristics.
When you try to send a flood of packets to your destination, consider several points:
v Sending packets puts a load on your system.
v Use the netstat -i command to monitor the status of your network interface during the experiment. You
may find that the system is dropping packets during a send by looking at the Oerrs output.
v You should also monitor other resources, such as mbufs and send/receive queue. It can be difficult to
place a heavy load onto the destination system. Your system might be overloaded before the other
system is.
v Consider the relativity of the results. If you want to monitor or test just one destination system, do the
same experiment on some other systems for comparison, because your network or router might have a
problem.
The ftp Command
You can use the ftp command to send a very large file by using /dev/zero as input and /dev/null as
output. This allows you to transfer a large file without involving disks (which might be a bottleneck) and
without having to cache the entire file in memory.
Use the following ftp subcommands (change count to increase or decrease the number of blocks read by
the dd command):
> bin
> put "|dd if=/dev/zero bs=32k count=10000" /dev/null
Remember, if you change the TCP send or receive space parameters, then for the ftp command, you
must refresh the inetd daemon with the refresh -s inetd command.
Make sure that tcp_senspace and tcp_recvspace are at least 65535 for the Gigabit Ethernet ″jumbo
frames″ and for the ATM with MTU 9180 or larger to get good performance due to larger MTU size.
An example to set the parameters is as follows:
# no -o tcp_sendspace=65535
# no -o tcp_recvspace=65535
# refresh -s inetd
0513-095 The request for subsystem refresh was completed successfully.
The ftp subcommands are as follows:
ftp> bin
200 Type set to I.
ftp> put "|dd if=/dev/zero bs=32k count=10000" /dev/null
200 PORT command successful.
150 Opening data connection for /dev/null.
10000+0 records in
10000+0 records out
226 Transfer complete.
327680000 bytes sent in 8.932 seconds (3.583e+04 Kbytes/s)
local: |dd if=/dev/zero bs=32k count=10000 remote: /dev/null
ftp> quit
221 Goodbye.
Chapter 10. Monitoring and Tuning Communications I/O Use
231
The netstat Command
The netstat command is used to show network status. Traditionally, it is used more for problem
determination than for performance measurement. However, the netstat command can be used to
determine the amount of traffic on the network to ascertain whether performance problems are due to
network congestion.
The netstat command displays information regarding traffic on the configured network interfaces, such as
the following:
v The address of any protocol control blocks associated with the sockets and the state of all sockets
v The number of packets received, transmitted, and dropped in the communications subsystem
v Cumulative statistics per interface
v Routes and their status
Using the netstat Command
The netstat command displays the contents of various network-related data structures for active
connections. In this chapter, only the options and output fields that are relevant for network performance
determinations are discussed. For all other options and columns, see the AIX 5L Version 5.2 Commands
Reference.
netstat -i: Shows the state of all configured interfaces.
The following example shows the statistics for a workstation with an integrated Ethernet and a Token-Ring
adapter:
# netstat -i
Name Mtu
Network
Address
lo0
16896 <Link>
lo0
16896 127
localhost
tr0
1492 <Link>10.0.5a.4f.3f.61
tr0
1492 9.3.1
ah6000d
en0
1500 <Link>8.0.5a.d.a2.d5
en0
1500 1.2.3
1.2.3.4
Ipkts Ierrs
144834
0
144834
0
658339
0
658339
0
0
0
0
0
Opkts Oerrs Coll
144946
0
0
144946
0
0
247355
0
0
247355
0
0
112
0
0
112
0
0
The count values are summarized since system startup.
Name Interface name.
Mtu
Maximum transmission unit. The maximum size of packets in bytes that are transmitted using the
interface.
Ipkts
Total number of packets received.
Ierrs
Total number of input errors. For example, malformed packets, checksum errors, or insufficient
buffer space in the device driver.
Opkts Total number of packets transmitted.
Oerrs Total number of output errors. For example, a fault in the local host connection or adapter output
queue overrun.
Coll
Number of packet collisions detected.
Note: The netstat -i command does not support the collision count for Ethernet interfaces (see The
entstat Command for Ethernet statistics).
Following are some tuning guidelines:
v If the number of errors during input packets is greater than 1 percent of the total number of input
packets (from the command netstat -i); that is,
Ierrs > 0.01 x Ipkts
232
Performance Management Guide
Then run the netstat -m command to check for a lack of memory.
v If the number of errors during output packets is greater than 1 percent of the total number of output
packets (from the command netstat -i); that is,
Oerrs > 0.01 x Opkts
Then increase the send queue size (xmt_que_size) for that interface. The size of the xmt_que_size
could be checked with the following command:
# lsattr -El adapter
v If the collision rate is greater than 10 percent, that is,
Coll / Opkts > 0.1
Then there is a high network utilization, and a reorganization or partitioning may be necessary. Use the
netstat -v or entstat command to determine the collision rate.
netstat -i -Z: This function of the netstat command clears all the statistic counters for the netstat -i
command to zero.
netstat -I interface interval: Displays the statistics for the specified interface. It offers information similar
to the netstat -i command for the specified interface and reports it for a given time interval. For example:
# netstat -I en0 1
input
(en0)
output
input
(Total)
output
packets errs packets errs colls packets errs packets
errs
0
0
27
0
0
799655
0
390669
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
78
0
254
0
0
0
0
0
0
200
0
62
0
0
0
1
0
0
0
0
2
0
colls
0
0
0
0
0
0
The previous example shows the netstat -I command output for the ent0 interface. Two reports are
generated side by side, one for the specified interface and one for all available interfaces (Total). The
fields are similar to the ones in the netstat -i example, input packets = Ipkts, input errs = Ierrs and so
on.
netstat -m: Displays the statistics recorded by the mbuf memory-management routines. The most useful
statistics in the output of the netstat -m command are the counters that show the requests for mbufs
denied and non-zero values in the failed column. If the requests for mbufs denied is not displayed, then
this must be an SMP system running operating system version 4.3.2 or later; for performance reasons,
global statistics are turned off by default. To enable the global statistics, set the no parameter
extended_netstats to 1. This can be done by changing the /etc/rc.net file and rebooting the system.
The following example shows the first part of the netstat -m output with extended_netstats set to 1:
# netstat -m
29 mbufs in use:
16 mbuf cluster pages in use
71 Kbytes allocated to mbufs
0 requests for mbufs denied
0 calls to protocol drain routines
Kernel malloc statistics:
******* CPU 0 *******
By size
inuse
calls failed
32
419
544702
0
64
173
22424
0
128
121
37130
0
256
1201 118326233
0
512
330
671524
0
1024
74
929806
0
delayed
0
0
0
0
0
0
free
221
19
135
239
14
82
hiwat
800
400
200
480
50
125
freed
0
0
4
138
54
2
Chapter 10. Monitoring and Tuning Communications I/O Use
233
2048
4096
8192
16384
32768
384
516
9
1
1
By type
inuse
1820884
1158445
5634
2953
1
0
0
0
0
0
calls failed delayed
0
0
0
0
0
memuse
8
46
1
24
0
125
150
12
30
1023
memmax
5605
21
27
41
0
mapb
Streams mblk statistic failures:
0 high priority mblk failures
0 medium priority mblk failures
0 low priority mblk failures
If global statistics are not on and you want to determine the total number of requests for mbufs denied,
add up the values under the failed columns for each CPU. If the netstat -m command indicates that
requests for mbufs or clusters have failed or been denied, then you may want to increase the value of
thewall by using the no -o thewall=NewValue command. See Overview of the mbuf Management Facility
for additional details about the use of thewall and maxmbuf.
Beginning with AIX 4.3.3, a delayed column was added. If the requester of an mbuf specified the M_WAIT
flag, then if an mbuf was not available, the thread is put to sleep until an mbuf is freed and can be used
by this thread. The failed counter is not incremented in this case; instead, the delayed column will be
incremented. Prior to AIX 4.3.3, the failed counter was also not incremented, but there was no delayed
column.
Also, if the currently allocated amount of network memory is within 85 percent of thewall, you may want to
increase thewall. If the value of thewall is increased, use the vmstat command to monitor total memory
use to determine if the increase has had a negative impact on overall memory performance.
If buffers are not available when a request is received, the request is most likely lost (to see if the adapter
actually dropped a package, see Adapter Statistics). Keep in mind that if the requester of the mbuf
specified that it could wait for the mbuf if not available immediately, this puts the requestor to sleep but
does not count as a request being denied.
If the number of failed requests continues to increase, the system might have an mbuf leak. To help track
down the problem, the no command parameter net_malloc_police can be set to 1, and the trace hook
with ID 254 can be used with the trace command.
After an mbuf/cluster is allocated and pinned, it can be freed by the application. Instead of unpinning this
buffer and giving it back to the system, it is left on a free-list based on the size of this buffer. The next time
that a buffer is requested, it can be taken off this free-list to avoid the overhead of pinning. After the
number of buffers on the free list reaches the highwater mark, buffers smaller than 4096 will be coalesced
together into page-sized units so that they can be unpinned and given back to the system. When the
buffers are given back to the system, the freed column is incremented. If the freed value consistently
increases, the highwater mark is too low. In AIX 4.3.2 and later, the highwater mark is scaled according to
the amount of RAM on the system.
netstat -v: The netstat -v command displays the statistics for each Common Data Link Interface
(CDLI)-based device driver that is in operation. Interface-specific reports can be requested using the
tokstat, entstat, fddistat, or atmstat commands.
Every interface has its own specific information and some general information. The following example
shows the Token-Ring and Ethernet part of the netstat -v command; other interface parts are similar. With
a different adapter, the statistics will differ somewhat. The most important output fields are highlighted.
# netstat -v
------------------------------------------------------------ETHERNET STATISTICS (ent0) :
Device Type: IBM 10/100 Mbps Ethernet PCI Adapter (23100020)
234
Performance Management Guide
Hardware Address: 00:60:94:e9:29:18
Elapsed Time: 9 days 19 hours 5 minutes 51 seconds
Transmit Statistics:
-------------------Packets: 0
Bytes: 0
Interrupts: 0
Transmit Errors: 0
Packets Dropped: 0
Max Packets on S/W Transmit Queue: 0
S/W Transmit Queue Overflow: 0
Current S/W+H/W Transmit Queue Length: 0
Broadcast Packets: 0
Multicast Packets: 0
No Carrier Sense: 0
DMA Underrun: 0
Lost CTS Errors: 0
Max Collision Errors: 0
Late Collision Errors: 0
Deferred: 0
SQE Test: 0
Timeout Errors: 0
Single Collision Count: 0
Multiple Collision Count: 0
Current HW Transmit Queue Length: 0
Receive Statistics:
------------------Packets: 0
Bytes: 0
Interrupts: 0
Receive Errors: 0
Packets Dropped: 0
Bad Packets: 0
Broadcast Packets: 0
Multicast Packets: 0
CRC Errors: 0
DMA Overrun: 0
Alignment Errors: 0
No Resource Errors: 0
Receive Collision Errors: 0
Packet Too Short Errors: 0
Packet Too Long Errors: 0
Packets Discarded by Adapter: 0
Receiver Start Count: 0
General Statistics:
------------------No mbuf Errors: 0
Adapter Reset Count: 0
Driver Flags: Up Broadcast Running
Simplex 64BitSupport PrivateSegment
IBM 10/100 Mbps Ethernet PCI Adapter Specific Statistics:
-----------------------------------------------Chip Version: 25
RJ45 Port Link Status : down
Media Speed Selected: 10 Mbps Half Duplex
Media Speed Running: Unknown
Receive Pool Buffer Size: 384
Free Receive Pool Buffers: 128
No Receive Pool Buffer Errors: 0
Inter Packet Gap: 96
Adapter Restarts due to IOCTL commands: 0
Packets with Transmit collisions:
1 collisions: 0
6 collisions: 0
11 collisions:
2 collisions: 0
7 collisions: 0
12 collisions:
3 collisions: 0
8 collisions: 0
13 collisions:
4 collisions: 0
9 collisions: 0
14 collisions:
5 collisions: 0
10 collisions: 0
15 collisions:
Excessive deferral errors: 0x0
------------------------------------------------------------TOKEN-RING STATISTICS (tok0) :
Device Type: IBM PCI Tokenring Adapter (14103e00)
Hardware Address: 00:20:35:7a:12:8a
Elapsed Time: 29 days 18 hours 3 minutes 47 seconds
Transmit Statistics:
-------------------Packets: 1355364
Bytes: 791555422
Interrupts: 902315
Transmit Errors: 0
Packets Dropped: 0
0
0
0
0
0
Receive Statistics:
------------------Packets: 55782254
Bytes: 6679991641
Interrupts: 55782192
Receive Errors: 1
Packets Dropped: 0
Chapter 10. Monitoring and Tuning Communications I/O Use
235
Max Packets on S/W Transmit Queue: 182
S/W Transmit Queue Overflow: 42
Current S/W+H/W Transmit Queue Length: 0
Broadcast Packets: 18878
Multicast Packets: 0
Timeout Errors: 0
Current SW Transmit Queue Length: 0
Current HW Transmit Queue Length: 0
Bad Packets: 0
Broadcast Packets: 54615793
Multicast Packets: 569
Receive Congestion Errors: 0
General Statistics:
------------------No mbuf Errors: 0
Lobe Wire Faults: 0
Abort Errors: 12
AC Errors: 0
Burst Errors: 1
Frame Copy Errors: 0
Frequency Errors: 0
Hard Errors: 0
Internal Errors: 0
Line Errors: 0
Lost Frame Errors: 0
Only Station: 1
Token Errors: 0
Remove Received: 0
Ring Recovered: 17
Signal Loss Errors: 0
Soft Errors: 35
Transmit Beacon Errors: 0
Driver Flags: Up Broadcast Running
AlternateAddress 64BitSupport ReceiveFunctionalAddr
16 Mbps
IBM PCI Tokenring Adapter (14103e00) Specific Statistics:
--------------------------------------------------------Media Speed Running: 16 Mbps Half Duplex
Media Speed Selected: 16 Mbps Full Duplex
Receive Overruns : 0
Transmit Underruns : 0
ARI/FCI errors : 0
Microcode level on the adapter :001PX11B2
Num pkts in priority sw tx queue : 0
Num pkts in priority hw tx queue : 0
Open Firmware Level : 001PXRS02
The highlighted fields are described as follows:
v Transmit and Receive Errors
Number of output/input errors encountered on this device. This field counts unsuccessful transmissions
due to hardware/network errors.
These unsuccessful transmissions could also slow down the performance of the system.
v Max Packets on S/W Transmit Queue
Maximum number of outgoing packets ever queued to the software transmit queue.
An indication of an inadequate queue size is if the maximal transmits queued equals the current queue
size (xmt_que_size). This indicates that the queue was full at some point.
To check the current size of the queue, use the lsattr -El adapter command (where adapter is, for
example, tok0 or ent0). Because the queue is associated with the device driver and adapter for the
interface, use the adapter name, not the interface name. Use the SMIT or the chdev command to
change the queue size.
v S/W Transmit Queue Overflow
Number of outgoing packets that have overflowed the software transmit queue. A value other than zero
requires the same actions as would be needed if the Max Packets on S/W Transmit Queue reaches the
xmt_que_size. The transmit queue size must be increased.
v Broadcast Packets
Number of broadcast packets received without any error.
If the value for broadcast packets is high, compare it with the total received packets. The received
broadcast packets should be less than 20 percent of the total received packets. If it is high, this could
236
Performance Management Guide
be an indication of a high network load; use multicasting. The use of IP multicasting enables a message
to be transmitted to a group of hosts, instead of having to address and send the message to each
group member individually.
v DMA Overrun
The DMA Overrun statistic is incremented when the adapter is using DMA to put a packet into system
memory and the transfer is not completed. There are system buffers available for the packet to be
placed into, but the DMA operation failed to complete. This occurs when the MCA bus is too busy for
the adapter to be able to use DMA for the packets. The location of the adapter on the bus is crucial in a
heavily loaded system. Typically an adapter in a lower slot number on the bus, by having the higher bus
priority, is using so much of the bus that adapters in higher slot numbers are not being served. This is
particularly true if the adapters in a lower slot number are ATM or SSA adapters.
v Max Collision Errors
Number of unsuccessful transmissions due to too many collisions. The number of collisions encountered
exceeded the number of retries on the adapter.
v Late Collision Errors
Number of unsuccessful transmissions due to the late collision error.
v Timeout Errors
Number of unsuccessful transmissions due to adapter reported timeout errors.
v Single Collision Count
Number of outgoing packets with single (only one) collision encountered during transmission.
v Multiple Collision Count
Number of outgoing packets with multiple (2 - 15) collisions encountered during transmission.
v Receive Collision Errors
Number of incoming packets with collision errors during reception.
v No mbuf Errors
Number of times that mbufs were not available to the device driver. This usually occurs during receive
operations when the driver must obtain memory buffers to process inbound packets. If the mbuf pool for
the requested size is empty, the packet will be discarded. Use the netstat -m command to confirm this,
and increase the parameter thewall.
The No mbuf Errors value is interface-specific and not identical to the requests for mbufs denied from
the netstat -m output. Compare the values of the example for the commands netstat -m and netstat -v
(Ethernet and Token-Ring part).
To determine network performance problems, check for any Error counts in the netstat -v output.
Additional guidelines:
v To check for an overloaded Ethernet network, calculate (from the netstat -v command):
(Max Collision Errors + Timeouts Errors) / Transmit Packets
If the result is greater than 5 percent, reorganize the network to balance the load.
v Another indication for a high network load is (from the command netstat -v):
If the total number of collisions from the netstat -v output (for Ethernet) is greater than 10 percent of
the total transmitted packets, as follows:
Number of collisions / Number of Transmit Packets > 0.1
netstat -p protocol: Shows statistics about the value specified for the protocol variable (udp, tcp, ip,
icmp), which is either a well-known name for a protocol or an alias for it. Some protocol names and aliases
are listed in the /etc/protocols file. A null response indicates that there are no numbers to report. If there
is no statistics routine for it, the program report of the value specified for the protocol variable is unknown.
The following example shows the output for the ip protocol:
Chapter 10. Monitoring and Tuning Communications I/O Use
237
# netstat -p ip
ip:
:
491351 total packets received
0 bad header checksums
0 with size smaller than minimum
0 with data size < data length
0 with header length < data size
0 with data length < header length
0 with bad options
0 with incorrect version number
25930 fragments received
0 fragments dropped (dup or out of space)
0 fragments dropped after timeout
12965 packets reassembled ok
475054 packets for this host
0 packets for unknown/unsupported protocol
0 packets forwarded
3332 packets not forwardable
0 redirects sent
405650 packets sent from this host
0 packets sent with fabricated ip header
0 output packets dropped due to no bufs, etc.
0 output packets discarded due to no route
5498 output datagrams fragmented
10996 fragments created
0 datagrams that can’t be fragmented
0 IP Multicast packets dropped due to no receiver
0 ipintrq overflows
The highlighted fields are described as follows:
v Total Packets Received
Number of total IP datagrams received.
v Bad Header Checksum or Fragments Dropped
If the output shows bad header checksum or fragments dropped due to dup or out of space, this
indicates either a network that is corrupting packets or device driver receive queues that are not large
enough.
v Fragments Received
Number of total fragments received.
v Dropped after Timeout
If the fragments dropped after timeout is other than zero, then the time to life counter of the ip
fragments expired due to a busy network before all fragments of the datagram arrived. To avoid this,
use the no command to increase the value of the ipfragttl network parameter. Another reason could be
a lack of mbufs; increase thewall.
v Packets Sent from this Host
Number of IP datagrams that were created and sent out from this system. This counter does not include
the forwarded datagrams (passthrough traffic).
v Fragments Created
Number of fragments created in this system when IP datagrams were sent out.
When viewing IP statistics, look at the ratio of packets received to fragments received. As a guideline for
small MTU networks, if 10 percent or more of the packets are getting fragmented, you should investigate
further to determine the cause. A large number of fragments indicates that protocols above the IP layer on
remote hosts are passing data to IP with data sizes larger than the MTU for the interface.
Gateways/routers in the network path might also have a much smaller MTU size than the other nodes in
the network. The same logic can be applied to packets sent and fragments created.
238
Performance Management Guide
Fragmentation results in additional CPU overhead so it is important to determine its cause. Be aware that
some applications, by their very nature, can cause fragmentation to occur. For example, an application that
sends small amounts of data can cause fragments to occur. However, if you know the application is
sending large amounts of data and fragmentation is still occurring, determine the cause. It is likely that the
MTU size used is not the MTU size configured on the systems.
The following example shows the output for the udp protocol:
# netstat -p udp
udp:
11521194 datagrams received
0 incomplete headers
0 bad data length fields
0 bad checksums
16532 dropped due to no socket
232850 broadcast/multicast datagrams dropped due to no socket
77 socket buffer overflows
11271735 delivered
796547 datagrams output
Statistics of interest are:
v Bad Checksums
Bad checksums could happen due to hardware card or cable failure.
v Dropped Due to No Socket
Number of received UDP datagrams of that destination socket ports were not opened. As a result, the
ICMP Destination Unreachable - Port Unreachable message must have been sent out. But if the
received UDP datagrams were broadcast datagrams, ICMP errors are not generated. If this value is
high, investigate how the application is handling sockets.
v Socket Buffer Overflows
Socket buffer overflows could be due to insufficient transmit and receive UDP sockets, too few nfsd
daemons, or too small nfs_socketsize, udp_recvspace and sb_max values.
If the netstat -p udp command indicates socket overflows, then you might need to increase the number of
the nfsd daemons on the server. First, check the affected system for CPU or I/O saturation, and verify the
recommended setting for the other communication layers by using the no -a command. If the system is
saturated, you must either to reduce its load or increase its resources.
The following example shows the output for the tcp protocol:
# netstat -p tcp
tcp:
63726 packets sent
34309 data packets (6482122 bytes)
198 data packets (161034 bytes) retransmitted
17437 ack-only packets (7882 delayed)
0 URG only packets
0 window probe packets
3562 window update packets
8220 control packets
71033 packets received
35989 acks (for 6444054 bytes)
2769 duplicate acks
0 acks for unsent data
47319 packets (19650209 bytes) received in-sequence
182 completely duplicate packets (29772 bytes)
4 packets with some dup. data (1404 bytes duped)
2475 out-of-order packets (49826 bytes)
0 packets (0 bytes) of data after window
0 window probes
800 window update packets
77 packets received after close
Chapter 10. Monitoring and Tuning Communications I/O Use
239
0 packets with bad hardware assisted checksum
0 discarded for bad checksums
0 discarded for bad header offset fields
0 connection request
3125 connection requests
1626 connection accepts
4731 connections established (including accepts)
5543 connections closed (including 31 drops)
62 embryonic connections dropped
38552 segments updated rtt (of 38862 attempts)
0 resends due to path MTU discovery
3 path MTU discovery terminations due to retransmits
553 retransmit timeouts
28 connections dropped by rexmit timeout
0 persist timeouts
464 keepalive timeouts
26 keepalive probes sent
1 connection dropped by keepalive
0 connections in timewait reused
0 delayed ACKs for SYN
0 delayed ACKs for FIN
0 send_and_disconnects
Statistics of interest are:
v Packets Sent
v Data Packets
v Data Packets Retransmitted
v Packets Received
v Completely Duplicate Packets
v Retransmit Timeouts
For the TCP statistics, compare the number of packets sent to the number of data packets retransmitted. If
the number of packets retransmitted is over 10-15 percent of the total packets sent, TCP is experiencing
timeouts indicating that network traffic may be too high for acknowledgments (ACKs) to return before a
timeout. A bottleneck on the receiving node or general network problems can also cause TCP
retransmissions, which will increase network traffic, further adding to any network performance problems.
Also, compare the number of packets received with the number of completely duplicate packets. If TCP on
a sending node times out before an ACK is received from the receiving node, it will retransmit the packet.
Duplicate packets occur when the receiving node eventually receives all the retransmitted packets. If the
number of duplicate packets exceeds 10-15 percent, the problem may again be too much network traffic or
a bottleneck at the receiving node. Duplicate packets increase network traffic.
The value for retransmit timeouts occurs when TCP sends a packet but does not receive an ACK in time.
It then resends the packet. This value is incremented for any subsequent retransmittals. These continuous
retransmittals drive CPU utilization higher, and if the receiving node does not receive the packet, it
eventually will be dropped.
netstat -s: The netstat -s command shows statistics for each protocol (while the netstat -p command
shows the statistics for the specified protocol).
netstat -s -s: The undocumented -s -s option shows only those lines of the netstat -s output that are not
zero, making it easier to look for error counts.
netstat -s -Z: This is an undocumented function of the netstat command. It clears all the statistic
counters for the netstat -s command to zero.
240
Performance Management Guide
netstat -r: Another option relevant to performance is the display of the discovered Path Maximum
Transmission Unit (PMTU).
For two hosts communicating across a path of multiple networks, a transmitted packet will become
fragmented if its size is greater than the smallest MTU of any network in the path. Because packet
fragmentation can result in reduced network performance, it is desirable to avoid fragmentation by
transmitting packets with a size no larger than the smallest MTU in the network path. This size is called
the path MTU.
Use the netstat -r command to display this value. In the following is example the netstat -r -f inet
command is used to display only the routing tables:
# netstat -r -f inet
Routing tables
Destination
Gateway
Flags
Refs
Route Tree for Protocol Family 2:
default
itsorusi
9.3.1
sv2019e
itsonv
sv2019e
itsorusi
sv2019e
ah6000d
sv2019e
ah6000e
sv2019e
sv2019e
sv2019e
coyote.ncs.mainz itsorusi
kresna.id.ibm.co itsorusi
9.184.104.111
kresna.id.ibm.com
127
localhost
UGc
Uc
UHW
UHW
UHW
UHW
UHW
UGHW
UGHW
UGc
U
1
25
0
1
1
0
4
1
0
0
3
Use
348
12504
235
883
184
209
11718
45
14
5
96
PMTU If
1492
1492
1492
1492
1492
-
tr0
tr0
tr0
tr0
tr0
tr0
tr0
tr0
tr0
tr0
lo0
Exp
-
Groups
-
netstat -D: The -D option allows you to see packets coming into and going out of each layer in the
communications subsystem along, with packets dropped at each layer.
# netstat -D
Source
Ipkts
Opkts
Idrops
Odrops
------------------------------------------------------------------------------tok_dev0
19333058
402225
3
0
ent_dev0
0
0
0
0
--------------------------------------------------------------Devices Total
19333058
402225
3
0
------------------------------------------------------------------------------tok_dd0
19333055
402225
0
0
ent_dd0
0
0
0
0
--------------------------------------------------------------Drivers Total
19333055
402225
0
0
------------------------------------------------------------------------------tok_dmx0
796966
N/A
18536091
N/A
ent_dmx0
0
N/A
0
N/A
--------------------------------------------------------------Demuxer Total
796966
N/A
18536091
N/A
------------------------------------------------------------------------------IP
694138
677411
7651
6523
TCP
143713
144247
0
0
UDP
469962
266726
0
812
--------------------------------------------------------------Protocols Total
1307813
1088384
7651
7335
------------------------------------------------------------------------------lo_if0
22088
22887
799
0
tr_if0
796966
402227
0
289
--------------------------------------------------------------Net IF Total
819054
425114
799
289
--------------------------------------------------------------------------------------------------------------------------------------------NFS/RPC Total
N/A
1461
0
0
------------------------------------------------------------------------------(Note: N/A -> Not Applicable)
Chapter 10. Monitoring and Tuning Communications I/O Use
241
The Devices layer shows number of packets coming into the adapter, going out of the adapter, and
number of packets dropped on input and output. There are various causes of adapter errors, and the
netstat -v command can be examined for more details.
The Drivers layer shows packet counts handled by the device driver for each adapter. Output of the
netstat -v command is useful here to determine which errors are counted.
The Demuxer values show packet counts at the demux layer, and Idrops here usually indicate that filtering
has caused packets to be rejected (for example, Netware or DecNet packets being rejected because these
are not handled by the system under examination).
Details for the Protocols layer can be seen in the output of the netstat -s command.
Note: In the statistics output, a N/A displayed in a field value indicates the count is not applicable. For the
NFS/RPC statistics, the number of incoming packets that pass through RPC are the same packets
which pass through NFS, so these numbers are not summed in the NFS/RPC Total field, hence the
N/A. NFS has no outgoing packet or outgoing packet drop counters specific to NFS and RPC.
Therefore, individual counts have a field value of N/A, and the cumulative count is stored in the
NFS/RPC Total field.
The netpmon Command
The netpmon command uses the trace facility to obtain a detailed picture of network activity during a time
interval. Because it uses the trace facility, the netpmon command can be run only by a root user or by a
member of the system group.
Also, the netpmon command cannot run together with any of the other trace-based performance
commands such as tprof and filemon. In its usual mode, the netpmon command runs in the background
while one or more application programs or system commands are being executed and monitored.
The netpmon command focuses on the following system activities:
v CPU usage
– By processes and interrupt handlers
– How much is network-related
– What causes idle time
v Network device driver I/O
– Monitors I/O operations through all Ethernet, Token-Ring, and Fiber-Distributed Data Interface (FDDI)
network device drivers.
– In the case of transmission I/O, the command monitors utilizations, queue lengths, and destination
hosts. For receive ID, the command also monitors time in the demux layer.
v Internet socket calls
– Monitors send(), recv(), sendto(), recvfrom(), sendmsg(), read(), and write() subroutines on
Internet sockets.
– Reports statistics on a per-process basis for the Internet Control Message Protocol (ICMP),
Transmission Control Protocol (TCP), and the User Datagram Protocol (UDP).
v NFS I/O
– On client: RPC requests, NFS read/write requests.
– On server: Per-client, per-file, read/write requests.
The following will be computed:
v Response times and sizes associated with transmit and receive operations at the device driver level.
v Response times and sizes associated with all types of Internet socket read and write system calls.
242
Performance Management Guide
v Response times and sizes associated with NFS read and write system calls.
v Response times associated with NFS remote procedure call requests.
To determine whether the netpmon command is installed and available, run the following command:
# lslpp -lI perfagent.tools
Tracing is started by the netpmon command, optionally suspended with the trcoff subcommand and
resumed with the trcon subcommand, and terminated with the trcstop subcommand. As soon as tracing
is terminated, the netpmon command writes its report to standard output.
Using netpmon
The netpmon command will start tracing immediately unless the -d option is used. Use the trcstop
command to stop tracing. At that time, all the specified reports are generated, and the netpmon command
exits. In the client-server environment, use the netpmon command to view how networking affects the
overall performance. It can be run on both client and server.
The netpmon command can read the I/O trace data from a specified file, instead of from the real-time
trace process. In this case, the netpmon report summarizes the network activity for the system and period
represented by the trace file. This offline processing method is useful when it is necessary to postprocess
a trace file from a remote machine or perform the trace data collection at one time and postprocess it at
another time.
The trcrpt -r command must be executed on the trace logfile and redirected to another file, as follows:
# gennames > gennames.out
# trcrpt -r trace.out > trace.rpt
At this point, an adjusted trace logfile is fed into the netpmon command to report on I/O activity captured
by a previously recorded trace session as follows:
# netpmon -i trace.rpt -n gennames.out | pg
In this example, the netpmon command reads file system trace events from the trace.rpt input file.
Because the trace data is already captured on a file, the netpmon command does not put itself in the
background to allow application programs to be run. After the entire file is read, a network activity report
will be displayed on standard output (which, in this example, is piped to the pg command).
If the trace command was run with the -C all flag, then run the trcrpt command also with the -C all flag
(see Formatting a Report from trace -C Output).
The following netpmon command running on an NFS server executes the sleep command and creates a
report after 400 seconds. During the measured interval, a copy to an NFS-mounted file system /nfs_mnt
is taking place.
# netpmon -o netpmon.out -O all; sleep 400; trcstop
With the -O option, you can specify the report type to be generated. Valid report type values are:
cpu
CPU usage
dd
Network device-driver I/O
so
Internet socket call I/O
nfs
NFS I/O
all
All reports are produced. The following is the default value.
# cat netpmon.out
Thu Jan 21 15:02:45 2000
System: AIX itsosmp Node: 4 Machine: 00045067A000
Chapter 10. Monitoring and Tuning Communications I/O Use
243
401.053 secs in measured interval
========================================================================
Process CPU Usage Statistics:
----------------------------Network
Process (top 20)
PID CPU Time
CPU %
CPU %
---------------------------------------------------------nfsd
12370
42.2210
2.632
2.632
nfsd
12628
42.0056
2.618
2.618
nfsd
13144
41.9540
2.615
2.615
nfsd
12886
41.8680
2.610
2.610
nfsd
12112
41.4114
2.581
2.581
nfsd
11078
40.9443
2.552
2.552
nfsd
11854
40.6198
2.532
2.532
nfsd
13402
40.3445
2.515
2.515
lrud
1548
16.6294
1.037
0.000
netpmon
15218
5.2780
0.329
0.000
gil
2064
2.0766
0.129
0.129
trace
18284
1.8820
0.117
0.000
syncd
3602
0.3757
0.023
0.000
swapper
0
0.2718
0.017
0.000
init
1
0.2201
0.014
0.000
afsd
8758
0.0244
0.002
0.000
bootpd
7128
0.0220
0.001
0.000
ksh
4322
0.0213
0.001
0.000
pcimapsvr.ip
16844
0.0204
0.001
0.000
netm
1806
0.0186
0.001
0.001
---------------------------------------------------------Total (all processes)
358.3152 22.336 20.787
Idle time
1221.0235 76.114
========================================================================
First Level Interrupt Handler CPU Usage Statistics:
--------------------------------------------------Network
FLIH
CPU Time
CPU %
CPU %
---------------------------------------------------------PPC decrementer
9.9419
0.620
0.000
external device
4.5849
0.286
0.099
UNKNOWN
0.1716
0.011
0.000
data page fault
0.1080
0.007
0.000
floating point
0.0012
0.000
0.000
instruction page fault
0.0007
0.000
0.000
---------------------------------------------------------Total (all FLIHs)
14.8083
0.923
0.099
========================================================================
Second Level Interrupt Handler CPU Usage Statistics:
---------------------------------------------------Network
SLIH
CPU Time
CPU %
CPU %
---------------------------------------------------------tokdd
12.4312
0.775
0.775
ascsiddpin
0.5178
0.032
0.000
---------------------------------------------------------Total (all SLIHs)
12.9490
0.807
0.775
========================================================================
Network Device-Driver Statistics (by Device):
------------------------------------------------------- Xmit ------------------ Recv --------Device
Pkts/s Bytes/s Util QLen
Pkts/s Bytes/s
Demux
-----------------------------------------------------------------------------token ring 0
31.61
4800 1.7% 0.046
200.93
273994 0.0080
========================================================================
244
Performance Management Guide
Network Device-Driver Transmit Statistics (by Destination Host):
---------------------------------------------------------------Host
Pkts/s Bytes/s
---------------------------------------ah6000c
31.57
4796
9.3.1.255
0.03
4
itsorusi
0.00
0
========================================================================
TCP Socket Call Statistics (by Process):
--------------------------------------------- Read --------- Write ----Process (top 20)
PID
Calls/s
Bytes/s
Calls/s
Bytes/s
-----------------------------------------------------------------------telnetd
18144
0.03
123
0.06
0
-----------------------------------------------------------------------Total (all processes)
0.03
123
0.06
0
========================================================================
NFS Server Statistics (by Client):
--------------------------------------- Read --------- Write ----Other
Client
Calls/s
Bytes/s
Calls/s
Bytes/s
Calls/s
-----------------------------------------------------------------------ah6000c
0.00
0
31.54
258208
0.01
-----------------------------------------------------------------------Total (all clients)
0.00
0
31.54
258208
0.01
========================================================================
Detailed Second Level Interrupt Handler CPU Usage Statistics:
------------------------------------------------------------SLIH: tokdd
count:
93039
cpu time (msec):
avg 0.134
min 0.026
max 0.541
sdev 0.051
SLIH: ascsiddpin
count:
8136
cpu time (msec):
avg 0.064
min 0.012
max 0.147
sdev 0.018
COMBINED (All SLIHs)
count:
101175
cpu time (msec):
avg 0.128
min 0.012
max 0.541
sdev 0.053
========================================================================
Detailed Network Device-Driver Statistics:
-----------------------------------------DEVICE: token ring 0
recv packets:
80584
recv sizes (bytes):
avg 1363.6 min 50
max 1520
sdev 356.3
recv times (msec):
avg 0.081
min 0.010
max 0.166
sdev 0.020
demux times (msec):
avg 0.040
min 0.008
max 0.375
sdev 0.040
xmit packets:
12678
xmit sizes (bytes):
avg 151.8
min 52
max 184
sdev 3.3
xmit times (msec):
avg 1.447
min 0.509
max 4.514
sdev 0.374
========================================================================
Detailed Network Device-Driver Transmit Statistics (by Host):
------------------------------------------------------------HOST: ah6000c
xmit packets:
12662
xmit sizes (bytes):
avg 151.9
min 52
max 184
sdev 2.9
xmit times (msec):
avg 1.448
min 0.509
max 4.514
sdev 0.373
HOST: 9.3.1.255
xmit packets:
xmit sizes (bytes):
xmit times (msec):
14
avg 117.0
avg 1.133
min 117
min 0.884
max 117
max 1.730
sdev 0.0
sdev 0.253
Chapter 10. Monitoring and Tuning Communications I/O Use
245
HOST: itsorusi
xmit packets:
xmit sizes (bytes):
xmit times (msec):
1
avg 84.0
avg 0.522
min 84
min 0.522
max 84
max 0.522
sdev 0.0
sdev 0.000
========================================================================
Detailed TCP Socket Call Statistics (by Process):
------------------------------------------------PROCESS: telnetd
PID: 18144
reads:
12
read sizes (bytes):
avg 4096.0 min 4096
max 4096
sdev 0.0
read times (msec):
avg 0.085
min 0.053
max 0.164
sdev 0.027
writes:
23
write sizes (bytes): avg 3.5
min 1
max 26
sdev 7.0
write times (msec):
avg 0.143
min 0.067
max 0.269
sdev 0.064
PROTOCOL: TCP (All Processes)
reads:
12
read sizes (bytes):
avg 4096.0 min 4096
max 4096
sdev 0.0
read times (msec):
avg 0.085
min 0.053
max 0.164
sdev 0.027
writes:
23
write sizes (bytes): avg 3.5
min 1
max 26
sdev 7.0
write times (msec):
avg 0.143
min 0.067
max 0.269
sdev 0.064
========================================================================
Detailed NFS Server Statistics (by Client):
------------------------------------------CLIENT: ah6000c
writes:
12648
write sizes (bytes): avg 8187.5 min 4096
max 8192
sdev 136.2
write times (msec):
avg 138.646 min 0.147
max 1802.067 sdev 58.853
other calls:
5
other times (msec):
avg 1.928
min 0.371
max 8.065
sdev 3.068
COMBINED (All Clients)
writes:
12648
write sizes (bytes): avg 8187.5 min 4096
max 8192
sdev 136.2
write times (msec):
avg 138.646 min 0.147
max 1802.067 sdev 58.853
other calls:
5
other times (msec):
avg 1.928
min 0.371
max 8.065
sdev 3.068
The output of the netpmon command is composed of two different types of reports: global and detailed.
The global reports list statistics as follows:
v Most active processes
v First-level interrupt handlers
v Second-level interrupt handlers
v
v
v
v
Network device drivers
Network device-driver transmits
TCP socket calls
NFS server or client statistics
The global reports are shown at the beginning of the netpmon output and are the occurrences during the
measured interval. The detailed reports provide additional information for the global reports. By default, the
reports are limited to the 20 most active statistics measured. All information in the reports is listed from top
to bottom as most active to least active.
Global Reports of netpmon
The reports generated by the netpmon command begin with a header, which identifies the date, the
machine ID, and the length of the monitoring period in seconds. The header is followed by a set of global
and detailed reports for all specified report types.
246
Performance Management Guide
Process CPU Usage Statistics: Each row describes the CPU usage associated with a process. Unless
the verbose (-v) option is specified, only the 20 most active processes are included in the list. At the
bottom of the report, CPU usage for all processes is totaled, and CPU idle time is reported. The idle time
percentage number is calculated from the idle time divided by the measured interval. The difference
between the CPU time totals and measured interval is due to Interrupt handlers.
The Network CPU % is the percentage of total time that this process spent executing network-related code.
If the -t flag is used, a thread CPU usage statistic is also present. Each process row described above is
immediately followed by rows describing the CPU usage of each thread owned by that process. The fields
in these rows are identical to those for the process, except for the name field. Threads are not named.
In the example report, the Idle time percentage number (76.114 percent) shown in the global CPU usage
report is calculated from the Idle time (1221.0235) divided by the measured interval times 4 (401.053
times 4), because there are four CPUs in this server. If you want to look at each CPU’s activity, you can
use sar, ps, or any other SMP-specific command. Similar calculation applies to the total CPU % that is
occupied by all processes. The Idle time is due to network I/O. The difference between the CPU Time
totals (1221.0235 + 358.315) and the measured interval is due to interrupt handlers and the multiple
CPUs. It appears that in the example report, the majority of the CPU usage was network-related: (20.787 /
22.336) = 93.07 percent. About 77.664 percent of CPU usage is either CPU idle or CPU wait time.
Note: If the result of total network CPU % divided by total CPU % is greater than 0.5 from Process CPU Usage
Statistics for NFS server, then the majority of CPU usage is network-related.
This method is also a good way to view CPU usage by process without tying the output to a specific
program.
First Level Interrupt Handler CPU Usage Statistics: Each row describes the CPU usage associated
with a first-level interrupt handler (FLIH). At the bottom of the report, CPU usage for all FLIHs is totaled.
CPU Time
Total amount of CPU time used by this FLIH
CPU %
CPU usage for this interrupt handler as a percentage of total time
Network CPU %
Percentage of total time that this interrupt handler executed on behalf of network-related events
Second Level Interrupt Handler CPU Usage Statistics: Each row describes the CPU usage associated
with a second-level interrupt handler (SLIH). At the bottom of the report, CPU usage for all SLIHs is
totaled.
Network Device-Driver Statistics (by Device): Each row describes the statistics associated with a
network device.
Device
Name of special file associated with device
Xmit Pkts/s
Packets per second transmitted through this device
Xmit Bytes/s
Bytes per second transmitted through this device
Xmit Util
Busy time for this device, as a percent of total time
Chapter 10. Monitoring and Tuning Communications I/O Use
247
Xmit Qlen
Number of requests waiting to be transmitted through this device, averaged over time, including
any transaction currently being transmitted
Recv Pkts/s
Packets per second received through this device
Recv Bytes/s
Bytes per second received through this device
Recv Demux
Time spent in demux layer as a fraction of total time
In this example, the Xmit QLen is only 0.046. This number is very small compared to its default size (30).
Its Recv Bytes/s is 273994, much smaller than the Token-Ring transmit speed (16 Mb/s). Therefore, in this
case, the network is not saturated, at least from this system’s view.
Network Device-Driver Transmit Statistics (by Destination Host): Each row describes the amount of
transmit traffic associated with a particular destination host, at the device-driver level.
Host
Destination host name. An asterisk (*) is used for transmissions for which no host name can be
determined.
Pkts/s Packets per second transmitted to this host.
Bytes/s
Bytes per second transmitted to this host.
TCP Socket Call Statistics for Each Internet Protocol (by Process): These statistics are shown for
each used Internet protocol. Each row describes the amount of read() and write() subroutine activity on
sockets of this protocol type associated with a particular process. At the bottom of the report, all socket
calls for this protocol are totaled.
NFS Server Statistics (by Client): Each row describes the amount of NFS activity handled by this
server on behalf of a particular client. At the bottom of the report, calls for all clients are totaled.
On a client machine, the NFS server statistics are replaced by the NFS client statistics (NFS Client
Statistics for each Server (by File), NFS Client RPC Statistics (by Server), NFS Client
Statistics (by Process)).
Detailed Reports of netpmon
Detailed reports are generated for all requested (-O) report types. For these report types, a detailed report
is produced in addition to the global reports. The detailed reports contain an entry for each entry in the
global reports with statistics for each type of transaction associated with the entry.
Transaction statistics consist of a count of the number of transactions for that type, followed by response
time and size distribution data (where applicable). The distribution data consists of average, minimum, and
maximum values, as well as standard deviations. Roughly two-thirds of the values are between average
minus standard deviation and average plus standard deviation. Sizes are reported in bytes. Response
times are reported in milliseconds.
Detailed Second-Level Interrupt Handler CPU-Usage Statistics: The output fields are described as
follows:
SLIH
Name of second-level interrupt handler
count Number of interrupts of this type
cpu time (msec)
CPU usage statistics for handling interrupts of this type
248
Performance Management Guide
Detailed Network Device-Driver Statistics (by Device): The output fields are described as follows:
DEVICE
Path name of special file associated with device
recv packets
Number of packets received through this device
recv sizes (bytes)
Size statistics for received packets
recv times (msec)
Response time statistics for processing received packets
demux times (msec)
Time statistics for processing received packets in the demux layer
xmit packets
Number of packets transmitted through this device
xmit sizes (bytes)
Size statistics for transmitted packets
xmit times (msec)
Response time statistics for processing transmitted packets
There are other detailed reports, such as Detailed Network Device-Driver Transmit Statistics (by
Host) and Detailed TCP Socket Call Statistics for Each Internet Protocol (by Process). For an
NFS client, there are the Detailed NFS Client Statistics for Each Server (by File), Detailed NFS
Client RPC Statistics (by Server), and Detailed NFS Client Statistics (by Process) reports. For an
NFS server, there is the Detailed NFS Server Statistics (by Client) report. They have similar output
fields as explained above.
In the example, the results from the Detailed Network Device-Driver Statistics lead to the following:
v recv bytes = 80584 packets * 1364 bytes/packet = 109,916,576 bytes
v xmit bytes = 12678 packets * 152 bytes/packet = 1,927,056 bytes
v total bytes exchanged = 109,916,576 + 1,927,056 = 111,843,632 bytes
v total bits exchanged = 111,843,632 * 8 bits/byte = 894,749,056 bits
v transmit speed = 894,749,056 / 401.053 = 2.23 Mb/s (assuming that the copy took the entire monitoring
period)
As in the global device driver report, you can conclude that this case is not network-saturated. The
average receive size is 1363.6 bytes, near to the default MTU (maximum transmission unit) value, which is
1492 when the device is a Token-Ring card. If this value is larger than the MTU (from lsattr -E -l interface,
replacing interface with the interface name, such as en0 or tr0, you could change the MTU or adapter
transmit-queue length value to get better performance with the following command:
# ifconfig tr0 mtu 8500
or
# chdev -l ’tok0’ -a xmt_que_size=’150’
If the network is congested already, changing the MTU or queue value will not help.
Notes:
1. If transmit and receive packet sizes are small on the device driver statistics report, then increasing the
current MTU size will probably result in better network performance.
2. If system wait time due to network calls is high from the network wait time statistics for the NFS client
report, the poor performance is due to the network.
Chapter 10. Monitoring and Tuning Communications I/O Use
249
Limitations of netpmon
The netpmon command uses the trace facility to collect the statistics. Therefore, it has an impact on the
system workload, as follows.
v In a moderate, network-oriented workload, the netpmon command increases overall CPU utilization by
3-5 percent.
v In a CPU-saturated environment with little I/O of any kind, the netpmon command slowed a large
compile by about 3.5 percent.
To alleviate these situations, use offline processing and on systems with many CPUs use the -C all flag
with the trace command.
The traceroute Command
While the ping command confirms IP network reachability, you cannot pinpoint and improve some isolated
problems. Consider the following situation:
v When there are many hops (for example, gateways or routes) between your system and the destination,
and there seems to be a problem somewhere along the path. The destination system may have a
problem, but you need to know where a packet is actually lost.
v The ping command hangs up and does not tell you the reasons for a lost packet.
The traceroute command can inform you where the packet is located and why the route is lost. If your
packets must pass through routers and links, which belong to and are managed by other organizations or
companies, it is difficult to check the related routers through the telnet command. The traceroute
command provides a supplemental role to the ping command.
Note: The traceroute command is intended for use in network testing, measurement, and management. It
should be used primarily for manual fault isolation. Because of the load it imposes on the network,
do not use the traceroute command during typical operations or from automated scripts.
Successful traceroute Examples
The traceroute command uses UDP packets and uses the ICMP error-reporting function. It sends a UDP
packet three times to each gateway or router on the way. It starts with the nearest gateway and expands
the search by one hop. Finally, the search gets to the destination system. In the output, you see the
gateway name, the gateway’s IP address, and three round-trip times for the gateway. See the following
example:
# traceroute wave
trying to get source for wave
source should be 9.53.155.187
traceroute to wave.austin.ibm.com (9.53.153.120) from 9.53.155.187 (9.53.155.187), 30 hops max
outgoing MTU = 1500
1 9.111.154.1 (9.111.154.1) 5 ms 3 ms 2 ms
2 wave (9.53.153.120) 5 ms 5 ms 5 ms
Following is another example:
# traceroute wave
trying to get source for wave
source should be 9.53.155.187
traceroute to wave.austin.ibm.com (9.53.153.120) from 9.53.155.187 (9.53.155.187), 30 hops max
outgoing MTU = 1500
1 9.111.154.1 (9.111.154.1) 10 ms 2 ms 3 ms
2 wave (9.53.153.120) 8 ms 7 ms 5 ms
After the address resolution protocol (ARP) entry expired, the same command was repeated. Note that the
first packet to each gateway or destination took a longer round-trip time. This is due to the overhead
caused by the ARP. If a public-switched network (WAN) is involved in the route, the first packet consumes
250
Performance Management Guide
a lot of memory due to a connection establishment and may cause a timeout. The default timeout for each
packet is 3 seconds. You can change it with the -w option.
The first 10 ms is due to the ARP between the source system (9.53.155.187) and the gateway
9.111.154.1. The second 8 ms is due to the ARP between the gateway and the final destination (wave). In
this case, you are using DNS, and every time before the traceroute command sends a packet, the DNS
server is searched.
Failed traceroute Examples
For a long path to your destination or complex network routes, you may see a lot of problems with the
traceroute command. Because many things are implementation-dependent, searching for the problem
may only waste your time. If all routers or systems involved are under your control, you may be able to
investigate the problem completely.
Gateway (Router) Problem: In the following example, packets were sent from the system 9.53.155.187.
There are two router systems on the way to the bridge. The routing capability was intentionally removed
from the second router system by setting the option ipforwarding of the no command to 0. See the
following example:
# traceroute lamar
trying to get source for lamar
source should be 9.53.155.187
traceroute to lamar.austin.ibm.com (9.3.200.141) from 9.53.155.187 (9.53.155.187), 30 hops max
outgoing MTU = 1500
1 9.111.154.1 (9.111.154.1) 12 ms 3 ms 2 ms
2 9.111.154.1 (9.111.154.1) 3 ms !H * 6 ms !H
If an ICMP error message, excluding Time Exceeded and Port Unreachable, is received, it is displayed as
follows:
!H
Host Unreachable
!N
Network Unreachable
!P
Protocol Unreachable
!S
Source route failed
!F
Fragmentation needed
Destination System Problem: When the destination system does not respond within a 3-second
time-out interval, all queries are timed out, and the results are displayed with an asterisk (*).
# traceroute chuys
trying to get source for chuys
source should be 9.53.155.187
traceroute to chuys.austin.ibm.com (9.53.155.188) from 9.53.155.187 (9.53.155.187), 30 hops max
outgoing MTU = 1500
1 * * *
2 * * *
3 * * *
^C#
If you think that the problem is due to a communication link, use a longer timeout period with the -w flag.
Although rare, all the ports queried might have been used. You can change the ports and try again.
Number of ″hops″ to Destination: Another output example might be as follows:
# traceroute mysystem.university.edu (129.2.130.22)
traceroute to mysystem.university.edu (129.2.130.22), 30 hops max
1 helios.ee.lbl.gov (129.3.112.1) 0 ms 0 ms 0 ms
2 lilac-dmc.university.edu (129.2.216.1) 39 ms 19 ms 39 ms
3 lilac-dmc.university.edu (129.2.215.1) 19 ms 39 ms 19 ms
Chapter 10. Monitoring and Tuning Communications I/O Use
251
4 ccngw-ner-cc.university.edu (129.2.135.23) 39 ms 40 ms 19 ms
5 ccn-nerif35.university.edu (129.2.167.35) 39 ms 39 ms 39 ms
6 csgw/university.edu (129.2.132.254) 39 ms 59 ms 39 ms
7 * * *
8 * * *
9 * * *
10 * * *
11 * * *
12 * * *
13 rip.university.EDU (129.2.130.22) 59 ms! 39 ms! 39 ms!
In this example, exactly half of the 12 gateway hops (13 is the final destination) are ″missing.″ However,
these hops were actually not gateways. The destination host used the time to live (ttl) from the arriving
datagram as the ttl in its ICMP reply; thus, the reply timed out on the return path. Because ICMPs are not
sent for ICMPs, no notice was received. The ! (exclamation mark) after each round-trip time indicates
some type of software incompatibility problem. (The cause was diagnosed after the traceroute command
issued a probe of twice the path length. The destination host was really only seven hops away.)
The iptrace daemon, and the ipreport and ipfilter Commands
You can use many tools for observing network activity. Some run under the operating system, others run
on dedicated hardware. One tool that can be used to obtain a detailed, packet-by-packet description of the
LAN activity generated by a workload is the combination of the iptrace daemon and the ipreport
command. To use the iptrace daemon with operating system version 4, you need the bos.net.tcp.server
fileset. The iptrace daemon is included in this fileset, as well as some other useful commands such as the
trpt and tcdump commands. The iptrace daemon can only be started by a root user.
By default, the iptrace daemon traces all packets. The option -a allows exclusion of address resolution
protocol (ARP) packets. Other options can narrow the scope of tracing to a particular source host (-s),
destination host (-d), or protocol (-p). Because the iptrace daemon can consume significant amounts of
processor time, be as specific as possible when you describe the packets you want traced.
Because iptrace is a daemon, start the iptrace daemon with the startsrc command rather than directly
from the command line. This method makes it easier to control and shut down cleanly. A typical example
would be as follows:
# startsrc -s iptrace -a "-i tr0 /home/user/iptrace/log1"
This command starts the iptrace daemon with instructions to trace all activity on the Token-Ring interface,
tr0, and place the trace data in /home/user/iptrace/log1. To stop the daemon, use the following:
# stopsrc -s iptrace
If you did not start the iptrace daemon with the startsrc command, you must use the ps command to find
its process ID with and terminate it with the kill command.
The ipreport command is a formatter for the log file. Its output is written to standard output. Options allow
recognition and formatting of RPC packets (-r), identifying each packet with a number (-n), and prefixing
each line with a 3-character string that identifies the protocol (-s). A typical ipreport command to format
the log1 file just created (which is owned by the root user) would be as follows:
# ipreport -ns log1 >log1_formatted
This would result in a sequence of packet reports similar to the following examples. The first packet is the
first half of a ping packet. The fields of most interest are as follows:
v The source (SRC) and destination (DST) host address, both in dotted decimal and in ASCII
v The IP packet length (ip_len)
v The indication of the higher-level protocol in use (ip_p)
252
Performance Management Guide
Packet Number 131
TOK: =====( packet transmitted on interface tr0 )=====Fri Jan 14 08:42:07 2000
TOK: 802.5 packet
TOK: 802.5 MAC header:
TOK: access control field = 0, frame control field = 40
TOK: [ src = 90:00:5a:a8:88:81, dst = 10:00:5a:4f:35:82]
TOK: routing control field = 0830, 3 routing segments
TOK: routing segments [ ef31 ce61 ba30 ]
TOK: 802.2 LLC header:
TOK: dsap aa, ssap aa, ctrl 3, proto 0:0:0, type 800 (IP)
IP:
< SRC = 129.35.145.140 > (alborz.austin.ibm.com)
IP:
< DST = 129.35.145.135 > (xactive.austin.ibm.com)
IP:
ip_v=4, ip_hl=20, ip_tos=0, ip_len=84, ip_id=38892, ip_off=0
IP:
ip_ttl=255, ip_sum=fe61, ip_p = 1 (ICMP)
ICMP:
icmp_type=8 (ECHO_REQUEST) icmp_id=5923 icmp_seq=0
ICMP: 00000000
2d088abf 00054599 08090a0b 0c0d0e0f
|-.....E.........|
ICMP: 00000010
10111213 14151617 18191a1b 1c1d1e1f
|................|
ICMP: 00000020
20212223 24252627 28292a2b 2c2d2e2f
| !"#$%&’()*+,-./|
ICMP: 00000030
30313233 34353637
|01234567
|
The next example is a frame from an ftp operation. Note that the IP packet is the size of the MTU for this
LAN (1492 bytes).
Packet Number 501
TOK: =====( packet received on interface tr0 )=====Fri Dec 10 08:42:51 1999
TOK: 802.5 packet
TOK: 802.5 MAC header:
TOK: access control field = 18, frame control field = 40
TOK: [ src = 90:00:5a:4f:35:82, dst = 10:00:5a:a8:88:81]
TOK: routing control field = 08b0, 3 routing segments
TOK: routing segments [ ef31 ce61 ba30 ]
TOK: 802.2 LLC header:
TOK: dsap aa, ssap aa, ctrl 3, proto 0:0:0, type 800 (IP)
IP:
< SRC = 129.35.145.135 > (xactive.austin.ibm.com)
IP:
< DST = 129.35.145.140 > (alborz.austin.ibm.com)
IP:
ip_v=4, ip_hl=20, ip_tos=0, ip_len=1492, ip_id=34233, ip_off=0
IP:
ip_ttl=60, ip_sum=5ac, ip_p = 6 (TCP)
TCP:
<source port=20(ftp-data), destination port=1032 >
TCP:
th_seq=445e4e02, th_ack=ed8aae02
TCP:
th_off=5, flags<ACK |>
TCP:
th_win=15972, th_sum=0, th_urp=0
TCP: 00000000
01df0007 2cd6c07c 00004635 000002c2
|....,..|..F5....|
TCP: 00000010
00481002 010b0001 000021b4 00000d60
|.H........!....`|
--------- Lots of uninteresting data omitted ----------TCP: 00000590
63e40000 3860000f 4800177d 80410014
|c...8`..H..}.A..|
TCP: 000005a0
82220008 30610038 30910020
|."..0a.80..
|
The ipfilter command extracts different operation headers from an ipreport output file and displays them
in a table. Some customized NFS information regarding requests and replies is also provided.
To determine whether the ipfilter command is installed and available, run the following command:
# lslpp -lI perfagent.tools
An example command is as follows:
# ipfilter log1_formatted
The operation headers currently recognized are: udp, nfs, tcp, ipx, icmp. The ipfilter command has three
different types of reports, as follows:
v A single file (ipfilter.all) that displays a list of all the selected operations. The table displays packet
number, Time, Source & Destination, Length, Sequence #, Ack #, Source Port, Destination Port,
Network Interface, and Operation Type.
v Individual files for each selected header (ipfilter.udp, ipfilter.nfs, ipfilter.tcp, ipfilter.ipx, ipfilter.icmp).
The information contained is the same as ipfilter.all.
Chapter 10. Monitoring and Tuning Communications I/O Use
253
v A file nfs.rpt that reports on NFS requests and replies. The table contains: Transaction ID #, Type of
Request, Status of Request, Call Packet Number, Time of Call, Size of Call, Reply Packet Number,
Time of Reply, Size of Reply, and Elapsed millisecond between call and reply.
Adapter Statistics
The commands in this section provide output comparable to the netstat -v command. They allow you to
reset adapter statistics (-r) and to get more detailed output (-d) than the netstat -v command output
provides.
The entstat Command
The entstat command displays the statistics gathered by the specified Ethernet device driver. The user
can optionally specify that the device-specific statistics be displayed in addition to the device-generic
statistics. Using the -d option will list any extended statistics for this adapter and should be used to ensure
all statistics are displayed. If no flags are specified, only the device-generic statistics are displayed.
The entstat command is also invoked when the netstat command is run with the -v flag. The netstat
command does not issue any entstat command flags.
# entstat ent0
------------------------------------------------------------ETHERNET STATISTICS (ent0) :
Device Type: IBM 10/100 Mbps Ethernet PCI Adapter (23100020)
Hardware Address: 00:60:94:e9:29:18
Elapsed Time: 0 days 0 hours 0 minutes 0 seconds
Transmit Statistics:
-------------------Packets: 0
Bytes: 0
Interrupts: 0
Transmit Errors: 0
Packets Dropped: 0
Max Packets on S/W Transmit Queue: 0
S/W Transmit Queue Overflow: 0
Current S/W+H/W Transmit Queue Length: 0
Broadcast Packets: 0
Multicast Packets: 0
No Carrier Sense: 0
DMA Underrun: 0
Lost CTS Errors: 0
Max Collision Errors: 0
Late Collision Errors: 0
Deferred: 0
SQE Test: 0
Timeout Errors: 0
Single Collision Count: 0
Multiple Collision Count: 0
Current HW Transmit Queue Length: 0
Receive Statistics:
------------------Packets: 0
Bytes: 0
Interrupts: 0
Receive Errors: 0
Packets Dropped: 0
Bad Packets: 0
Broadcast Packets: 0
Multicast Packets: 0
CRC Errors: 0
DMA Overrun: 0
Alignment Errors: 0
No Resource Errors: 0
Receive Collision Errors: 0
Packet Too Short Errors: 0
Packet Too Long Errors: 0
Packets Discarded by Adapter: 0
Receiver Start Count: 0
General Statistics:
------------------No mbuf Errors: 0
Adapter Reset Count: 0
Driver Flags: Up Broadcast Running
Simplex 64BitSupport
In the above report, you may want to concentrate on:
Transmit Errors
Number of output errors encountered on this device. This is a counter for unsuccessful
transmissions due to hardware/network errors.
254
Performance Management Guide
Receive Errors
Number of input errors encountered on this device. This is a counter for unsuccessful reception
due to hardware/network errors.
Packets Dropped
Number of packets accepted by the device driver for transmission which were not (for any reason)
given to the device.
Max Packets on S/W Transmit Queue
Maximum number of outgoing packets ever queued to the software transmit queue.
S/W Transmit Queue Overflow
Number of outgoing packets that have overflowed the transmit queue.
No Resource Errors
Number of incoming packets dropped by the hardware due to lack of resources. This error usually
occurs because the receive buffers on the adapter were exhausted. Some adapters may have the
size of the receive buffers as a configurable parameter. Check the device configuration attributes
(or SMIT helps) for possible tuning information.
Single Collision Count/Multiple Collision Count
Number of collisions on an Ethernet network. These collisions are accounted for here rather than
in the collision column of the output of the netstat -i command.
Notice in this example, the Ethernet adapter is behaving well because there are no Receive Errors. These
errors are sometimes caused when a saturated network only transmits partial packets. The partial packets
are eventually retransmitted successfully but are recorded as receive errors.
If you receive S/W Transmit Queue Overflow errors, the value of Max Packets on S/W Transmit Queue will
correspond to the transmit queue limit for this adapter (xmt_que_size).
Note: These values can represent the hardware queue if the adapter does not support a software transmit
queue. If there are transmit-queue overflows, then increased the hardware or software queue limits
for the driver.
If there are not enough receive resources, this would be indicated by Packets Dropped: and depending on
the adapter type, would be indicated by Out of Rcv Buffers or No Resource Errors: or some similar
counter.
The elapsed time displays the real-time period that has elapsed since the last time the statistics were
reset. To reset the statistics, use the entstat -r adapter_name command.
Similar output can be displayed for Token-Ring, FDDI, and ATM interfaces using the tokstat, fddistat, and
atmstat commands.
The tokstat Command
The tokstat command displays the statistics gathered by the specified Token-Ring device driver. The user
can optionally specify that the device-specific statistics be displayed in addition to the device driver
statistics. If no flags are specified, only the device driver statistics are displayed.
This command is also invoked when the netstat command is run with the -v flag. The netstat command
does not issue any tokstat command flags.
The output produced by the tokstat tok0 command and the problem determination are similar to that
described in The entstat Command.
Chapter 10. Monitoring and Tuning Communications I/O Use
255
The fddistat Command
The fddistat command displays the statistics gathered by the specified FDDI device driver. The user can
optionally specify that the device-specific statistics be displayed in addition to the device driver statistics. If
no flags are specified, only the device driver statistics are displayed.
This command is also invoked when the netstat command is run with the -v flag. The netstat command
does not issue any fddistat command flags.
The output produced by the fddistat fddi0 command and the problem determination are similar to that
described in The entstat Command.
The atmstat Command
The atmstat command displays the statistics gathered by the specified ATM device driver. The user can
optionally specify that the device-specific statistics be displayed in addition to the device driver statistics. If
no flags are specified, only the device driver statistics are displayed.
The output produced by the atmstat atm0 command and the problem determination are similar to that
described in The entstat Command.
The no Command
Use the no command to display current network values and to change options.
-a
Prints all options and current values (example: no -a)
-d
Sets options back to default (example: no -d thewall)
-o
option=NewValue (example: no -o thewall=16384)
For a listing of all attributes for the no command, see Network Option Tunable Parameters.
Some network attributes are run-time attributes that can be changed at any time. Others are load-time
attributes that must be set before the netinet kernel extension is loaded.
Note: When the no command is used to change parameters, the change is in effect only until the next
system boot. At that point, all parameters are initially reset to their defaults. To make the change
permanent, put the appropriate no command in the /etc/rc.net file.
If your system uses Berkeley-style network configuration, set the attributes near the top of the
/etc/rc.bsdnet file. If you use an SP system, edit the tuning.cust file as documented in the RS/6000 SP:
Installation and Relocation manual.
Note: The no command performs no-range checking. If it is used incorrectly, the no command can cause
your system to become inoperable.
The following tuning sections discuss some of the no command attributes and how to adjust them.
Tuning TCP and UDP Performance
The optimal settings of the tunable communications parameters vary with the type of LAN, as well as with
the communications-I/O characteristics of the predominant system and application programs. The following
sections describe the global principles of communications tuning, followed by specific recommendations for
the different types of LAN.
Overall Recommendations
You can choose to tune primarily either for maximum throughput or for minimum memory use. Some
recommendations apply to one or the other; some apply to both. Recommended application block sizes for
different adapter devices are as follows:
256
Performance Management Guide
Device Name
Application Block Size
Ethernet
Multiples of 4096
Token-Ring (4 Mb)
Multiples of 4096
Token-Ring (16 Mb)
Multiples of 4096
FDDI (tcp)
Multiples of 4096
SOCC (tcp)
28672 bytes
HIPPI
65536 bytes
ATM
Multiples of 4096
Maximizing Throughput
Use the following recommendations to tune for maximum throughput:
Request-Response Protocols: Follow these guidelines:
v For maximum number of transactions per second, use the smallest feasible messages.
v For maximum bytes per second, use messages that are at least 1000 bytes and equal to or just less
than a multiple of 4096 bytes.
v If the requests and responses are fixed-size and fit into one datagram, use UDP.
– If possible, make the write sizes equal to the following:
(a multiple of the MTU size minus 28 bytes to allow for standard IP and UDP headers).
– In general, it is more efficient for the application to write large messages and have them fragmented
and reassembled by IP, than to have the application write multiple times.
– Whenever possible, use the connect() subroutine to associate an address with the UDP socket. This
may not be possible on a server that is communicating with a number of clients through a single
socket.
v If the requests or responses are variable-size, use TCP with the TCP_NODELAY option. Measurements
indicate that the overhead of TCP compared with UDP is negligible, especially if optimum write sizes
are used.
– To avoid data copies in the kernel, make write sizes greater than 512 bytes.
– Make writes equal to or slightly less than, a multiple of MTU size. This will avoid the sending of a
segment (packet) with just a few bytes in it.
Streaming: Follow these guidelines:
v TCP provides higher throughput than UDP and ensures reliable delivery.
v Writes should be in multiples of 16384 bytes. If possible, writes should be the size of the MSS (see
Tuning TCP Maximum Segment Size).
Minimizing Memory
Use the following recommendations to tune for minimizing memory usage:
v If your traffic is predominantly local, use the largest MTU size that is supported by your LAN type. This
minimizes the fragmentation of packets exchanged by local systems. The offsetting cost is
fragmentation in gateways that connect your LAN to other LANs with smaller MTUs (see Tuning TCP
Maximum Segment Size).
v Whenever possible, application programs should read and write in quantities of either:
– Less than or equal to 512 bytes
OR
– Slightly less than or equal to 4096 bytes (or multiples thereof)
Chapter 10. Monitoring and Tuning Communications I/O Use
257
If the applications were using TCP, both time and memory would be wasted. TCP tries to form outbound
data into MTU-sized packets. If the MTU of the LAN were larger than 14976 bytes, TCP would put the
sending thread to sleep when the tcp_sendspace limit was reached.To force the data to be written, a
timeout ACK from the receiver would be required.
Adapter Transmit and Receive Queue Tuning
Most communication drivers provide a set of tunable parameters to control transmit and receive resources.
These parameters typically control the transmit queue and receive queue limits, but may also control the
number and size of buffers or other resources. These parameters limit the number of buffers or packets
that may be queued for transmit or limit the number of receive buffers that are available for receiving
packets. These parameters can be tuned to ensure enough queueing at the adapter level to handle the
peak loads generated by the system or the network.
Following are some general guidelines:
v Tune transmit queues when the CPU is faster than the network (more common on multi-processor
systems where many CPUs are transmitting to a single adapter).
v Tune transmit queues when socket buffer sizes are large.
v Tune receive queues when there is very bursty traffic.
v Tune transmit and receive queues when there is high rate of small-sized packets.
Transmit Queues
For transmit, the device drivers may provide a transmit queue limit. There may be both hardware queue
and software queue limits, depending on the driver and adapter. Some drivers have only a hardware
queue; some have both hardware and software queues. Some drivers internally control the hardware
queue and only allow the software queue limits to be modified. Generally, the device driver will queue a
transmit packet directly to the adapter hardware queue. If the system CPU is fast relative to the speed of
the network, or on an SMP system, the system may produce transmit packets faster than they can be
transmitted on the network. This will cause the hardware queue to fill. After the hardware queue is full,
some drivers provide a software queue and they will then queue to the software queue. If the software
transmit queue limit is reached, then the transmit packets are discarded. This can affect performance
because the upper-level protocols must then time out and retransmit the packet.
Prior to AIX 4.2.1, the upper limits on the transmit queues were in the range of 150 to 250, depending on
the specific adapter. The system default values were low, typically 30. With AIX 4.2.1 and later, the
transmit queue limits were increased on most of the device drivers to 2048 buffers. The default values
were also increased to 512 for most of these drivers. The default values were increased because the
faster CPUs and SMP systems can overrun the smaller queue limits.
Following are examples of MCA adapter transmit queue sizes:
MCA Adapter Type
Default
Range
Ethernet
512
20 - 2048
10/100 Ethernet
64
16,32,64,128,256
Token-Ring
99 or 512
32 - 2048
FDDI
512
3 - 2048
ATM / 155 ATM
512
0 - 2048
Following are examples of PCI adapter transmit queue sizes:
PCI Adapter Type
Default
Range
Ethernet
64
16 - 256
10/100 Ethernet
256, 512, or 2048
16 -16384
258
Performance Management Guide
Token-Ring
96, 512, or 2048
32 - 16384
FDDI
30 or 2048
3 - 16384
155 ATM
100 or 2048
0 - 16384
For adapters that provide hardware queue limits, changing these values will cause more real memory to
be consumed on receives because of the associated control blocks and buffers associated with them.
Therefore, raise these limits only if needed or for larger systems where the increase in memory use is
negligible. For the software transmit queue limits, increasing these limits does not increase memory usage.
It only allows packets to be queued that were already allocated by the higher layer protocols.
Receive Queues
Some adapters allow you to configure the number of resources used for receiving packets from the
network. This might include the number of receive buffers (and even their size) or may be a receive queue
parameter (which indirectly controls the number of receive buffers).
The receive resources may need to be increased to handle peak bursts on the network. The network
interface device driver places incoming packets on a receive queue. If the receive queue is full, packets
are dropped and lost, resulting in the sender needing to retransmit. The receive queue is tunable using the
SMIT or chdev commands (see How to Change the Parameters). The maximum queue size is specified to
each type of communication adapter (see Tuning MCA and PCI Adapters).
For the Micro Channel adapters and the PCI adapters, receive queue parameters typically control the
number of receive buffers that are provided to the adapter for receiving input packets.
Device-Specific Buffers
AIX 4.1.4 and later support device-specific mbufs. This allows a driver to allocate its own private set of
buffers and have them pre-setup for Direct Memory Access (DMA). This can provide additional
performance because the overhead to set up the DMA mapping is done one time. Also, the adapter can
allocate buffer sizes that are best suited to its MTU size. For example, ATM, High Performance Parallel
Interface (HIPPI), and the SP switch support a 64 K MTU (packet) size. The maximum system mbuf size
is 16 KB. By allowing the adapter to have 64 KB buffers, large 64 K writes from applications can be copied
directly into the 64 KB buffers owned by the adapter, instead of copying them into multiple 16 K buffers
(which has more overhead to allocate and free the extra buffers).
The adapters that support Device Specific mbufs are:
v MCA ATM
v MCA HIPPI
v Various SP high speed switch adapters
Device-specific buffers add an extra layer of complexity for the system administrator. The system
administrator must use device-specific commands to view the statistics relating to the adapter’s buffers
and then change the adapter’s parameters as necessary. If the statistics indicate that packets were
discarded because not enough buffer resources were available, then those buffer sizes need to be
increased.
Due to differences between drivers and the utilities used to alter these parameters, the utilities and
parameters are not fully described here. The MCA ATM parameters are listed in Micro Channel Adapter
(MCA). Use the atmstat -d atm0 command to view the ATM statistics (substitute your ATM interface
number as needed).
When to Increase the Receive/Transmit Queue Parameters
Following are some guidelines to help you determine when to increase the receive/transmit queue
parameters:
Chapter 10. Monitoring and Tuning Communications I/O Use
259
1. When the CPU is much faster than the network and multiple applications may be using the same
network. This would be common on a larger multi-processor system (SMP).
2. When running with large values for tcp_sendspace or tcp_recvspace as set in the no options or
running applications that might use system calls to increase the TCP send and receive socket buffer
space. These large values can cause the CPU to send down large numbers of packets to the adapter,
which will need to be queued. Procedures are similar for udp_sendspace and udp_recvspace for
UDP applications.
3. When there is very bursty traffic.
4. A high-traffic load of small packets can consume more resources than a high traffic load of large
buffers. Because large buffers take more time to send on the network. The packet rate will therefore be
slower for larger packets.
Commands to Query and Change the Queue Parameters
Several status utilities can be used to show the transmit queue high-water limits and number of queue
overflows. You can use the command netstat -v, or go directly to the adapter statistics utilities (entstat for
Ethernet, tokstat for Token-Ring, fddistat for FDDI, atmstat for ATM, and so on).
For an entstat example output, see The entstat Command. Another method is to use the netstat -i utility.
If it shows non-zero counts in the Oerrs column for an interface, then this is typically the result of output
queue overflows.
How to See the Settings
You can use the lsattr -E -l adapter-name command or you can use the SMIT command (smitty
commodev) to show the adapter configuration.
Different adapters have different names for these variables. For example, they may be named
sw_txq_size, tx_que_size, or xmt_que_size for the transmit queue parameter. The receive queue size
and receive buffer pool parameters may be named rec_que_size, rx_que_size, or rv_buf4k_min for
example.
Following is the output of a lsattr -E -l atm0 command on an IBM PCI 155 Mbs ATM adapter. This output
shows the sw_txq_size is set to 250 and the rv_buf4K_min receive buffers set to x30.
# lsattr -E -l atm0
dma_mem
0x400000
N/A
False
regmem
0x1ff88000 Bus Memory address of Adapter Registers
False
virtmem
0x1ff90000 Bus Memory address of Adapter Virtual Memory False
busintr
3
Bus Interrupt Level
False
intr_priority 3
Interrupt Priority
False
use_alt_addr
no
Enable ALTERNATE ATM MAC address
True
alt_addr
0x0
ALTERNATE ATM MAC address (12 hex digits) True
sw_txq_size 250
Software Transmit Queue size
True
max_vc
1024
Maximum Number of VCs Needed
True
min_vc
32
Minimum Guaranteed VCs Supported
True
rv_buf4k_min 0x30
Minimum 4K-byte pre-mapped receive buffers
True
interface_type 0
Sonet or SDH interface
True
adapter_clock 1
Provide SONET Clock
True
uni_vers
auto_detect N/A
True
Following is an example of a Micro Channel 10/100 Ethernet settings using the lsattr -E -l ent0 command.
This output shows the tx_que_size and rx_que_size both set to 256.
# lsattr -E -l ent0
bus_intr_lvl 11
intr_priority 3
dma_bus_mem
0x7a0000
bus_io_addr
0x2000
dma_lvl
7
tx_que_size 256
rx_que_size 256
use_alt_addr no
260
Bus interrupt level
False
Interrupt priority
False
Address of bus memory used for DMA False
Bus I/O address
False
DMA arbitration level
False
TRANSMIT queue size
True
RECEIVE queue size
True
Enable ALTERNATE ETHERNET address True
Performance Management Guide
alt_addr
media_speed
ip_gap
0x
ALTERNATE ETHERNET address
100_Full_Duplex Media Speed
96
Inter-Packet Gap
True
True
True
How to Change the Parameters
The following are some of the parameters that are user-configurable:
v Transmit Queue Size (tx_que_size)
The device driver supports a user-configurable transmit queue. This is the queue the adapter uses (not
an extension of the adapter’s queue). It is configurable among the values of 16, 32, 64, 128 and 256,
with a default of 256.
Because of the configurable size of the adapter’s hardware queue, the driver does not support a
software queue.
v Receive Queue Size (rx_que_size)
The device driver supports a user-configurable receive queue. This is the queue the adapter uses (not
an extension of the adapter’s queue). It is configurable among the values of 16, 32, 64, 128 and 256,
with a default of 256.
v Receive Buffer Pool Size (rxbuf_pool_size)
The device driver supports a user-configurable receive buffer pool size. The buffer is the number of
preallocated mbufs for receiving packets. The minimum size of the buffer is the receive queue size and
the maximum is 2 KB (the default value of 384).
v Media Speed (media_speed)
The device driver supports speeds of 10 (10 Mbps, half-duplex), 20 (10 Mbps, full-duplex), 100 (100
Mbps, half-duplex), 200 (100 Mbps, full-duplex), and auto-negotiate on twisted pair. On the AUI port, the
device driver supports speeds of 10 (10 Mbps, half-duplex) and 20 (10 Mbps, full-duplex). The bayonet
Niell-Concelman (BNC) port will only support 10 (10 Mbps, half-duplex). This attribute is
user-configurable, with a default of auto-negotiate on twisted pair.
v Enable Alternate Address (use_alt_addr)
The device driver supports a configuration option to toggle use of an alternate network address. The
values are yes and no, with a default of no. When this value is set to yes, the alt_addr parameter
defines the address.
v Alternate Network Address (alt_addr)
For the network address, the device driver accepts the adapter’s hardware address or a configured
alternate network address. When the use_alt_addr configuration option is set to yes, this alternate
address is used. Any valid individual address can be used, but a multicast address cannot be defined
as a network address.
v Inter-Packet Gap (ip_gap)
The inter-packet gap (IPG) bit rate setting controls the aggressiveness of the adapter on the network. A
smaller number will increase the aggressiveness of the adapter, while a larger number will decrease the
aggressiveness (and increase the fairness) of the adapter. If the adapter statistics show a large number
of collisions and deferrals, increase this number. Valid values range from 96 to 252, in increments of 4.
The default value of 96 results in IPG of 9.6 microseconds for 10 Mb and 0.96 microseconds for 100
Mb media speeds. Each unit of bit rate introduces an IPG of 100 ns at 10 Mb and 10 ns at 100 Mb
media speed.
To change any of the parameter values, do the following:
1. Detach the interface by running the following command:
# ifconfig en0 detach
where en0 represents the adapter name.
2. Use SMIT to display the adapter settings. Select Devices -> Communications -> adapter type ->
Change/Show...
Chapter 10. Monitoring and Tuning Communications I/O Use
261
3. Move the cursor to the field you want to change, and press F4 to see the minimum and maximum
ranges for the field (or the specific set of sizes that are supported).
4. Select the appropriate size, and press Enter to update the ODM database.
5. Reattach the adapter by running the following command:
# ifconfig en0 hosthame up
An alternative method to change these parameter values is to run the following command:
# chdev -l [ifname] -a [attribute-name]=newvalue
For example, to change the above tx_que_size on en0 to 128, use the following sequence of commands.
Note that this driver only supports four different sizes, so it is better to use the SMIT command to see
these values.
# ifconfig en0 detach
# chdev -l ent0 -a tx_que_size=128
# ifconfig en0 hostname up
Tuning MCA and PCI Adapters
The following information is provided to document the various adapter-tuning parameters. These
parameters and values are for AIX 4.3.1 and are provided to aid you in understanding the various tuning
parameters, or when a system is not available to view the parameters.
These parameter names, defaults, and range values were obtained from the ODM database. The
comment field was obtained from the lsattr -E -l interface-name command.
The Notes field provides additional comments.
Micro Channel Adapter (MCA):
Feature Code: 2980
Ethernet High-Performance LAN Adapter (8ef5)
Parameter
Default Range
Comment
------------- -------- -------- --------------------------xmt_que_size
512
20-2048
TRANSMIT queue size
rec_que_size
30
20-150
RECEIVE queue size
rec_pool_size 37
16-64
RECEIVE buffer pool size
Notes
----------------SW TX queue
See Note 1
On Adapter
Feature Code: 2992
Ethernet High-Performance LAN Adapter (8f95)
Parameter
Default
Range
Comment
Notes
------------- --------- -------- ------------------- ---------xmt_que_size
512
20-2048 TRANSMIT queue size SW queue
Feature Code: 2994
IBM 10/100 Mbps Ethernet TX MCA Adapter (8f62)
Parameter
------------tx_que_size
rx_que_size
Default
-------64
32
Range
---------------16,32,64,128,256
16,32,64,128,256
Comment
Notes
--------------------- ----------TRANSMIT queue size
HW queue
RECEIVE queue size
HW queue
Feature Code: 2970
Token-Ring High-Performance Adapter (8fc8)
Parameter
262
Default
Range
Performance Management Guide
Comment
Notes
------------- -------- -------- --------------------xmt_que_size 99
32-2048
TRANSMIT queue size
rec_que_size 30
20-150
RECEIVE queue size
-----------SW queue
See Note 1
Feature Code: 2972
Token-Ring High-Performance Adapter (8fa2)
Parameter
------------xmt_que_size
rx_que_size
Default
-------512
32
Range
-------32-2048
32-160
Comment
Notes
---------------------------- ---------TRANSMIT queue size
SW queue
HARDWARE RECEIVE queue size HW queue
Feature Code: 2727
FDDI Primary Card, Single Ring Fiber
Parameter
------------tx_que_size
rcv_que_size
Default
-------512
30
Range
Comment
Notes
-------- ------------------------------ -------------------3-2048
Transmit Queue Size (in mbufs)
20-150
Receive Queue
See Note 1
Feature Code: 2984
100 Mbps ATM Fiber Adapter (8f7f)
Parameter
Default Range Comment
Notes
--------------- ----- --------- -------------------------- ----sw_queue
512 0-2048
Software transmit queue len. SW Queue
dma_bus_width
0x1000000 0x800000-0x40000000,0x100000
Amount of memory to map for DMA See Note 3
max_sml_bufs
50
40-400
Maximum Small ATM mbufs
Max 256 byte buffers
max_med_bufs
100 40-1000
Maximum Medium ATM mbufs Max 4KB buffers
max_lrg_bufs
300 75-1000
Maximum Large ATM mbufs
Max 8KB buffers See Note 2
max_hug_bufs
50
0-400
Maximum Huge ATM mbufs
Max 16KB buffers
max_spec_bufs
4
0-400
Maximum ATM MTB mbufs Max of max_spec_buf size
spec_buf_size
64
32-1024 Max Transmit Block (MTB) size (kbytes)
sml_highwater
20
10-200
Minimum Small ATM mbufs
Min 256 byte buffers
med_highwater
30
20-300
Minimum Medium ATM mbufs Min 4KB buffers
lrg_highwater
70
65-400
Minimum Large ATM mbufs
Min 8KB buffers
hug_highwater
10
4-300
Minimum Huge ATM mbufs
Min 16KB buffers
spec_highwater 20
0-300
Minimum ATM MTB mbufs
Min 64KB buffers
best_peak_rate 1500 1-155000 Virtual Circuit Peak Segamentation Rate
Feature Code: 2989
155 Mbps ATM Fiber Adapter (8f67)
Parameter
Default
Range
------------- -------- -------(same as ATM 100 adapter above)
Comment
Notes
---------- -------
Notes:
1. A software receive queue is provided only for compatibility with operating system version 3.2.x
applications that use the network device driver interface to read packets directly from the driver. This
queue limits how many input packets are queued for these applications to receive. This parameter is
defined only if bos.compat is installed.
This queue is not used by the typical TCP/IP stack.
Chapter 10. Monitoring and Tuning Communications I/O Use
263
2. MCA ATM: The receive side also uses the large (8 K) buffers. The receive logic only uses the 8 K
buffers, so if this size runs low, it affects receive performance.
The other buffers sizes are only for transmit buffers.
3. MCA ATM: If you need to increase the total number of buffers, you may need to change the
dma_bus_width (= 0x1000000) parameter. DMA bus memory width controls the total amount of
memory used for ATM buffers. Increase this parameter if you get an error while you are increasing the
maximum buffers or high-water limits.
PCI Adapters:
Feature Code 2985
IBM PCI Ethernet Adapter (22100020)
Parameter
------------tx_que_size
rx_que_size
Default
-------64
32
Range
----------------16,32,64,128,256
16,32,64,128,256
Comment
------------------TRANSMIT queue size
RECEIVE queue size
Notes
--------HW Queues
HW Queues
Featue Code 2968
IBM 10/100 Mbps Ethernet PCI Adapter (23100020)
Parameter
---------------tx_que_size
rx_que_size
rxbuf_pool_size
Default
------256
256
384
Range
---------------16,32,64,128,256
16,32,64,128,256
16-2048
Comment
--------------------TRANSMIT queue size
RECEIVE queue size
# buffers in receive
buffer pool
Notes
-------------------HW Queue Note 1
HW Queue Note 2
Dedicat. receive
buffers Note 3
Feature Code: 2969
Gigabit Ethernet-SX PCI Adapter (14100401)
Parameter
------------tx_que_size
rx_que_size
receive_proc
Default
------512
512
6
Range
-------512-2048
512
0-128
Comment
----------------------------------Software Transmit Queueu size
Receive queue size
Minimum Receive Buffer descriptiors
Notes
--------SW Queue
HW Queue
Feature Code: 2986
3Com 3C905-TX-IBM Fast EtherLink XL NIC
Parameter
-------------tx_wait_q_size
rx_wait_q_size
Default
-------32
32
Range
-----4-128
4-128
Comment
---------------------------Driver TX Waiting Queue Size
Driver RX Waiting Queue Size
Notes
---------HW Queues
HW Queues
Feature Code: 2742
SysKonnect PCI FDDI Adapter (48110040)
Parameter
------------tx_queue_size
RX_buffer_cnt
Default
-------30
42
Range
-------3-250
1-128
Comment
------------------Transmit Queue Size
Receive frame count
Notes
--------------SW Queue
Rcv buffer pool
Feature Code: 2979
IBM PCI Tokenring Adapter (14101800)
Parameter
------------xmt_que_size
rx_que_size
264
Default
-------96
32
Range
------32-2048
32-160
Comment
--------------------------TRANSMIT queue size
HARDWARE RECEIVE queue size
Performance Management Guide
Notes
-------SW Queue
HW queue
Feature Code: 2979
IBM PCI Tokenring Adapter (14103e00)
Parameter
------------xmt_que_size
rx_que_size
Default
-------512
64
Range
-------32-2048
32-512
Comment
-------------------TRANSMIT queue size
RECEIVE queue size
Notes
-------SW Queue
HW Queue
Feature Code: 2988
IBM PCI 155 Mbps ATM Adapter (14107c00)
Parameter
------------sw_txq_size
rv_buf4k_min
Default
--------100
48 (0x30)
Range
-----------0-4096
0-512 (x200)
Comment
Notes
-------------------------------- -------Software Transmit Queue size
SW Queue
Minimum 4K-byte pre-mapped receive buffers
Notes on the IBM 10/100 Mbps Ethernet PCI Adapter:
1. Prior to AIX 4.3.2, default tx_queue_size was 64.
2. Prior to AIX 4.3.2, default rx_que_size was 32.
3. In AIX 4.3.2 and later, the driver added a new parameter to control the number of buffers dedicated to
receiving packets.
Enabling Thread Usage on LAN Adapters (dog threads)
Drivers, by default, call IP directly, which calls up the protocol stack to the socket level while running on
the interrupt level. This minimizes instruction path length, but increases the interrupt hold time. On an SMP
system, a single CPU can become the bottleneck for receiving packets from a fast adapter. By enabling
the dog threads, the driver queues the incoming packet to the thread and the thread handles calling IP,
TCP, and the socket code. The thread can run on other CPUs which may be idle. Enabling the dog
threads can increase capacity of the system in some cases.
Note: This feature is not supported on uniprocessors, because it would only add path length and slow
down performance.
This is a feature for the input side (receive) of LAN adapters. It can be configured at the interface level
with the ifconfig command (ifconfig interface thread or ifconfig interface hostname up thread).
To disable the feature, use the ifconfig interface -thread command.
Guidelines when considering using dog threads are as follows:
v More CPUs than adapters need to be installed. Typically, at least two times more CPUs than adapters
are recommended.
v Systems with faster CPUs benefit less. Machines with slower CPU speed may be helped the most.
v This feature is most likely to enhance performance when there is high input packet rate. It will enhance
performance more on MTU 1500 compared to MTU 9000 (jumbo frames) on Gigabit as the packet rate
will be higher on small MTU networks.
The dog threads run best when they find more work on their queue and do not have to go back to sleep
(waiting for input). This saves the overhead of the driver waking up the thread and the system
dispatching the thread.
v The dog threads can also reduce the amount of time a specific CPU spends with interrupts masked.
This can release a CPU to resume typical user-level work sooner.
v The dog threads can also reduce performance by about 10 percent if the packet rate is not fast enough
to allow the thread to keep running. The 10 percent is an average amount of increased CPU overhead
needed to schedule and dispatch the threads.
Chapter 10. Monitoring and Tuning Communications I/O Use
265
Tuning TCP Maximum Segment Size
The TCP protocol includes a mechanism for both ends of a connection to negotiate the maximum segment
size (MSS) to be used over the connection. Each end uses the OPTIONS field in the TCP header to
advertise a proposed MSS. The MSS that is chosen is the smaller of the values provided by the two ends.
The purpose of this negotiation is to avoid the delays and throughput reductions caused by fragmentation
of the packets when they pass through routers or gateways and reassembly at the destination host.
The value of MSS advertised by the TCP software during connection setup depends on whether the other
end is a local system on the same physical network (that is, the systems have the same network number)
or whether it is on a different (remote) network.
Local Network
If the other end of the connection is local, the MSS advertised by TCP is based on the MTU (maximum
transfer unit) of the local network interface, as follows:
TCP MSS = MTU - TCP header size - IP header size.
Because this is the largest possible MSS that can be accommodated without IP fragmentation, this value
is inherently optimal, so no MSS-tuning is required for local networks.
Remote Network
When the other end of the connection is on a remote network, this operating system’s TCP defaults to
advertising an MSS of 512 bytes. This conservative value is based on a requirement that all IP routers
support an MTU of at least 576 bytes.
The optimal MSS for remote networks is based on the smallest MTU of the intervening networks in the
route between source and destination. In general, this is a dynamic quantity and could only be ascertained
by some form of path MTU discovery. The TCP protocol, by default, does not provide a mechanism for
doing path MTU discovery, which is why a conservative MSS value is the default. However, it is possible
to enable the TCP PMTU discovery by using the following command:
# no -o tcp_pmtu_discover=1
MTU path discovery was added to AIX 4.2.1, but the default is off. With AIX 4.3.3 and later, the default is
on.
A typical side effect of this setting is to see the routing table increasing (one more entry per each active
TCP connection). The no option route_expire should be set to a non-zero value, in order to have any
unused cached route entry removed from the table, after route_expire time of inactivity.
While the conservative default is appropriate in the general Internet, it can be unnecessarily restrictive for
private Intranets within an administrative domain. In such an environment, MTU sizes of the component
physical networks are known, and the minimum MTU and optimal MSS can be determined by the
administrator. The operating system provides several ways in which TCP can be persuaded to use this
optimal MSS. Both source and destination hosts must support these features. In a heterogeneous,
multi-vendor environment, the availability of the feature on both systems can determine the choice of
solution.
Static Routes: The default MSS of 512 can be overridden by specifying a static route to a specific
remote network. Use the -mtu option of the route command to specify the MTU to that network. In this
case, you would specify the actual minimum MTU of the route, rather than calculating an MSS value.
In a small, stable environment, this method allows precise control of MSS on a network-by-network basis.
The disadvantages of this approach are as follows:
v It does not work with dynamic routing.
266
Performance Management Guide
v It becomes impractical when the number of remote networks increases.
v Static routes must be set at both ends to ensure that both ends negotiate with a larger-than-default
MSS.
When to Use the tcp_mssdflt Option of the no Command: This parameter is used to set the
maximum packet size for communication with remote networks. However, only one value can be set even
if there are several adapters with different MTU sizes. The default value of 512 that TCP uses for remote
networks can be changed via the no command. This change is a systemwide change.
To override the MSS default specify a value that is the minimum MTU value less 40 to allow for the typical
length of the TCP and IP headers.
The size is the same as the MTU for communication across a local network with one exception: the
tcp_mssdflt size is only for the size of the data in a packet. Reduce the tcp_mssdflt for the size of any
headers so that you send full packets instead of a full packet and a fragment. Calculate this as follows:
MTU of interface - TCP header size - IP header size - rfc1323 header size
which is:
MTU - 20 - 20 - 12, or MTU - 52
Limiting data to (MTU - 52) bytes ensures that, where possible, only full packets will be sent.
In an environment with a larger-than-default MTU, this method has the advantage in that the MSS does
not need to be set on a per-network basis. The disadvantages are as follows:
v Increasing the default can lead to IP router fragmentation if the destination is on a network that is truly
remote and the MTUs of the intervening networks are not known.
v The tcp_mssdflt parameter must be set to the same value on the destination host.
Subnetting and the subnetsarelocal Option of the no Command: Several physical networks can be
made to share the same network number by subnetting. The no option subnetsarelocal specifies, on a
systemwide basis, whether subnets are to be considered local or remote networks. With the command no
-o subnetsarelocal=1 (the default), Host A on subnet 1 considers Host B on subnet 2 to be on the same
physical network.
The consequence is that when Host A and Host B establish a connection, they negotiate the MSS
assuming they are on the same network. Each host advertises an MSS based on the MTU of its network
interface, usually leading to an optimal MSS being chosen.
The advantages to this approach are as follows:
v It does not require any static bindings; MSS is automatically negotiated.
v It does not disable or override the TCP MSS negotiation, so that small differences in the MTU between
adjacent subnets can be handled appropriately.
The disadvantages to this approach are as follows:
v Potential IP router fragmentation when two high-MTU networks are linked through a lower-MTU
network. The following figure illustrates this problem.
Chapter 10. Monitoring and Tuning Communications I/O Use
267
Figure 26. Inter-Subnet Fragmentation. This illustration show a data path from Host A, through an FDDI with an
MTU=4352, through Router 1, to the Ethernet with an MTU=1500. From there it goes to Router 2 and another FDDI
with an MTU=4352 and out to Host B. An explanation of how fragmentation occurs in this example is described in the
text immediately following the illustration.
v In this scenario, Hosts A and B would establish a connection based on a common MTU of 4352. A
packet going from A to B would be fragmented by Router 1 and defragmented by Router 2. The reverse
would occur going from B to A.
v Source and destination must both consider subnets to be local.
UDP Socket Buffer Tuning
UDP is a datagram protocol. Being a datagram, the entire message (datagram) must be copied into the
kernel on a send operation as one atomic operation. The maximum amount of data that UDP can send at
one time is limited by the size of the memory buffer assigned to a specific UDP socket, and the maximum
packet size that the IP layer can handle in each packet.
udp_sendspace
Set this parameter to 65536, because any value greater than 65536 is ineffective. Because UDP transmits
a packet as soon as it gets any data, and because IP has an upper limit of 65536 bytes per packet,
anything beyond 65536 runs the small risk of being discarded by IP. The IP protocol will fragment the
datagram into smaller packets if needed, based on the MTU size of the interface the packet will be sent
on. For example, sending an 8 K datagram, IP would fragment this into 1500 byte packets if sent over
Ethernet. Because UDP does not implement any flow control, all packets given to UPD are passed to IP
(where they may be fragmented) and then placed directly on the device drivers transmit queue.
udp_recvspace
On the receive side, the incoming datagram (or fragment if the datagram is larger than the MTU size) will
first be received into a buffer by the device driver. This will typically go into a buffer that is large enough to
hold the largest possible packet from this device.
The setting of udp_recvspace is harder to compute because it varies by network adapter type, UDP
sizes, and number of datagrams queued to the socket. Set the udp_recvspace larger rather than smaller,
because packets will be discarded if it is too small.
For example, Ethernet might use 2 K receive buffers. Even if the incoming packet is maximum MTU size
of 1500 bytes, it will only use 73 percent of the buffer. IP will queue the incoming fragments until a full
UDP datagram is received. It will then be passed to UDP. UDP will put the incoming datagram on the
receivers socket. However, if the total buffer space in use on this socket exceeds udp_recvspace, then
the entire datagram will be discarded. This is indicated in the output of the netstat -s command as
dropped due to full socket buffers errors.
Because the communication subsystem accounts for buffers used, and not the contents of the buffers, you
must account for this when setting udp_recvspace. In the above example, the 8 K datagram would be
fragmented into 6 packets which would use 6 receive buffers. These will be 2048 byte buffers for Ethernet.
So, the total amount of socket buffer consumed by this one 8 K datagram is as follows:
6*2048=12,288 bytes
Thus, you can see that the udp_recvspace must be adjusted higher depending on how efficient the
incoming buffering is. This will vary by datagram size and by device driver. Sending a 64 byte datagram
would consume a 2 K buffer for each 64 byte datagram.
268
Performance Management Guide
Then, you must account for the number of datagrams that may be queued onto this one socket. For
example, NFS server receives UDP packets at one well-known socket from all clients. If the queue depth
of this socket could be 30 packets, then you would use 30 * 12,288 = 368,640 for the udp_recvspace if
NFS is using 8 K datagrams. NFS Version 3 allows up to 32K datagrams.
A suggested starting value for udp_recvspace is 10 times the value of udp_sendspace, because UDP
may not be able to pass a packet to the application before another one arrives. Also, several nodes can
send to one node at the same time. To provide some staging space, this size is set to allow 10 packets to
be staged before subsequent packets are discarded. For large parallel applications using UDP, the value
may have to be increased.
Note: The value of sb_max, which specifies the maximum socket buffer size for any socket buffer, should
be at least twice the size of the largest of the UDP and TCP send and receive buffers.
TCP Socket Buffer Tuning
The following table shows some suggested minimum sizes for socket buffers based on the type of adapter
and the MTU size. Note that setting these values too high can hurt performance. In addition, there is the
Nagle Black hole problem that can cause very low throughput for large MTU adapters, such as ATM if the
TCP send and receive space parameters are not chosen correctly.
Device
Speed
MTU
tcp_sendspace
tcp_recvspace
sb_max
rfc1323
Ethernet
10 Mbit
1500
16384
16384
32768
0
Ethernet
100 Mbit
1500
65536
65536
65536
0
Ethernet
Gigabit
1500
131072
65536
131072
0
Ethernet
Gigabit
9000
131072
65535 (Note 1)
262144
0
Ethernet
Gigabit
9000
262144
131072 (Note 1)
524288
1
ATM
155 Mbit
1500
16384
16384
131072
0
ATM
155 Mbit
9180
65535
65535 (Note 2)
131072
0
ATM
155 Mbit
65527
655360
655360 (Note 3)
1310720
1
FDDI
100 Mbit
4352
45056
45056
90012
0
Notes:
1. In the case of Gigabit Ethernet with a 9000 byte MTU, the performance was the same for both given
sets of buffer sizes.
2. Certain combinations of TCP send and receive space will result in very low throughput (1 Mbit or less).
This problem is described in detail in How a Large ATM MTU Causes Deadlocks in TCP Data
Transfers, IEEE/ACM Transactions on Networking, Vol. 3, No.4 August 1995 and TCP Buffering and
Performance over an ATM Network, Internetworking: Research and Experience, Vol. 6 1-13, 1995.
To avoid this problem, set the tcp_sendspace to a minimum of 3 times the MTU size or equal or
larger than the receivers tcp_recvspace. For example, on ATM with MTU 9180, a tcp_sendspace of
16384 and a tcp_recvspace of 32768 or 65536 resulted in very poor performance. However, setting
both to 65536 resulted in excellent performance. Also, setting both equal to 16384 resulted in
acceptable performance (the equal or larger rule).
3. TCP has only a 16-bit value to use for its window size. This translates to a maximum window size of
65536 bytes. For adapters that have large MTU sizes (32 K or 64 K for example), TCP streaming
performance may be very poor. For example, on a device with a 64 K MTU size, and with a
tcp_recvspace set to 64 K, TCP can only send one packet and then its window will close. It must wait
for an ACK back from the receiver before it can send again. This problem can be solved in two ways:
Chapter 10. Monitoring and Tuning Communications I/O Use
269
v One option is to enable rfc1323. This option enhances TCP and allows it to overcome the 16 bit
limit so that it can use a window size larger than 64 Kb. You can then set the tcp_recvspace to a
large value such as 10 times the MTU size which will allow TCP to stream data and give good
performance.
v The second option is to reduce the MTU size of the adapter. For example, use the command
ifconfig at0 mtu 16384 to set the ATM MTU size to 16 K. This will cause TCP to compute a smaller
MSS. With a 16 K MTU size, it could still send 4 packets for a 64 K window size.
Following are some general guidelines:
v Set the TCP send/recv space to at least 10 times the MTU size.
v MTU sizes above 16 K should use rfc1323=1 to allow larger TCP recvspace values.
v For high-speed adapters, larger TCP send/receive space values help performance.
v The window size is the receiver’s window size; therefore, rfc1323 affects only the receiver.
v For the Gigabit Ethernet adapter increase the tcp_sendspace. If the application gets blocked and is put
to sleep due to a small tcp_sendspace, there is too much latency on wake up and sending the packets
again to keep the adapter busy.
The ftp and rcp commands are examples of TCP applications that benefit from tuning the tcp_sendspace
and tcp_recvspace variables.
tcp_sendspace
TCP send buffer size can limit how much data the application can send before the application is put to
sleep. The TCP socket send buffer is used to buffer the application data in the kernel using mbufs/clusters
before it is sent beyond the socket and TCP layer. The default size of this buffer is specified by the
parameter tcp_sendspace, but you can use the setsockopt() subroutine to override it.
If the amount of data that the application wants to send is smaller than the send buffer size and also
smaller than the maximum segment size and if TCP_NODELAY is not set, then TCP will delay up to 200
ms, until enough data exists to fill the send buffer or the amount of data is greater than or equal to the
maximum segment size, before transmitting the packets.
If TCP_NODELAY is set, then the data is sent immediately (useful for request/response type of
applications). If the send buffer size is less than or equal to the maximum segment size (ATM and SP
switches can have 64 K MTUs), then the application’s data will be sent immediately and the application
must wait for an ACK before sending another packet (this prevents TCP streaming and could reduce
throughput).
Note: To maintain a steady stream of packets, increase the socket send buffer size so that it is greater
than the MTU (3-10 times the MTU size could be used as a starting point).
If an application does nonblocking I/O (specified O_NDELAY or O_NONBLOCK on the socket), then if the
send buffer fills up, the application will return with an EWOULDBLOCK/EAGAIN error rather than being put
to sleep. Applications must be coded to handle this error (suggested solution is to sleep for a short while
and try to send again).
When you are changing send/recv space values, in some cases you must stop/restart the inetd process
as follows:
# stopsrc -s inetd; startsrc -s inetd
tcp_recvspace
TCP receive-buffer size limits how much data the receiving system can buffer before the application reads
the data. The TCP receive buffer is used to accommodate incoming data. When the data is read by the
TCP layer, TCP can send back an acknowledgment (ACK) for that packet immediately or it can delay
before sending the ACK. Also, TCP tries to piggyback the ACK if a data packet was being sent back
anyway. If multiple packets are coming in and can be stored in the receive buffer, TCP can acknowledge
270
Performance Management Guide
all of these packets with one ACK. Along with the ACK, TCP returns a window advertisement to the
sending system telling it how much room remains in the receive buffer. If not enough room remains, the
sender will be blocked until the application has read the data. Smaller values will cause the sender to
block more. The size of the TCP receive buffer can be set using the setsockopt() subroutine or by the
tcp_recvspace parameter.
rfc1323
The TCP window size by default is limited to 65536 bytes (64 K) but can be set higher if rfc1323 is set to
1. If you are setting tcp_recvspace to greater than 65536, set rfc1323=1 on each side of the connection.
Without having rfc1323 set on both sides, the effective value for tcp_recvspace will be 65536.
If you are sending data through adapters that have large MTU sizes (32 K or 64 K for example), TCP
streaming performance may not be optimal because the packet or packets will be sent and the sender will
have to wait for an acknowledgment. By enabling the rfc1323 option using the command no -o rfc1323=1,
TCP’s window size can be set as high as 4 GB. However, on adapters that have 64 K or larger MTUs,
TCP streaming performance can be degraded if the receive buffer can only hold 64 K. If the receiving
machine does not support rfc1323, then reducing the MTU size is one way to enhance streaming
performance.
After setting the rfc1323 option to 1, you can increase the tcp_recvspace parameter to something much
larger, such as 10 times the size of the MTU.
sb_max
This parameter controls how much buffer space is consumed by buffers that are queued to a sender’s
socket or to a receiver’s socket. The system accounts for socket buffers used based on the size of the
buffer, not on the contents of the buffer.
If a device driver puts 100 bytes of data into a 2048-byte buffer, then the system considers 2048 bytes of
socket buffer space to be used. It is common for device drivers to receive buffers into a buffer that is large
enough to receive the adapters maximum size packet. This often results in wasted buffer space but it
would require more CPU cycles to copy the data to smaller buffers.
Because there are so many different network device drivers, increase the sb_max value much higher
rather than making it the same as the largest TCP or UDP socket buffer size parameters. After the total
number of mbufs/clusters on the socket reaches the sb_max limit, no additional buffers can be queued to
the socket until the application has read the data.
Note: When you are setting buffer size parameters to larger than 64 K, you must also increase the value
of sb_max, which specifies the maximum socket buffer size for any socket buffer.
One guideline would be to set it to twice as large as the largest TCP or UDP receive space.
Interface-Specific Network Options (ISNO)
In AIX 4.3.3, a feature called Interface-Specific Network Options (ISNO) was introduced that allows IP
network interfaces to be custom-tuned for the best performance. Values set for an individual interface take
precedence over the systemwide values set with the no command. The feature is enabled (the default) or
disabled for the whole system with the no command use_isno option. This single-point ISNO disable
option is included as a diagnostic tool to eliminate potential tuning errors if the system administrator needs
to isolate performance problems.
Programmers and performance analysts should note that the ISNO values will not show up in the socket
(meaning they cannot be read by the getsockopt() system call) until after the TCP connection is made.
The interface this socket will actually be using is not known until the connection is complete, so the socket
reflects the system defaults from the no command. After the connection is accepted, ISNO values are put
into the socket.
Chapter 10. Monitoring and Tuning Communications I/O Use
271
The following five parameters have been added for each supported network interface:
v rfc1323
v tcp_nodelay
v tcp_sendspace
v tcp_recvspace
v tcp_mssdflt
When set for a specific interface, these values override the corresponding no option values set for the
system. These parameters are available for all of the mainstream TCP/IP interfaces (Token-Ring, FDDI,
10/100 Ethernet, and Gigabit Ethernet), except the css# IP interface on the SP switch. As a simple
workaround, SP switch users can set the tuning options appropriate for the switch using the systemwide
no command, then use the ISNOs to set the values needed for the other system interfaces. ATM is
supported and works correctly with AIX 4.3.3 (a software update is needed) and later.
These options are set for the TCP/IP interface (such as en0 or tr0), and not the network adapter (ent0 or
tok0).
The five new ISNO parameters cannot be displayed or changed using SMIT. Following are commands that
can be used first to verify system and interface support and then to set and verify the new values.
v Make sure the use_isno option is enabled by using the following command:
# no -a | grep isno
use_isno = 1
v Make sure the interface supports the five new ISNOs by using the lsattr -El command:
# lsattr -E -l en0 -H
attribute
value description
:
rfc1323
N/A
tcp_nodelay
N/A
tcp_sendspace
N/A
tcp_recvspace
N/A
tcp_mssdflt
N/A
user_settable
True
True
True
True
True
v Set the interface-specific values, using either the ifconfig or chdev command. The ifconfig command
sets values temporarily (best used for testing). The chdev command alters the ODM, so custom values
return after system reboots.
For example, to set the tcp_recvspace and tcp_sendspace to 64K and enable tcp_nodelay, use one
of the following methods:
# ifconfig en0 tcp_recvspace 65536 tcp_sendspace 65536 tcp_nodelay 1
or
# chdev -l en0 -a tcp_recvspace=65536 -a tcp_sendspace=65536 -a tcp_nodelay=1
v Verify the settings using the ifconfig or lsattr command:
# ifconfig en0
en0: flags=e080863<UP,BROADCAST,NOTRAILERS,RUNNING,SIMPLEX,MULTICAST,GROUPRT,64BIT>
inet 9.19.161.100 netmask 0xffffff00 broadcast 9.19.161.255
tcp_sendspace 65536 tcp_recvspace 65536 tcp_nodelay 1
or
# lsattr -El en0
rfc1323
tcp_nodelay
1
tcp_sendspace 65536
tcp_recvspace 65536
tcp_mssdflt
272
N/A
N/A
N/A
N/A
N/A
Performance Management Guide
True
True
True
True
True
IP Protocol Performance Tuning Recommendations
At the IP layer, the only tunable parameter is ipqmaxlen, which controls the length of the IP input queue
discussed in IP Layer. In AIX Version 4, in general, interfaces do not do queuing. Packets can arrive very
quickly and overrun the IP input queue. You can use the netstat -s or netstat -p ip command to view an
overflow counter (ipintrq overflows).
If the number returned is greater than 0, overflows have occurred. Use the no command to set the
maximum length of this queue. For example:
# no -o ipqmaxlen=100
This example allows 100 packets to be queued up. The exact value to use is determined by the maximum
burst rate received. If this cannot be determined, using the number of overflows can help determine what
the increase should be. No additional memory is used by increasing the queue length. However, an
increase may result in more time spent in the off-level interrupt handler, because IP will have more
packets to process on its input queue. This could adversely affect processes needing CPU time. The
tradeoff is reduced packet-dropping versus CPU availability for other processing. It is best to increase
ipqmaxlen by moderate increments if the tradeoff is a concern in your environment.
Ethernet Performance Tuning Recommendations
Ethernet is one of the contributors to the ″least common denominator″ algorithm of MTU choice. If a
configuration includes Ethernets and other LANs, and there is extensive traffic among them, the MTUs of
all of the LANs may need to be set to 1500 bytes to avoid fragmentation when data enters an Ethernet.
Following are some guidelines:
v Do not change the default (and maximum) MTU of 1500 bytes.
v Set the application block size in multiples of 4096 bytes.
v Keep socket space settings at the default values.
v If the workload includes extensive use of services that use UDP, such as NFS or RPC, increase
sb_max to allow for the fact that each 1500-byte MTU uses a 4096-byte buffer.
Token-Ring (4 MB) Performance Tuning Recommendations
The default MTU of 1492 bytes is appropriate for Token-Rings that interconnect to Ethernets or to
heterogeneous networks in which the minimum MTU is not known. Following are some guidelines:
v Unless the LAN has extensive traffic to outside networks, raise the MTU to the maximum of 3900 bytes.
v Application block size should be in multiples of 4096 bytes.
v Socket space settings can be left at the default values.
v If the workload includes extensive use of services that use UDP, such as NFS or RPC, increase
sb_max to allow for the fact that each 1492-byte MTU uses a 4096-byte buffer.
Token-Ring (16 MB) Performance Tuning Recommendations
The default MTU of 1492 bytes is appropriate for Token-Rings that interconnect to Ethernets or to
heterogeneous networks in which the minimum MTU is not known. Following are some guidelines:
v Unless the LAN has extensive traffic to outside networks, raise the MTU to 8500 bytes. This allows NFS
8 KB packets to fit in one MTU. Further increasing the MTU to the maximum of 17000 bytes seldom
results in corresponding throughput improvement.
v Application block size should be in multiples of 4096 bytes.
v Socket space settings can be left at the default values.
v If the workload includes extensive use of services that use UDP, such as NFS or RPC, and the MTU
must be left at the default because of interconnections, increase sb_max to allow for the fact that each
1492-byte MTU uses a 4096-byte buffer.
Chapter 10. Monitoring and Tuning Communications I/O Use
273
FDDI Performance Tuning Recommendations
Despite the comparatively low MTU, this high-speed medium benefits from substantial increases in socket
buffer size. Following are some guidelines:
v Unless the LAN has extensive traffic to outside networks, the default MTU of 4352 bytes should be
retained.
v Where possible, an application using TCP should write multiples of 4096 bytes at a time (preferably 8
KB or 16 KB) for maximum throughput.
v Use no -o *_*space=NewSize to set the TCP and UDP socket send and receive space defaults to
NewSize bytes. NewSize should be at least 57344 bytes (56 KB).
v Use no -o sb_max=(2*NewSize) to increase the maximum number of socket buffer space.
v For RS/6000 Model *90 or faster, use no -o rfc1323=1 to allow socket buffer sizes to be set to more
than 65536. Then use the previous procedure with NewSize of at least 128 KB.
ATM Performance Tuning Recommendations
Following are some guidelines:
v Unless the LAN has extensive traffic to outside networks, retain the default MTU of 9180 bytes. Where
possible, an application using TCP should write multiples of 4096 bytes at a time (preferably 8 KB or 16
KB) for maximum throughput.
v Use no -o *_*space=NewSize to set the TCP and UDP socket send and receive space defaults to
NewSize bytes. NewSize should be at least 57344 bytes (56 KB).
v Use no -o sb_max=(2*NewSize) to increase the maximum number of socket buffer space.
v For RS/6000 Model *90 or faster, use no -o rfc1323=1 to allow socket buffer sizes to be set to more
than 65536. Then use the previous procedure with NewSize of at least 128 KB.
SOCC Performance Tuning Recommendations
Following are some guidelines:
v The default MTU 61428 bytes should not be changed.
v Where possible, an application using TCP should write 28672 bytes (28 KB) at a time for maximum
throughput.
v Set TCP and UDP socket send and receive space defaults to 57344 bytes.
HIPPI Performance Tuning Recommendations
Following are some guidelines:
v
v
v
v
The default MTU of 65536 bytes should not be changed.
Where possible, an application using TCP should write 65536 bytes at a time for maximum throughput.
Set sb_max to a value greater than 2 * 655360.
TCP and UDP socket send and receive space defaults should be set to 655360 bytes. Use no -o
rfc1323=1 to allow socket buffer sizes to be set to more than 65536.
Tuning mbuf Pool Performance
The network subsystem uses a memory management facility that revolves around a data structure called
an mbuf. Mbufs are mostly used to store data for incoming and outbound network traffic. Having mbuf
pools of the right size can have a very positive effect on network performance. If the mbuf pools are
configured incorrectly, both network and system performance can suffer. In AIX Version 4, the mbuf
parameters are automatically tuned based on the system configuration, particularly the number of
communications adapters, and network requirements.
274
Performance Management Guide
Overview of the mbuf Management Facility
The mbuf management facility controls different buffer sizes that can range from 32 bytes up to 16384
bytes. The pools are created from system memory by making an allocation request to the Virtual Memory
Manager (VMM). The pools consist of pinned pieces of kernel virtual memory in which they always reside
in physical memory and are never paged out. The result is that the real memory available for paging in
application programs and data has been decreased by the amount that the mbuf pools have been
increased.
The network memory pool is split evenly among each processor. Each sub-pool is then split up into
buckets of 32-16384 bytes. Each bucket can borrow memory from other buckets on the same processor
but a processor cannot borrow memory from another processor’s network memory pool. When a network
service needs to transport data, it can call a kernel service such as m_get() to obtain a memory buffer. If
the buffer is already available and pinned, it can get it immediately. If the upper limit has not been reached
and the buffer is not pinned, then a buffer is allocated and pinned. Once pinned, the memory stays pinned
but can be freed back to the network pool. If the number of free buffers reaches a high-water mark, then a
certain number is unpinned and given back to the system for general use. This unpinning is done by the
netm() kernel process. The caller of the m_get() subroutine can specify whether to wait for a network
memory buffer. If M_DONTWAIT is specified and no pinned buffers are available at that time, a failed
counter is incremented. If M_WAIT is specified, the process is put to sleep until the buffer can be allocated
and pinned.
Tuning Network Memory
Use the command netstat -m to detect shortages of network memory (mbufs/clusters).
An upper limit is placed on how much of RAM can be used for network memory. You can tune this
parameter by setting maxmbuf or thewall. The unit of thewall is in 1 KB, so that 16384 indicates 16 MB
of RAM.
The default value for thewall in operating system version
v
v
v
v
4.2.1
4.3.0
4.3.1
4.3.2
is 1/8 of RAM or 16384 (16 MB), whichever is smaller
is 1/8 of RAM or 131072 (128 MB), whichever is smaller
is 1/2 of RAM or 131072 (128 MB), whichever is smaller
and later is 1/2 of RAM or 1048576 (1 GB), whichever is smaller
The maximum value of thewall is 64 MB prior to operating system version 4.3.0, 128 MB in operating
system version AIX 4.3.0 and version AIX 4.3.1, and 1 GB in AIX 4.3.2 and higher (systems that are not
Common Hardware Reference Platform (CHRP) are limited to 256 MB).
The value of maxmbuf is also used to limit how much real memory can be used by the communications
subsystem. You can view this by running the command lsattr -E -l sys0. If maxmbuf is greater than 0
(can be changed using the chdev or smitty commands), the maxmbuf value is used regardless of the
value of thewall. Previously, the upper limit on mbufs was the higher value of maxmbuf or thewall.
maxmbuf is set to 0. The value of 0 indicates that the system default (thewall) should be used.
The limit on network memory (thewall or maxmbuf) is also used to limit how much memory can be used
for streams. The tunable parameter called strthresh (default value of 85 percent of thewall/maxmbuf)
specifies that once the total amount of allocated memory has reached 85 percent, no more memory can
be given to STREAMS.
Similarly, another threshold called sockthresh (also defaults to 85 percent) specifies that once the total
amount of allocated memory has reached 85 percent of thewall/maxmbuf, no new socket connections
can occur (the socket() and socketpair() system calls return with ENOBUFS). Use the no command to tune
these thresholds.
Chapter 10. Monitoring and Tuning Communications I/O Use
275
Tuning Asynchronous Connections for High-Speed Transfers
Async ports permit the connection to a computer of optional devices such as terminals, printers, fax
machines, and modems. Async ports are provided by adapter devices such as the 8-, 16-, or 64-port IBM
adapters or the 128-port Digiboard adapter, which reside on the Micro Channel and provide multiple
asynchronous connections, typically RS232 or RS422. Many adapters, such as the three IBM async
adapters mentioned above, were originally designed for servicing terminals and printers, and so are
optimized for output (sends). Input processing (receives) is not as well optimized, perhaps because the
assumption was once made that users could not type very fast. This is not a great concern when data
transmission is slow and irregular, as with keyboard input. It becomes a problem with raw-mode
applications, where massive chunks of input are transmitted by other computers and by devices such as
fax machines. In raw-mode processing, data is treated as a continuous stream; input bytes are not
assembled into lines, and erase and kill processing are disabled.
Async Port Tuning Techniques
While some adapters have inherent limitations, the guidelines and techniques described here can provide
better performance from these adapters for raw-mode transfers.
A shell script containing appropriate stty commands to implement most of the following tuning techniques
is given at the end of the section.
v For input processing, using the echo option is expensive, as it increases the time per character.
Character echo is useful for canonical user input but is probably not necessary for most raw-mode
applications. Example:
# stty -echo < /dev/ttyn
v Increase the value of the vmin variable for each TTY from the default of 4. The vmin variable value is
the minimum number of bytes that should be received when the read is successful. The value chosen
for the vmin variable should be the lesser of the application data-block size or 255 (the maximum
allowed). If the application block size is variable or unknown, vmin should be set to 255. Setting the
vmin variable to 255 will result in fewer read executions and will reduce CPU utilization by 15 to 20
percent for file-transfer programs.
v Except on the 128-port adapter, set vtime > 0 to prevent an indefinite block on read. If the vtime
variable is set to zero on the 128-port adapter, POSIX line-discipline processing will be offloaded to the
adapter hardware, significantly reducing CPU processing.
v For raw-mode sends where output translations are not needed, turn off the opost option in the POSIX
line discipline. This will help the CPU performance by reducing the output path length. For file-transfer
applications, which move large amounts of data on TTY lines, this can reduce CPU utilization by 3
times. Example:
# stty -opost < /dev/ttyn
v Because the 64-port adapter is prone to unpredictable data overruns at higher baud rates when
XON/XOFF is used for pacing, avoid the risk of losing data by using RTS/CTS hardware pacing instead.
v Because the 64-port-adapter concentrator boxes have a limited bandwidth and saturate at higher baud
rates, adding more ports to a saturated concentrator will decrease the performance of all ports
connected. Instead, add another concentrator and continue until it is saturated or you have run out of
CPU.
Also, concentrator saturation is a concern because, if the concentrator box approaches overload, no
additional throughput is accepted. The effective baud rate is lowered, and there is a noticeable slowdown
in work.
Shell Script fastport.sh for Fast File Transfers
The fastport.sh script is intended to condition a TTY port for fast file transfers in raw mode; for example,
when a fax machine is to be connected. Using the script may improve CPU performance by a factor of 3
at 38,400 baud. The fastport.sh script is not intended for the canonical processing that is used when
276
Performance Management Guide
interacting with a user at an async terminal, because canonical processing cannot be easily buffered. The
bandwidth of the canonical read is too small for the fast-port settings to make a perceptible difference.
Any TTY port can be configured as a fast port. The improved performance is the result of reducing the
number of interrupts to the CPU during the read cycle on a given TTY line. To configure a TTY port as a
fast port, do the following:
1. Create a TTY for the port using the command smitty -> Devices -> TTY -> Add a TTY), with Enable
LOGIN=disable and BAUD rate=38,400.
2. Create the Korn shell script named fastport.sh, as follows:
#****************************************************************
#
#
Configures a fastport for "raw" async I/O.
#
#****************************************************************
set -x
sync;sync
i=$1
if [ $i -le 100 ]
then
# for the native async ports and the 8-, 16-, and 64-port adapters
# set vmin=255 and vtime=0.5 secs with the following stty
stty -g </dev/tty$i |awk ’ BEGIN { FS=":";OFS=":" }
{ $5="ff";$6=5;print $0 } ’ >foo
# for a 128-port adapter, remove the preceding stty, then
# uncomment and use the
# following stty instead to
# set vmin=255 and vtime=0 to offload line discipline processing
# stty -g </dev/tty$i |awk ’ BEGIN { FS=":";OFS=":" }
# { $5="ff";$6=0;print $0 } ’ >foo
stty `cat foo ` </dev/tty$i
sleep 2
# set raw mode with minimal input and output processing
stty -opost -icanon -isig -icrnl -echo -onlcr</dev/tty$i
rm foo
sync;sync
else
echo "Usage is fastport.sh < TTY number >"
fi
3. Run the script for TTY number with the following command:
# fastport.sh ttynumber
Tuning Name Resolution
TCP/IP attempts to obtain an Internet Protocol (IP) address from a host name in a process known as
name resolution. The process of translating an Internet Protocol address into a host name is known as
reverse name resolution. A resolver routine is used to resolve names. It queries DNS, NIS and finally the
local /etc/hosts file to find the required information.
You can accelerate the process of name resolution by overwriting the default search order, if you know
how you want names to be resolved. This is done through the use of the /etc/netsvc.conf file or the
NSORDER environment variable.
v If both the /etc/netsvc.conf file and the NSORDER are used, NSORDER overrides the
/etc/netsvc.conf file. To specify host ordering with /etc/netsvc.conf, create the file and include the
following line:
hosts=value,value,value
where value may be (lowercase only) bind, local, nis, bind4, bind6, local4, local6, nis4, or nis6 (for
/etc/hosts). The order is specified on one line with values separated by commas. White spaces are
permitted between the commas and the equal sign.
Chapter 10. Monitoring and Tuning Communications I/O Use
277
The values specified and their ordering is dependent on the network configuration. For example, if the
local network is organized as a flat network, then only the /etc/hosts file is needed. The
/etc/netsvc.conf file would contain the following line:
hosts=local
The NSORDER environment variable would be set as:
NSORDER=local
v If the local network is a domain network using a name server for name resolution and an /etc/hosts file
for backup, specify both services. The /etc/netsvc.conf file would contain the following line:
hosts=bind,local
The NSORDER environment variable would be set as:
NSORDER=bind,local
The algorithm will attempt the first source in the list. The algorithm will then determine to try another
specified service based on:
v Current service is not running; therefore, it is unavailable.
v Current service could not find the name and is not authoritative.
Improving telnetd/rlogind Performance
In AIX 4.2, the rlogind daemon was changed to move portions of its code into the pseudo tty driver
(ptydd) in the kernel. This change eliminated context switches needed to transfer a single character from
an application to a terminal. This feature is enabled by default and cannot be configured.
In AIX 4.3, this change was extended to the telnetd daemon. Data coming in goes directly from the socket
layer to the pty device driver’s slave side. This change eliminated the need for a context switch of the
telnetd daemon to write incoming data to the PTS (pseudo-terminal slave).
This option is turned on by default in AIX 4.3.2 and later releases. Before AIX 4.3.2, you must add a -a
flag to the telnetd daemon in /etc/inetd.conf and refresh the inetd daemon.
Expected performance improvements for the rlogind and telnetd daemons with this feature is about 50
percent.
Tuning the SP Network
This section provides information about network tunable parameters that need special attention in an SP
environment. For a more detailed discussion on SP tuning, see RS/6000 SP System Performance Tuning.
SP Switch Statistics
Note: The commands in this section are SP-specific.
The estat Command
The unsupported and undocumented estat command can be helpful in determining SP Switch problems.
The entstat command is located in the /usr/lpp/ssp/css directory and produces output similar to the
entstat command. The output contains sections for transmit, receive, and general statistics.
The output that is helpful in determining SP Switch problems is Transmit Errors, Receive Errors, Packets
Dropped, and No mbuf Errors. The second line in the output indicates how long the adapter has been
online.
278
Performance Management Guide
The vdidlxxxx Commands
These unsupported and undocumented commands display the SP Switch pool usage since the SP Switch
was last started. There are several commands called vdidlxxxx (where vdidl3 is for an MCA-based node,
vdidl3mx for a 332 MHz node, and vdidl3pci for the S70 and S7A). These commands are found in the
/usr/lpp/ssp/css directory on each node. For the SP Switch, only the send pool is used because
microcode in the adapter manages the receive pool.
Following is an example for the vdidl3 command:
# /usr/lpp/ssp/css/vdidl3 -i
send pool: size=524288 [email protected]=0x50002c00
bkt
allocd
free success
fail
12
0
0
0
0
13
0
0
0
0
14
0
0
0
0
15
0
0
0
0
16
0
8
0
0
rsvd pool: size=262144 [email protected]=0x50002000
bkt
allocd
free success
fail
12
0
0
0
0
13
0
0
0
0
14
0
0
0
0
15
0
0
0
0
16
0
4
0
0
recv pool: size=524288 [email protected]=0x50002e00
bkt
allocd
free success
fail
12
0
0
0
0
13
0
0
0
0
14
0
0
0
0
15
0
0
0
0
16
0
0
0
0
[email protected]=0x50dc0000
split
comb
1
0
0
0
0
0
0
0
0
0
[email protected]=0x50e40000
split
comb
0
0
0
0
0
0
0
0
0
0
[email protected]=0x50e80000
split
comb
0
0
0
0
0
0
0
0
0
0
[email protected]=0x50001d00
freed
0
0
0
0
0
[email protected]=0x50b84680
freed
0
0
0
0
0
[email protected]=0x50001e00
freed
0
0
0
0
0
Interpret the output carefullt because some of the statistics have several meanings. For the SP Switch,
only the send pool is used because the receive pool is managed by microcode in the adapter. Each
column is described as follows:
bkt
Lists the pool allocation in powers of 2 for the line it is on. The line starting with 12 means 2 to the
12th or 4 KB allocations, and the line starting with 16 means 2 to the 16th or 64 KB allocations.
allocd Lists the current allocations at the time the command was run, for each of the size allocations in
the first column. This ″snapshot″ value fluctuates between successive executions of the command.
free
Lists the number of buffers of each allocation that are allocated and unused at the time the
command was run. In the above example, eight 64 K allocations are free for use. This
instantaneous value canfluctuates between successive executions of the command.
success
This counter increments every time an allocation of the given size succeeded. This counter is
cumulative, so it shows the number of successes since the adapter was last initialized.
fail
This counter is incremented every time an allocation is not available for the size requested.
However, it is possible that the allocation is made by splitting up a larger allocation, or combining
smaller ones into the size needed. A fail does not necessarily mean a packet was dropped. It is
also possible that an allocation was split from a larger allocation without incurring a failed count.
This counter is cumulative, so it shows the number of fails since the adapter was last initialized.
split
This counter indicates how many times the allocation size was extracted from the pool by carving
the size needed from a larger allocation in the pool. This counter is cumulative, so it shows the
number of splits since the adapter was last initialized.
comb This field is currently not used.
freed
This field is currently not used.
Chapter 10. Monitoring and Tuning Communications I/O Use
279
SP System-Specific Tuning Recommendations
The following are specific details for setting the network tunables for the SP system.
arptab_bsiz
Number of ARP cache entries in each bucket (default = 7). See the following table for
recommendations.
arptab_nb
Number of ARP cache buckets (default = 25). See the following table for recommendations.
CSS MTU size
The recommended MTU of a switch is 65520. However, under some circumstances (for example,
when you want to avoid the Nagle algorithm causing very slow traffic), it may be necessary to
reduce this value. You can reduce the MTU of a switch to 32678 with only a 2 to 10 percent loss
in throughput. However, CPU utilization will be slightly higher due to the per-packet overhead. To
reduce the MTU of the switch, run the following:
ifconfig css0 mtu new_size
This command takes effect immediately and must be run as root user. Always use the same MTU
across all nodes in an SP.
rfc1323
See rfc1323.
sb_max
See TCP Socket Buffer Tuning.
tcp_mssdflt
See When to Use the tcp_mssdflt Option of the no Command.
tcp_sendspace and tcp_recvspace
See TCP Socket Buffer Tuning. These parameters should never be higher than the major network
adapter transmit queue limit. To calculate this limit, use:
(major adapter queue size) * (major network adapter MTU)
thewall
See Tuning Network Memory.
udp_sendspace and udp_recvspace
See UDP Socket Buffer Tuning.
By default, the maximum number of ARP entries allowed is 175 (25 * 7). This default value of 175 might
not be large enough in SP environments with many nodes. An inadequate number of slots in the ARP
cache will slow the performance of nodes in the system. Use the following table to estimate optimal values
for the arptab_nb and arptab_bsiz variables.
Number of Nodes
arptab_nb
Number of Interfaces
arptab_bsiz
1 - 64
25
1-3
7
65 - 128
64
4
8
129 - 256
128
5
10
257 - 512
256
more...
2 times number of interfaces
In general, arptab_nb increases monotonically with the number of nodes and arptab_bsiz with the
number of IP interfaces in an SP system.
These parameters must be placed in the first section of the /etc/rc.net file in front of the configuration
methods.
280
Performance Management Guide
Managing Tunable SP Parameters
The SP usually requires that tunable settings be changed from the default values in order to achieve
optimal performance of the entire system. Placement of these tunable values is crucial. If they are not set
in the correct places, subsequent rebooting of the nodes or other changes can cause them to change or
be lost.
For all dynamically tunable values (those that take effect immediately), the setting for each node should be
set in the tuning.cust file. This file is found in the /tftpboot directory on each node. There is also a copy
of the file in this same directory on the Control Work Station (CWS). Tunable parameters changed using
the no, nfso or vmtune command can be included in this file. Even though the sample files do not include
nfso and vmtune commands, the commands can be added.
A small number of tuning recommendations that are not dynamically tunable values need to be changed in
the rc.net file. These tunable parameters are for ARP cache-tuning and setting the number of adapter
types per interface. The following are the only tunable parameters that should be added to rc.net:
v arptab_nb
v arptab_bsize
v arpqsize
v ifsize
Using the sample tuning.cust settings selected as part of the installation is a sufficient starting point for
the SP nodes in the environment type selected.
If the system has nodes that require different tuning settings, it is recommended that a copy of each
setting be saved on the CWS. When nodes with specific tuning settings are installed, that version of
tuning.cust must be moved into /tftpboot on the CWS.
Another option is to create one tuning.cust file that determines the node number, and based on that node
number, sets the appropriate tuning values.
Initial Settings of SP Tunable Parameters
When a node is installed, migrated or customized, and that node’s boot/install server does not have a
/tftpboot/tuning.cust file, a default file with performance tuning variable settings in
/usr/lpp/ssp/install/tuning.default is copied to /tftpboot/tuning.cust on that node. You can choose from
one of the four sample tuning files, or you can create and customize your own. The existing files are
located in the /usr/lpp/ssp/install/config directory and are as follows:
tuning.commercial
Contains initial performance tuning parameters for a typical commercial environment.
tuning.development
Contains initial performance tuning parameters for a typical interactive or development
environment. These are the default tuning parameters.
tuning.scientific
Contains initial performance tuning parameters for a typical engineering/scientific environment.
tuning.server
Contains initial performance tuning parameters for a typical server environment.
The other option is to create and select your own alternate tuning file. While this may not be the initial
choice, it certainly must be the choice at some point in time. On the CWS, create a tuning.cust file, or
you can begin with one of the sample files. Edit the tuning.cust file and proceed to the installation of
nodes. This tuning.cust file is then propagated to each node’s /tftpboot/tuning.cust file from the
boot/install server when the node is installed, migrated, or customized. The tuning file is maintained across
reboots.
Chapter 10. Monitoring and Tuning Communications I/O Use
281
Tuning the SP Network for Specific Workloads
The following table provides a combined overview of tunable parameters for different environments. The
settings given are only initial settings and are not guaranteed to be optimized for your environment.
Examine your specific implementation and adjust your tuning settings accordingly.
Parameter
Commercial
Environment
Server Environment Scientific
Environment
Development
Environment
thewall
16384
65536
16384
16384
sb_max
1310720
1310720
1310720
1310720
subnetsarelocal
1
1
1
1
ipforwarding
1
1
1
1
tcp_sendspace
262144
65536
655360
65536
tcp_recvspace
262144
65536
655360
65536
udp_sendspace
65536
65536
65536
32768
udp_recvspace
655360
655360
655360
65536
rfc1323
1
1
1
1
tcp_mssdflt
1448
1448
Varies depending on
other network types
1448
tcp_mtu_discover (AIX
4.2.1 and later)
1
1
1
1
udp_mtu_discover (AIX
4.2.1 and later)
1
1
1
1
282
Performance Management Guide
Chapter 11. Monitoring and Tuning NFS Use
This chapter discusses Network File System (NFS) monitoring and tuning. It contains the following major
sections:
v NFS Overview
v Analyzing NFS Performance
v Tuning for NFS Performance
NFS Overview
NFS allows programs on one system to access files on another system transparently by mounting the
remote directory. Usually, when the server is booted, directories are made available by the exportfs
command, and the daemons to handle remote access (nfsd daemons) are started. Similarly, the mounts
of the remote directories and the initiation of the appropriate numbers of NFS block I/O daemons (biod
daemon) to handle remote access are performed during client system boot.
Prior to AIX 4.2.1, nfsd and biod daemons were processes that required tuning the number of nfsd and
biod daemons. Now, there is a single nfsd and biod daemon, each of which is multithreaded (multiple
kernel threads within the process). Also, the number of threads is self-tuning in that it creates or deletes
threads as needed.
The following figure illustrates the structure of the dialog between NFS clients and a server. When a thread
in a client system attempts to read or write a file in an NFS-mounted directory, the request is redirected
from the usual I/O mechanism to one of the client’s biod daemon. The biod daemon sends the request to
the appropriate server, where it is assigned to one of the server’s NFS daemons (nfsd daemon). While
that request is being processed, neither the biod nor the nfsd daemon involved do any other work.
Figure 27. NFS Client-Server Interaction. This illustration shows two clients and one server on a network that is laid
out in a typical star topology. Client A is running thread m in which data is directed to one of its biod daemons. Client
B is running thread n and directing data to its biod daemons. The respective daemons send the data across the
network to server Z where it is assigned to one of the server’s NFS (nfsd) daemons.
NFS uses Remote Procedure Calls (RPC) to communicate. RPCs are built on top of the External Data
Representation (XDR) protocol which transforms data to a generic format before transmitting and allowing
machines with different architectures to exchange information. The RPC library is a library of procedures
that allows a local (client) process to direct a remote (server) process to execute a procedure call as if the
© Copyright IBM Corp. 1997, 2002
283
local (client) process had executed the procedure call in its own address space. Because the client and
server are two separate processes, they no longer have to exist on the same physical system.
Figure 28. The Mount and NFS Process. This illustration is a three column table with Client Activity, Client, and Server
as the three column headings. The first client activity is mount. An rpc call goes from the client to the server’s
portmapper mountd. The second client activity is open/close read/write. There is two-way interaction between the
client’s biod daemon and the server’s nfsd daemon.
The portmap daemon, portmapper, is a network service daemon that provides clients with a standard
way of looking up a port number associated with a specific program. When services on a server are
requested, they register with portmap daemon as an available server. The portmap daemon then
maintains a table of program-to-port pairs.
When the client initiates a request to the server, it first contacts the portmap daemon to see where the
service resides. The portmap daemon listens on a well-known port so the client does not have to look for
it. The portmap daemon responds to the client with the port of the service that the client is requesting.
The client, upon receipt of the port number, is able to make all of its future requests directly to the
application.
The mountd daemon is a server daemon that answers a client request to mount a server’s exported file
system or directory. The mountd daemon determines which file system is available by reading the
/etc/xtab file. The mount process takes place as follows:
1. Client mount makes call to server’s portmap daemon to find the port number assigned to the mountd
daemon.
2. The portmap daemon passes the port number to the client.
3. The client mount command then contacts the server mountd daemon directly and passes the name of
the desired directory.
4. The server mountd daemon checks /etc/xtab (built by the exportfs -a command, which reads
/etc/exports) to verify availability and permissions on the requested directory.
5. If all is verified, the server mountd daemon gets a file handle (pointer to file system directory) for the
exported directory and passes it back to the client’s kernel.
The client only contacts the portmap daemon on its very first mount request after a system restart. Once
the client knows the port number of the mountd daemon, the client goes directly to that port number for
any subsequent mount request.
The biod daemon is the block input/output daemon and is required in order to perform read-ahead and
write-behind requests, as well as directory reads. The biod daemon threads improve NFS performance by
filling or emptying the buffer cache on behalf of the NFS clients. When a user on a client system wants to
read or write to a file on a server, the biod threads send the requests to the server. The following NFS
operations are sent directly to the server from the operating system’s NFS client kernel extension and do
not require the use of the biod daemon:
v getattr()
284
Performance Management Guide
v
v
v
v
v
setattr()
lookup()
readlink()
create()
remove()
v
v
v
v
v
v
v
rename()
link()
symlink()
mkdir()
rmdir()
readdir()
readdirplus()
v fsstat()
The nfsd daemon is the active agent providing NFS services from the NFS server. The receipt of any one
NFS protocol request from a client requires the dedicated attention of an nfsd daemon thread until that
request is satisfied and the results of the request processing are sent back to the client.
NFS Network Transport
Prior to AIX 4.2.1, UDP was the exclusive transport protocol for NFS packets. TCP was added as an
alternative protocol in AIX 4.2.1 and is now the default. UDP works efficiently over clean or efficient
networks and responsive servers. For wide area networks or for busy networks or networks with slower
servers, TCP may provide better performance because its inherent flow control can minimize retransmits
on the network.
NFS Version 3
While the operating system supports both NFS Version 2 and Version 3 on the same machine, NFS
Version 3, which was introduced in AIX 4.2.1, can enhance performance in many ways:
Write Throughput
Applications running on client systems may periodically write data to a file, changing the file’s contents.
The amount of time an application waits for its data to be written to stable storage on the server is a
measurement of the write throughput of a distributed file system. Write throughput is therefore an
important aspect of performance. All distributed file systems, including NFS, must ensure that data is
safely written to the destination file while at the same time minimizing the impact of server latency on write
throughput.
The NFS Version 3 protocol offers a better alternative to increasing write throughput by eliminating the
synchronous write requirement of NFS Version 2 while retaining the benefits of close-to-open semantics.
The NFS Version 3 client significantly reduces the latency of write operations to the server by writing the
data to the server’s cache file data (main memory), but not necessarily to disk. Subsequently, the NFS
client issues a commit operation request to the server that ensures that the server has written all the data
to stable storage. This feature, referred to as safe asynchronous writes, can vastly reduce the number of
disk I/O requests on the server, thus significantly improving write throughput.
The writes are considered ″safe″ because status information on the data is maintained, indicating whether
it has been stored successfully. Therefore, if the server crashes before a commit operation, the client will
know by looking at the status indication whether to resubmit a write request when the server comes back
up.
Chapter 11. Monitoring and Tuning NFS Use
285
Read Throughput
NFS sequential read throughput as measured at the client is enhanced via the VMM read-ahead and
caching mechanisms. Read-ahead allows file data to be transferred to the client from the NFS server in
anticipation of that data being requested by an NFS client application. By the time the request for data is
issued by the application, it is possible that the data resides already in the client’s memory, and thus the
request can be satisfied immediately. VMM caching allows re-reads of file data to occur instantaneously,
assuming that the data was not paged out of client memory which would necessitate retrieving the data
again from the NFS server.
In addition, CacheFS, see Cache File System (CacheFS), may be used to further enhance read
throughput in environments with memory-limited clients, very large files, and/or slow network segments by
adding the potential to satisfy read requests from file data residing in a local disk cache on the client.
Reduced Requests for File Attributes
Because read data can sometimes reside in the cache for extended periods of time in anticipation of
demand, clients must check to ensure their cached data remains valid if a change is made to the file by
another application. Therefore, the NFS client periodically acquires the file’s attributes, which includes the
time the file was last modified. Using the modification time, a client can determine whether its cached data
is still valid.
Keeping attribute requests to a minimum makes the client more efficient and minimizes server load, thus
increasing scalability and performance. Therefore, NFS Version 3 was designed to return attributes for all
operations. This increases the likelihood that the attributes in the cache are up to date and thus reduces
the number of separate attribute requests.
Efficient Use of High Bandwidth Network Technology
NFS Version 2 has an 8 KB maximum buffer-size limitation, which restricts the amount of NFS data that
can be transferred over the network at one time. In NFS Version 3, this limitation has been relaxed. The
default read/write size is 32KB for this operating system’s NFS and the maximum is 64KB, enabling NFS
to construct and transfer larger chunks of data. This feature allows NFS to more efficiently use high
bandwidth network technologies such as FDDI, 100baseT and 1000baseT Ethernet, and the SP Switch,
and contributes substantially to NFS performance gains in sequential read/write performance.
Reduced Directory ″Lookup″ Requests
A full directory listing (such as that produced by the ls -l command) requires that name and attribute
information be acquired from the server for all entries in the directory listing. NFS Version 2 clients query
the server separately for the file and directory names list and attribute information for all directory entries in
a lookup request. With NFS Version 3, names list and attribute information is returned at one time,
relieving both client and server from performing multiple tasks. However, in some environments, the NFS
Version 3 READDIRPLUS operation might cause slower performance.
Changes in AIX 5.2
Beginning with AIX 5.2, the nfs_v3_server_readdirplus nfso option may be used in such environments to
disable the use of READDIRPLUS. However, this is not generally recommended because this does not
comply with the NFS Version 3 standard.
Also in AIX 5.2, support was added for caching of longer filenames (greater than 31 characters) in the
NFS client directory name lookup cache (dnlc). Implementation of this feature will benefit NFS client work
loads using very long filenames, which previously would cause excessive NFS LOOKUP operations due to
dnlc misses. For example:
# ls -l
286
Performance Management Guide
Analyzing NFS Performance
NFS gathers statistics on types of NFS operations performed, along with error information and
performance indicators. You can use the following commands to identify network problems and observe
the type of NFS operations taking place on your system.
The nfsstat Command
The nfsstat command displays statistical information about the NFS and the RPC interface to the kernel
for clients and servers. This command could also be used to reinitialize the counters for these statistics
(nfsstat -z). For performance issues, the RPC statistics (-r option) are the first place to look. The NFS
statistics show you how the applications use NFS.
RPC Statistics
The nfsstat command displays statistical information about RPC calls, such as:
v Total number of RPC calls received or rejected
v Total number of RPC calls sent or rejected by a server
v Number of times no RPC packet was available when trying to receive
v Number of packets that were too short or had malformed headers
v Number of times a call had to be transmitted again
v
v
v
v
Number
Number
Number
Number
of
of
of
of
times
times
times
times
a reply did not match the call
a call timed out
a call had to wait on a busy client handle
authentication information had to be refreshed
The NFS part of the nfsstat command output is divided into Version 2 and Version 3 statistics of NFS.
The RPC part is divided into Connection oriented (TCP) and Connectionless (UDP) statistics.
NFS Server Information
The NFS server displays the number of NFS calls received (calls) and rejected (badcalls) due to
authentication, as well as the counts and percentages for the various kinds of calls made.
The following example shows the server part of the nfsstat command output specified by the -s option, as
follows:
# nfsstat -s
Server rpc:
Connection oriented:
calls
badcalls nullrecv badlen
15835
0
0
0
Connectionless:
calls
badcalls nullrecv badlen
0
0
0
0
Server nfs:
calls
badcalls
public_v2
15835
0
0
Version 2: (0 calls)
null
getattr
setattr
0 0%
0 0%
0 0%
wrcache
write
create
0 0%
0 0%
0 0%
mkdir
rmdir
readdir
0 0%
0 0%
0 0%
Version 3: (15835 calls)
null
getattr
setattr
7 0%
3033 19%
55 0%
write
create
mkdir
xdrcall
0
dupchecks dupreqs
772
0
xdrcall
0
dupchecks dupreqs
0
0
public_v3
0
root
0 0%
remove
0 0%
statfs
0 0%
lookup
0 0%
rename
0 0%
readlink
0 0%
link
0 0%
read
0 0%
symlink
0 0%
lookup
1008 6%
symlink
access
1542 9%
mknod
readlink
20 0%
remove
read
9000 56%
rmdir
Chapter 11. Monitoring and Tuning NFS Use
287
175 1%
rename
87 0%
commit
97 0%
185 1%
link
0 0%
0 0%
readdir
1 0%
0 0%
readdir+
150 0%
0 0%
fsstat
348 2%
120 0%
fsinfo
7 0%
0 0%
pathconf
0 0%
RPC output for server (-s) is as follows:
calls
Total number of RPC calls received from clients
badcalls
Total number of calls rejected by the RPC layer
nullrecv
Number of times an RPC call was not available when it was thought to be received
badlen
Packets truncated or damaged (number of RPC calls with a length shorter than a minimum-sized
RPC call)
xdrcall
Number of RPC calls whose header could not be External Data Representation (XDR) decoded
dupchecks
Number of RPC calls looked up in the duplicate request cache
dupreqs
Number of duplicate RPC calls found
The output also displays a count of the various kinds of calls and their respective percentages.
Duplicate checks are performed for operations that cannot be performed twice with the same result. The
classic example is the rm command. The first rm command will succeed, but if the reply is lost, the client
will retransmit it. We want duplicate requests like these to succeed, so the duplicate cache is consulted,
and if it is a duplicate request, the same (successful) result is returned on the duplicate request as was
generated on the initial request.
By looking at the percentage of calls for different types of operations (such as getattr(), read(), write(), or
readdir()), you can decide what type of tuning to use. For example, if the percentage of getattr() calls is
very high, then tuning attribute caches may be advantageous. If the percentage of write() calls is very
high, then disk and LVM tuning is important. If the percentage of read() calls is very high, then using more
RAM for caching files could improve performance.
NFS Client Information
The NFS client displays the number of calls sent and rejected, as well as the number of times a client
handle was received (clgets) and a count of the various kinds of calls and their respective percentages.
The following example shows the nfsstat output specified for clients using the -c option, as follows:
# nfsstat -c
Client rpc:
Connection oriented
calls
badcalls badxids timeouts
0
0
0
0
nomem
cantconn interrupts
0
0
0
Connectionless
calls
badcalls retrans
badxids
6553
0
0
0
timers
nomem
cantsend
0
0
0
Client nfs:
288
Performance Management Guide
newcreds badverfs
0
0
timers
timeouts newcreds
0
0
badverfs
0
0
calls
badcalls
clgets
6541
0
0
Version 2: (6541 calls)
null
getattr
setattr
0 0%
590 9%
414 6%
wrcache
write
create
0 0%
2482 37%
276 4%
mkdir
rmdir
readdir
6 0%
6 0%
30 0%
Version 3: (0 calls)
null
getattr
setattr
0 0%
0 0%
0 0%
write
create
mkdir
0 0%
0 0%
0 0%
rename
link
readdir
0 0%
0 0%
0 0%
commit
0 0%
cltoomany
0
root
0 0%
remove
277 4%
statfs
5 0%
lookup
2308 35%
rename
147 2%
readlink
0 0%
link
0 0%
read
0 0%
symlink
0 0%
lookup
0 0%
symlink
0 0%
readdir+
0 0%
access
0 0%
mknod
0 0%
fsstat
0 0%
readlink
0 0%
remove
0 0%
fsinfo
0 0%
read
0 0%
rmdir
0 0%
pathconf
0 0%
RPC output for the client (-c) is as follows:
calls
Total number of RPC calls made to NFS.
badcalls
Total number of calls rejected by the RPC layer.
retrans
Number of times a call had to be retransmitted due to a timeout while waiting for a reply from the
server. This is applicable only to RPC over connection-less transports.
badxid
Number of times a reply from a server was received that did not correspond to any outstanding
call. This means the server is taking too long to reply.
timeouts
Number of times a call timed-out while waiting for a reply from the server.
newcreds
Number of times authentication information had to be refreshed.
badverfs
Number of times a call failed due to a bad verifier in the response.
timers Number of times the calculated timeout value was greater than or equal to the minimum specified
timeout value for a call.
nomem
Number of times a call failed due to a failure to allocate memory.
cantconn
Number of times a call failed due to a failure to make a connection to the server.
interrupts
Number of times a call was interrupted by a signal before completing.
cantsend
Number of times a send failed due to a failure to make a connection to the client.
The output also displays a count of the various kinds of calls and their respective percentages.
For performance monitoring, the nfsstat -c command provides information on whether the network is
dropping UDP packets. A network may drop a packet if it cannot handle it. Dropped packets can be the
result of the response time of the network hardware or software or an overloaded CPU on the server.
Dropped packets are not actually lost, because a replacement request is issued for them.
Chapter 11. Monitoring and Tuning NFS Use
289
The retrans column in the RPC section displays the number of times requests were retransmitted due to
a timeout in waiting for a response. This situation is related to dropped UDP packets. If the retrans
number consistently exceeds five percent of the total calls in column one, it indicates a problem with the
server keeping up with demand. Use the vmstat, netpmon, and iostat commands on the server machine
to check the load.
A high badxid count implies that requests are reaching the various NFS servers, but the servers are too
loaded to send replies before the client’s RPC calls time out and are retransmitted. The badxid value is
incremented each time a duplicate reply is received for a transmitted request (an RPC request retains its
XID through all transmission cycles). Excessive retransmissions place an additional strain on the server,
further degrading response time. If badxid and timeouts are greater than five percent of the total calls,
increase the timeo parameter of the NFS-mount options by using the smitty chnfsmnt command. If
badxid is 0, but retrans and timeouts are sizable, attempt to decrease the NFS buffer size (that is, the
rsize and wsize options of the mount command).
If the server is CPU-bound, NFS and its daemons are affected. To improve the situation, the server must
be tuned or upgraded, or the user can localize the application files. If the server is I/O-bound, the server
file systems can be reorganized, or localized files can be used.
If the number of retransmits and timeouts are close to the same value, it is certain that packets are being
dropped. Packets are rarely dropped on the client. Usually, packets are dropped on either the network or
on the server. The server could drop packets if it overflows its interface driver’s transmit queue or if the
server’s User Datagram Protocol (UDP) socket buffer was overflown (nfs_socketsize). If there are no
socket buffer overflows or Oerrs on the server, and the client is getting many retransmits and timeouts,
packets are possibly being dropped on the network. Problems could occur in media and network devices,
such as routers, bridges, or concentrators. Network sniffers and other tools can be used to debug such
problems. See Dropped Packets for further discussion.
In some instances, an application or user experiences poor performance, yet examination of the nfsstat -c
output indicates no or very few timeouts and retransmits. This means that the client is receiving responses
from the server as fast as it is asking for them. The first thing to check is that there is an appropriate
number of biod daemons running on the client machine. This can also be observed when an application is
doing remote file locking. When remote file locks are set on a file served over NFS, the client goes into a
fully synchronous mode of operation that will turn off all data and attribute caching for the file. The result is
very slow performance and is, unfortunately, normal. Locking packets can be identified in ipreport output
by looking for NLM requests.
nfsstat -m: The nfsstat -m command displays the server name and address, mount flags, current read
and write sizes, retransmission count, and the timers used for dynamic retransmission for each NFS mount
on the client, as follows:
# nfsstat -m
/SAVE from /SAVE:itsorus.austin.ibm.com
Flags:
vers=2,proto=udp,auth=unix,soft,intr,dynamic,rsize=8192,wsize=8192,retrans=5
Lookups: srtt=27 (67ms), dev=17 (85ms), cur=11 (220ms)
Reads:
srtt=16 (40ms), dev=7 (35ms), cur=5 (100ms)
Writes: srtt=42 (105ms), dev=14 (70ms), cur=12 (240ms)
All:
srtt=27 (67ms), dev=17 (85ms), cur=11 (220ms)
The numbers in parentheses in the example output are the actual times in milliseconds. The other values
are unscaled values kept by the operating system kernel. You can ignore the unscaled values. Response
times are shown for lookups, reads, writes, and a combination of all of these operations (All). Other
definitions used in this output are as follows:
srtt
Smoothed round-trip time
dev
Estimated deviation
cur
Current backed-off timeout value
290
Performance Management Guide
The netpmon Command
See The netpmon Command for a discussion of this command and its output.
The nfso Command
The nfso command can be used to configure NFS attributes. It sets or displays network options in the
currently running kernel. Therefore, the command must run after each system startup or network
configuration.
Note: The nfso command performs no range-checking. If it is used incorrectly, the nfso command can
make your system inoperable.
The nfso parameters and their values can be displayed by using the nfso -a command, as follows:
(tremor:) # nfso -a
portcheck= 0
udpchecksum= 1
nfs_socketsize= 60000
nfs_tcp_socketsize= 60000
nfs_setattr_error= 0
nfs_gather_threshold= 4096
nfs_repeat_messages= 0
nfs_udp_duplicate_cache_size= 5000
nfs_tcp_duplicate_cache_size= 5000
nfs_server_base_priority= 0
nfs_dynamic_retrans= 1
nfs_iopace_pages= 0
nfs_max_connections= 1024
nfs_max_threads= 128
nfs_use_reserved_ports= 0
nfs_device_specific_bufs= 1
nfs_server_clread= 1
nfs_rfc1323= 1
nfs_max_write_size= 0
nfs_max_read_size= 0
nfs_allow_all_signals= 0
nfs_v2_pdts= 1
nfs_v3_pdts= 1
nfs_v2_vm_bufs= 1000
nfs_v3_vm_bufs= 1000
nfs_securenfs_authtimeout= 0
nfs_v3_server_readdirplus= 1
For a description of these attributes, see Network Tunable Parameters. Most NFS attributes are run-time
attributes that can be changed at any time. Load time attributes, such as nfs_socketsize, need NFS to be
stopped first and restarted afterwards.
To display or change a specific parameter, use the nfso -o command, as follows:
# nfso -o portcheck
portcheck= 0
# nfso -o portcheck=1
The parameters can be reset to their default value by using the -d option, as follows:
# nfso -d portcheck
# nfso -o portcheck
portcheck= 0
NFS References
Following is a summary of NFS-related files, commands, daemons, and subroutines. See the AIX 5L
Version 5.2 System Management Guide: Communications and Networks and the AIX 5L Version 5.2
Commands Reference for details.
Chapter 11. Monitoring and Tuning NFS Use
291
List of Network File System (NFS) Files
Following is a list of NFS files containing configuration information:
bootparams
Lists clients that diskless clients can use for booting
exports
Lists the directories that can be exported to NFS clients
networks
Contains information about networks on the Internet network
pcnfsd.conf
Provides configuration options for the rpc.pcnfsd daemon
rpc
Contains database information for Remote Procedure Call (RPC) programs
xtab
Lists directories that are currently exported
/etc/filesystems
Lists all the file systems that are attempted to be mounted at system restart
List of NFS Commands
Following is a list of NFS commands:
chnfs Starts a specified number of biod and nfsd daemons
mknfs Configures the system to run NFS and starts NFS daemons
nfso
Configures NFS network options
automount
Mounts an NFS file system automatically
chnfsexp
Changes the attributes of an NFS-exported directory
chnfsmnt
Changes the attributes of an NFS-mounted directory
exportfs
Exports and unexports directories to NFS clients
lsnfsexp
Displays the characteristics of directories that are exported with NFS
lsnfsmnt
Displays the characteristics of mounted NFS systems
mknfsexp
Exports a directory using NFS
mknfsmnt
Mounts a directory using NFS
rmnfs Stops the NFS daemons
rmnfsexp
Removes NFS-exported directories from a server’s list of exports
rmnfsmnt
Removes NFS-mounted file systems from a client’s list of mounts
List of NFS Daemons
Following is a list of NFS locking daemons:
lockd Processes lock requests through the RPC package
292
Performance Management Guide
statd
Provides crash-and-recovery functions for the locking services on NFS
Following is a list of network service daemons and utilities:
biod
Sends the client’s read and write requests to the server
mountd
Answers requests from clients for file system mounts
nfsd
Starts the daemons that handle a client’s request for file system operations
pcnfsd
Handles service requests from PC-NFS clients
nfsstat
Displays information about a machine’s ability to receive calls
on
Executes commands on remote machines
portmap
Maps RPC program numbers to Internet port numbers
rexd
Accepts request to run programs from remote machines
rpcgen
Generates C code to implement an RPC protocol
rpcinfo
Reports the status of RPC servers
rstatd Returns performance statistics obtained from the kernel
rup
Shows the status of a remote host on the local network
rusers
Reports a list of users logged on to the remote machines
rusersd
Responds to queries from the rusers command
rwall
Sends messages to all users on the network
rwalld Handles requests from the rwall command
showmount
Displays a list of all clients that have mounted remote file systems
spray Sends a specified number of packets to a host
sprayd
Receives packets sent by the spray command
Following is a list of secure networking daemons and utilities:
chkey Changes the user’s encryption key
keyenvoy
Provides an intermediary between user processes and the key server
keylogin
Decrypts and stores the user’s secret key
keyserv
Stores public and private keys
mkkeyserv
Starts the keyserv daemon and uncomments the appropriate entries in the /etc/rc.nfs file
Chapter 11. Monitoring and Tuning NFS Use
293
newkey
Creates a new key in the public key file
rmkeyserv
Stops the keyserv daemon and comments the entry for the keyserv daemon in the /etc/rc.nfs file
ypupdated
Updates information in Network Information Service (NIS) maps
Following is a diskless client support configuration file:
bootparamd
Provides information necessary for booting to diskless clients
Following is a list of NFS subroutines:
cbc_crypt(), des_setparity(), or ecb_crypt()
Implements Data Encryption Standard (DES) routines.
Tuning for NFS Performance
In an NFS network, the server is the primary target for tuning, but a few things can be tuned on the client
as well.
How Many biod and nfsd Daemons Are Needed?
Because biod and nfsd daemons handle one request at a time, and because NFS response time is often
the largest component of overall response time, it is undesirable to have threads blocked for lack of a biod
or nfsd daemon.
Note: There is a single nfsd daemon and a single biod daemon, each of which is multithreaded (multiple
kernel threads within the process). Also, the number of threads is self-tuning in that it creates
additional threads as needed. You can, however, tune the maximum number of nfsd threads by
using the nfs_max_threads parameter of the nfso command. You can also tune the maximum
number of biod threads per mount via the biod mount option.
The general considerations for configuring NFS daemons are as follows:
v Increasing the number of daemons cannot compensate for inadequate client or server processor power
or memory, or inadequate server disk bandwidth. Before changing the number of daemons, you should
check server and client resource-utilization levels with the iostat and vmstat commands.
v If the CPU or disk subsystem is already at near-saturation level, an increase in the number of daemons
will not yield better performance.
v NFS daemons are comparatively inexpensive. The biod or nfsd daemon costs a few pages in memory
(some of them pinned). Of course, the unpinned pages are only in real memory if the nfsd or biod
daemon has been active recently. Further, idle nfsd or biod daemons do not consume CPU time.
v All NFS requests go through an nfsd daemon; only reads and writes go through a biod daemon.
Choosing Initial Numbers of nfsd and biod daemons
Determining the best numbers of nfsd and biod daemons is an iterative process. Guidelines can give you
no more than a reasonable starting point.
The defaults are a good starting point for small systems, but should probably be increased for client
systems with more than two users or servers with more than two clients. A few guidelines are as follows:
v In each client, estimate the maximum number of files that will be written simultaneously. Configure at
least two biod daemons per file. If the files are large (more than 32 KB), you may want to start with four
biod daemons per file to support read-ahead or write-behind activity. It is common for up to five biod
daemons to be busy writing to a single large file.
294
Performance Management Guide
v In each server, start by configuring as many nfsd daemons as the sum of the numbers of biod
daemons that you have configured on the clients to handle files from that server. Add 20 percent to
allow for non-read/write NFS requests.
v If you have fast client workstations connected to a slower server, you may have to constrain the rate at
which the clients generate NFS requests.The best solution is to reduce the number of biod daemons on
the clients, with due attention to the relative importance of each client’s workload and response time.
Tuning the Numbers of nfsd and biod daemons
After you have arrived at an initial number of biod and nfsd daemons, or have changed one or the other,
do the following:
v First, recheck the affected systems for CPU or I/O saturation with the vmstat and iostat commands. If
the server is now saturated, you must reduce its load or increase its power, or both.
v Use the command netstat -s to determine if any system is experiencing UDP socket buffer
overflows. If so, use the command no -a to verify that the recommendations in Tuning Other Layers to
Improve NFS Performance have been implemented. If so, and the system is not saturated, increase the
number of biod or nfsd daemons.
v Examine the nullrecv column in the nfsstat -s output. If the number starts to grow, it may mean there
are too many nfsd daemons. However, this is less likely on this operating system’s NFS servers than it
is on other platforms. The reason for that is that all nfsd daemons are not awakened at the same time
when an NFS request comes into the server. Instead, the first nfsd daemon wakes up, and if there is
more work to do, this daemon wakes up the second nfsd daemon, and so on.
To change the number of nfsd daemons, you can use the chnfs command, or set the nfso
nfs_max_threads parameter as mentioned earlier.
To change the number of nfsd daemons on a server to 10, both immediately and at each subsequent
system boot, use the following:
# chnfs -n 10
To change the number of nfsd daemons on a system to 9, with the change delayed until the next system
boot, run the following command:
# chnfs -I -n 9
To change the number of biod daemons per mount, use the biod mount option.
Increasing the number of biod daemons on the client worsens server performance because it allows the
client to send more request at once, further loading the network and the server. In extreme cases of a
client overrunning the server, it may be necessary to reduce the client to one biod daemon, as follows:
# stopsrc -s biod
This leaves the client with the kernel process biod still running.
Performance Implications of Hard or Soft NFS Mounts
One of the choices you make when configuring NFS-mounted directories is whether the mounts will be
hard (-o hard) or soft (-o soft). When, after a successful mount, an access to a soft-mounted directory
encounters an error (typically, a timeout), the error is immediately reported to the program that requested
the remote access. When an access to a hard-mounted directory encounters an error, NFS retries the
operation.
A persistent error accessing a hard-mounted directory can escalate into a perceived performance problem
because the default number of retries (1000) and the default timeout value (0.7 second), combined with an
algorithm that increases the timeout value for successive retries, mean that NFS continues to try to
complete the operation.
Chapter 11. Monitoring and Tuning NFS Use
295
It is technically possible to reduce the number of retries, or increase the timeout value, or both, using
options of the mount command. Unfortunately, changing these values sufficiently to remove the perceived
performance problem might lead to unnecessary reported hard errors. Instead, use the intr option to
mount the hard-mounted directories, which allows the user to interrupt from the keyboard a process that is
in a retry loop.
Although soft-mounting the directories causes the error to be detected sooner, it runs a serious risk of data
corruption. In general, read/write directories should be hard-mounted.
Other mount Options That Affect Performance
The mount command provides some NFS-tuning options that are often ignored or used incorrectly
because of a lack of understanding of their use.
Before you start adjusting mount options, make certain that you know what you are trying to achieve with
respect to packet delivery and packet turnaround on the server or network. You would use most of the
NFS-specific mount options if your goal is to decrease the load on the NFS server, or to work around
network problems.
The NFS performance-specific mount options are all specified as a list entry on the -o option for the
mount command. Separate the options for the -o option on the command line only by a comma, not by a
comma and a space.
The rsize and wsize Options
The most useful options are those for changing the read and write size values. These options define the
maximum sizes of each RPC for read and write. Often, the rsize and wsize options of the mount
command are decreased to decrease the read/write packet that is sent to the server. There can be two
reasons why you might want to do this:
1. The server may not be capable of handling the data volume and speeds inherent in transferring the
read/write packets (8 KB for NFS Version 2 and 32 KB for NFS Version 3). This might be the case if a
NFS client is using a PC as an NFS server. The PC will likely have limited memory available for
buffering large packets.
2. If a read/write size is decreased, there may be a subsequent reduction in the number of IP fragments
generated by the call. If you are dealing with a faulty network, the chances of a call/reply pair
completing with a two-packet exchange are greater than if there must be seven packets successfully
exchanged. Likewise, if you are sending NFS packets across multiple networks with different
performance characteristics, the packet fragments may not all arrive before the timeout value for IP
fragments.
Reducing the rsize and wsize may improve the NFS performance in a congested network by sending
shorter package trains for each NFS request. But a side effect is that more packets are needed to send
data across the network, increasing total network traffic, as well as CPU utilization on both the server and
client.
If your NFS file system is mounted across a high-speed network, such as the SP Switch, then larger read
and write packet sizes would enhance NFS file system performance. With NFS Version 3, rsize and wsize
can be set as high as 65536. The default is 32768. With NFS Version 2, the largest that rsize and wsize
can be is 8192, which is also the default.
Improving sequential I/O throughput
On AIX 5.1 and later, you can improve the performance of sequential I/O operations on very large files in
NFS exported file systems by using a mechanism called release-behind.
For more information on release-behind, see Monitoring and Tuning File Systems.
296
Performance Management Guide
Improving client throughput for large-file I/O
On special cases over NFS Version 3 where you are trying to perform sequential write operations on files
that are larger than client memory, you can improve performance by using commit-behind. Under normal
circumstances, writing large files causes heavy page replacement activity on the client. This forces commit
operations to be performed one page at a time. Commit-behind enables a more aggressive logic for
commiting client pages and returning those pages to the free list.
You can enable commit-behind when mounting the file system by specifying the combehind flag with the
mount command. You also need to set an appropriate value for the numclust variable. This variable
specifies the number of 16-Kilobyte clusters processed by the sequential write-behind algorithm of the
Virtual Memory Manager (VMM). When the I/O pattern is sequential, use a large value for numclust in
order to keep more pages in RAM before scheduling them for I/O. Increase the value for numclust if
striped logical volumes or disk arrays are being used. On AIX 5.1, this parameter can be specified as a
mount option. On previous versions of AIX, use the vmtune command to specify a value for numclust.
Disabling Unused NFS ACL Support
If your workload does not use the NFS access control list (ACL) support on a mounted file system, you
can reduce the workload on both client and server to some extent by specifying the noacl option. This can
be done as follows:
options=noacl
Set this option as part of the client’s /etc/filesystems stanza for that file system.
Tuning to Avoid Retransmits
Related to the hard-versus-soft mount question is the question of the appropriate timeout duration for a
given network configuration. If the server is heavily loaded, is separated from the client by one or more
bridges or gateways, or is connected to the client by a WAN, the default timeout criterion may be
unrealistic. If so, both server and client are burdened with unnecessary retransmits. For example, if the
following command:
# nfsstat -c
reports a significant number (greater 5 percent of the total) of both timeouts and badxids, you could
increase the timeo parameter with the following SMIT fast path:
# smitty chnfsmnt
Identify the directory you want to change, and enter a new value on the line NFS TIMEOUT. In tenths of
a second.
The default time is 0.7 second (timeo=7), but this value is manipulated in the NFS kernel extension
depending on the type of call. For read calls, for instance, the value is doubled to 1.4 seconds.
To achieve control over the timeo value for operating system version 4 clients, you must set the
nfs_dynamic_retrans option of the nfso command to 0. There are two directions in which you can
change the timeo value, and in any given case, there is only one right way to change it. The correct way,
making the timeouts longer or shorter, depends on why the packets are not arriving in the allotted time.
If the packet is only late and does finally arrive, then you may want to make the timeo variable longer to
give the reply a chance to return before the request is retransmitted.
However, if the packet has been dropped and will never arrive at the client, then any time spent waiting for
the reply is wasted time, and you want to make the timeo shorter.
One way to estimate which option to take is to look at a client’s nfsstat -cr output and see if the client is
reporting lots of badxid counts. A badxid value means that an RPC client received an RPC call reply that
Chapter 11. Monitoring and Tuning NFS Use
297
was for a different call than the one it was expecting. Generally, this means that the client received a
duplicate reply for a previously retransmitted call. Packets are thus arriving late and the timeo should be
lengthened.
Also, if you have a network analyzer available, you can apply it to determine which of the two situations is
occurring. Lacking that, you can try setting the timeo option higher and lower and see what gives better
overall performance. In some cases, there is no consistent behavior. Your best option then is to track down
the actual cause of the packet delays/drops and fix the real problem; that is, server or network/network
device.
For LAN-to-LAN traffic through a bridge, try 50 (tenths of seconds). For WAN connections, try 200. Check
the NFS statistics again after waiting at least one day. If the statistics still indicate excessive retransmits,
increase timeo by 50 percent and try again. You will also want to examine the server workload and the
loads on the intervening bridges and gateways to see if any element is being saturated by other traffic.
Dropped Packets
Given that dropped packets are detected on an NFS client, the real challenge is to find out where they are
being lost. Packets can be dropped at the client, the server, and somewhere on the network.
Packets Dropped by the Client
Packets are rarely dropped by a client. Because each client RPC call is self-pacing, that is, each call must
get a reply before going on, there is little opportunity for overrunning system resources. The most stressful
operation is probably reading, where there is a potential for 1 MB+/sec of data flowing into the machine.
While the data volume can be high, the actual number of simultaneous RPC calls is fairly small and each
biod daemon has its own space allocated for the reply. Thus, it is very unusual for a client to drop
packets.
Packets are more commonly dropped either by the network or by the server.
Packets Dropped by the Server
Two situations exist where servers will drop packets under heavy loads:
1. Adapter Driver
When an NFS server is responding to a very large number of requests, the server will sometimes
overrun the interface driver output queue. You can observe this by looking at the statistics that are
reported by the netstat -i command. Examine the columns marked Oerrs and look for any counts.
Each Oerrs is a dropped packet. This is easily tuned by increasing the problem device driver’s transmit
queue size. The idea behind configurable queues is that you do not want to make the transmit queue
too long, because of latencies incurred in processing the queue. But because NFS maintains the same
port and XID for the call, a second call can be satisfied by the response to the first call’s reply.
Additionally, queue-handling latencies are far less than UDP retransmit latencies incurred by NFS if the
packet is dropped.
2. Socket Buffers
The second common place where a server will drop packets is the UDP socket buffer. Dropped
packets here are counted by the UDP layer and the statistics can be seen by using the netstat -p udp
command. Examine the statistics marked UDP: for the socket buffer overflows statistic.
NFS packets will usually be dropped at the socket buffer only when a server has a lot of NFS write
traffic. The NFS server uses a UDP socket attached to NFS port 2049 and all incoming data is
buffered on that UDP port. The default size of this buffer is 60,000 bytes. You can divide that number
by the size of the default NFS write packet (8192) to find that it will take only eight simultaneous write
packets to overflow that buffer. This overflow could occur with just two NFS clients (with the default
configurations).
In this situation there is either high volume or high burst traffic on the socket.
298
Performance Management Guide
v If there is high volume, a mixture of writes plus other possibly non-write NFS traffic, there may not
be enough nfsd daemons to take the data off the socket fast enough to keep up with the volume.
Recall that it takes a dedicated nfsd daemon to service each NFS call of any type.
v In the high burst case, there may be enough nfsd daemons, but the speed at which packets arrive
on the socket is such that they cannot wake up fast enough to keep it from overflowing.
Each of the two situations is handled differently.
v In the case of high volume, it may be sufficient to just increase the number of nfsd daemons
running on the system. Because there is no significant penalty for running with more nfsd daemons
on a machine, try this solution first. Also, see How Many biod and nfsd Daemons Are Needed?.
v In the case of high burst traffic, the only solution is to enlarge the socket, in the hope that some
reasonable size will be sufficiently large enough to give the nfsd daemons time to catch up with the
burst. Memory dedicated to this socket will not be available for any other use, so it must be noted
that a tuning objective of total elimination of socket buffer overflows by making the socket larger
may result in this memory being underutilized for the vast majority of the time. A cautious
administrator will watch the socket buffer overflow statistic, correlate it with performance problems,
and determine how large to make the socket buffer. See Increasing NFS Socket Buffer Size for
details on how to manipulate the NFS socket buffer.
You might see cases where the server has been tuned and no dropped packets are arriving for either the
socket buffer or the driver Oerrs, but clients are still experiencing timeouts and retransmits. Again, this is a
two-case scenario. If the server is heavily loaded, it may be that the server is just overloaded and the
backlog of work for nfsd daemons on the server is resulting in response times beyond the default timeout
set on the client. See NFS Tuning Checklist for hints on how to determine if this is the problem. The other
possibility, and the most likely problem if the server is known to be otherwise idle, is that packets are being
dropped on the network.
Dropped Packets On the Network
If there are no socket buffer overflows or Oerrs on the server, the client is getting lots of timeouts and
retransmits and the server is known to be idle, then packets are most likely being dropped on the network.
What is meant by when network is mentioned? It means a large variety of things including media and
network devices such as routers, bridges, concentrators, and the whole range of things that can implement
a transport for packets between the client and server.
Anytime a server is not overloaded and is not dropping packets, but NFS performance is bad, assume that
packets are being dropped on the network. Much effort can be expended proving this and finding exactly
how the network is dropping the packets. The easiest way of determining the problem depends mostly on
the physical proximity involved and resources available.
Sometimes the server and client are in close enough proximity to be direct-connected, bypassing the
larger network segments that may be causing problems. Obviously, if this is done and the problem is
resolved, then the machines themselves can be eliminated as the problem. More often, however, it is not
possible to wire up a direct connection, and the problem must be tracked down in place.
Tuning the NFS File-Attribute Cache
NFS maintains a cache on each client system of the attributes of recently accessed directories and files.
Five parameters that can be set in the /etc/filesystems file control how long a given entry is kept in the
cache. They are as follows:
actimeo
Absolute time for which file and directory entries are kept in the file-attribute cache after an
update. If specified, this value overrides the following *min and *max values, effectively setting
them all to the actimeo value.
acregmin
Minimum time after an update that file entries will be retained. The default is 3 seconds.
Chapter 11. Monitoring and Tuning NFS Use
299
acregmax
Maximum time after an update that file entries will be retained. The default is 60 seconds.
acdirmin
Minimum time after an update that directory entries will be retained. The default is 30 seconds.
acdirmax
Maximum time after an update that directory entries will be retained. The default is 60 seconds.
Each time the file or directory is updated, its removal is postponed for at least acregmin or acdirmin
seconds. If this is the second or subsequent update, the entry is kept at least as long as the interval
between the last two updates, but not more than acregmax or acdirmax seconds.
Tuning for Maximum Caching of NFS Data
NFS does not have a data-caching function, but the Virtual Memory Manager (VMM) caches pages of NFS
data just as it caches pages of disk data. If a system is essentially a dedicated NFS server, it may be
appropriate to permit the VMM to use as much memory as necessary for data caching. For a server
exporting JFS file systems, this is accomplished by setting the maxperm parameter, which controls the
maximum percentage of memory occupied by JFS file pages, to 100 percent.
For example:
# vmtune -P 100
On a server exporting Enhanced JFS file systems, this is accomplished by setting the maxclient
parameter. The maxclient parameter controls the maximum percentage of memory occupied by
client-segment pages which is where Enhanced JFS file data is cached.
For example:
# vmtune -t 100
The same technique could be used on NFS clients, but would only be appropriate if the clients were
running workloads that had very little need for working-segment pages.
Cache File System (CacheFS)
The Cache File System can be used to enhance performance of remote file systems or slow devices such
as CD-ROM. When a file system is cached, the data read from the remote file system or CD-ROM is
stored in a cache on the local system, thereby avoiding the use of the network and NFS server when the
same data is accessed for the second time. CacheFS is designed as a layered file system; this means
that CacheFS provides the ability to cache one file system (the NFS file system, also called the back-file
system) on another (your local file system, also called the front-file system), as shown in the following
figure:
300
Performance Management Guide
Figure 29. Cache File System (CacheFS). This illustration show a client machine and a server that are connect by a
network. The storage media on the server contains the back file system. The storage media on the client contains the
cached file system or the front file system.
Note that in AIX 4.3, only NFS Version 2 and NFS Version 3 are supported as back-file systems, and JFS
is the only supported front-file system.
CacheFS works as follows:
1. After creating a CacheFS file system on a client system, the system administrator specifies which file
systems are to be mounted in the cache.
2. When a user on the client attempts to access files that are part of the back file system, those files are
placed in the cache. The cache does not get filled until a user requests access to a file or files.
Therefore, the initial request to access a file will be at typical NFS speeds, but subsequent accesses to
the same file will be at local JFS speeds.
3. To ensure that the cached directories and files are kept up to date, CacheFS periodically checks the
consistency of files stored in the cache. It does so by comparing the current modification time to the
previous modification time.
4. If the modification times are different, all data and attributes for the directory or file are purged from the
cache, and new data and attributes are retrieved from the back file system.
An example where CacheFS would be suitable is in a CAD environment where master-copies of drawing
components can be held on the server and cached-copies on the client workstation when in use.
CacheFS does not allow reads and writes on files that are 2 GB or larger in size.
CacheFS Benefits
Because NFS data is cached on the local disk once it is read from the server, read requests to the NFS
file system can be satisfied much faster than if the data had to be retrieved over the net again. Depending
on the memory size and usage of the client, a small amount of data might be held and retrieved from
memory, so that the benefits of cached data on the disk applies to a large amount of data that cannot be
kept in memory. An additional benefit is that data on the disk cache will be held at system shutdown,
whereas data cached in memory will have to be retrieved from the server again after the reboot.
Chapter 11. Monitoring and Tuning NFS Use
301
Other potential NFS bottlenecks are a slow or busy network and a weak performing server with too many
NFS clients to serve. Therefore, access from the client system to the server is likely to be slow. CacheFS
will not prevent you from having to do the first read over the network and to access the server, but you
can avoid reads over the network for further requests for the same data.
If more read requests can be satisfied from the client’s local disk, the amount of NFS accesses to the
server will decrease. This means that more clients can be served by the server; thus, the client per server
ratio will increase.
Fewer read requests over the network will decrease your network load and, therefore, allow you to get
some relief on very busy networks or space for other data transfers.
Not every application will benefit from CacheFS. Because CacheFS will only speed up read performance,
mainly applications that have huge read requests for the same data over and over again will benefit from
CacheFS. Large CAD applications will certainly benefit from CacheFS, because of the often very large
models that they have to load for their calculations.
Performance tests showed that sequential reads from the CacheFS file system are 2.4 times to 3.4 times
faster than reads from the NFS server’s memory or disk.
What CacheFS Does Not Do
CacheFS will not increase the write performance to NFS file systems. However, you have two write
options to choose as parameters to the -o option of the mount command, when mounting a CacheFS.
They will influence the subsequent read performance to the data. The write options are as follows:
write-around
The write-around mode (the default) handles writes the same way that NFS does; that is, writes
are made to the back file system, and the affected file is purged from the cache. This means that
write-around voids the cache and new data must be obtained back from the server after the write.
non-shared
You can use the non-shared mode when you are certain that no one else will be writing to the
cached file system. In this mode, all writes are made to both the front and the back file system,
and the file remains in the cache. This means that future read accesses can be done to the cache,
rather than going to the server.
Small reads might be kept in memory anyway (depending on your memory usage); so there is no benefit
in also caching the data on the disk. Caching of random reads to different data blocks does not help,
unless you will access the same data over and over again.
The initial read request still has to go to the server because only by the time a user attempts to access
files that are part of the back file system will those files be placed in the cache. For the initial read request,
you will see typical NFS speed. Only for subsequent accesses to the same data, you will see local JFS
access performance.
The consistency of the cached data is only checked at intervals. Therefore, it is dangerous to cache data
that is frequently changed. CacheFS should only be used for read-only or read-mostly data.
Write Performance over a cached NFS file system differs from NFS Version 2 to NFS Version 3.
Performance tests have shown that:
v Sequential writes to a new file over NFS Version 2 to a CacheFS mount point can be 25 percent slower
than writes directly to the NFS Version 2 mount point.
v Sequential writes to a new file over NFS Version 3 to a CacheFS mount point can be 6 times slower
than writes directly to the NFS Version 3 mount point.
302
Performance Management Guide
Configuring CacheFS
CacheFS is not implemented by default or prompted at the time of the creation of an NFS file system.
Make sure to install fileset level 4.3.1.1 for enhanced CacheFS write performance over NFS in AIX 4.3.1.
The system administrator must specify explicitly which file systems are to be mounted in the cache as
follows:
1. Create the local cache file system by using the cfsadmin command:
# cfsadmin -c -o parameters cache-directory
where parameters specify the resource parameters and cache-directory is the name of the directory
where the cache should be created.
2. Mount the back file system onto the cache:
# mount -V cachefs -o backfstype=nfs,cachedir=/cache-directory remhost:/rem-directory local-mount-point
where rem-directory is the name of the remote host and file system where the data resides, and
local-mount-point is the mount point on the client where the remote file system should be mounted.
3. Alternately, you could administer CacheFS using the SMIT command (use the smitty cachefs fast
path).
Several parameters can be set at creation time, as follows:
maxblocks
Sets the maximum number of blocks that CacheFS is allowed to claim within the front file system.
Default = 90 percent.
minblocks
Sets the minimum number of blocks that CacheFS is allowed to claim within the front file system.
Default = 0 percent.
threshblocks
Sets the number of blocks that must be available in the JFS file system on the client side before
CacheFS can claim more than the blocks specified by minblocks. Default = 85 percent.
maxfiles
Maximum number of files that CacheFS can use, expressed as a percentage of the total number
of i-nodes in the front file system. Default = 90 percent.
minfiles
Minimum number of files that CacheFS is always allowed to use, expressed as a percentage of
the total number of i-nodes in the front file system. Default = 0 percent.
maxfilesize
Largest file size, expressed in megabytes, that CacheFS is allowed to cache. Default = 3.
RPC Tuning for NFS
The rpc.lockd daemon is multi-threaded and, by default, can create up to 33 threads. In situations where
there is heavy remote procedure call (RPC) file locking activity, the rpc.lockd daemon might become a
bottleneck once it reaches the maximum number of threads. When that maximum is reached, any
subsequent requests will have to wait, which may result in other timeouts. If there are more than one
client, the NFS server should have more lockd threads than the client side. The number of lockd threads
can be adjusted to a limit of 511 with the following:
# chsys -s rpc.lockd -a <# of threads>
# stopsrc -s rpc.lockd; startsrc -s rpc.lockd
Chapter 11. Monitoring and Tuning NFS Use
303
Tuning Other Layers to Improve NFS Performance
NFS uses UDP or TCP to perform its network I/O. Ensure that you have applied the tuning techniques
described in Tuning TCP and UDP Performance and Tuning mbuf Pool Performance. In particular, you
should:
v Ensure that the LAN adapter transmit and receive queues are set to the maximum (see Adapter
Transmit and Receive Queue Tuning).
v Increase the maximum socket buffer size (sb_max) to at least 131072 bytes. If the MTU size is not
4096 bytes or larger, set sb_max to at least 262144. Also set the UDP socket buffer sizes
(udp_sendspace and udp_recvspace) to 131072 bytes. If you are using TCP, set tcp_sendspace and
tcp_recvspace to this value.
v If possible, increase the MTU size on the LAN. On a 16 Mb Token-Ring, for example, an increase in
MTU size from the default 1492 bytes to 8500 bytes allows a complete 8 KB NFS read or write request
to be transmitted without fragmentation. It also makes much more efficient use of mbuf space, reducing
the probability of overruns.
v A tunable parameter called nfs_tcp_socketsize provides default window sizes for TCP ports used in
NFS connections. There is only one connect per client/server pair no matter how many mounts there
are. Do not set the nfs_tcp_socketsize value to less than 60,000. Large or busy servers should have
larger values until TCP NFS traffic shows no packets dropped from the output of the netstat -s -p tcp
command.
v The maximum number of TCP connections allowed into the server can be controlled by the new option
nfs_max_connections. The default of 0 indicates that there is no limit. The client will close TCP
connections that have been idle for approximately 5 minutes, and the connection is reestablished when
use warrants it. The server will close connections that have been idle for approximately 6 minutes.
v The operating system provides an option to turn off the UDP checksum for NFS only. You can use the
nfso command option, called udpchecksum. The default is 1 (checksum enabled). Slight performance
gains can be realized by turning it off, at the expense of increased chance of data corruption.
Increasing NFS Socket Buffer Size
In the course of tuning UDP, you may find that the netstat -s command indicates a significant number of
UDP socket buffer overflows. As with ordinary UDP tuning, increase the sb_max value. You also need
to increase the value of nfs_socketsize, which specifies the size of the NFS socket buffer. Following is an
example:
# no -o sb_max=131072
# nfso -o nfs_socketsize=130972
The previous example sequence sets sb_max to a value at least 100 bytes larger than the desired value
of nfs_socketsize and sets nfs_socketsize to 130972.
Note: In AIX Version 4, the socketsize is set dynamically. Configurations using the no and nfso command
must be repeated every time the machine is booted. Add them in the /etc/rc.net or /etc/rc.nfs file
immediately before the nfsd daemons are started and after the biod daemons are started. The
position is crucial.
NFS Server Disk Configuration
NFS servers that experience high levels of write activity can benefit from configuring the journal logical
volume on a separate physical volume from the data volumes. See Disk Preinstallation Guidelines for
further details.
It is often necessary to achieve high parallelism on data access. Concurrent access to a single file system
on a server by multiple clients or multiple client processes can result in throughput being bottlenecked on
the disk I/O for a particular device. You can use the iostat command to evaluate disk loading.
304
Performance Management Guide
For large NFS servers, the general strategy should be to divide evenly the disk I/O demand across as
many disk and disk adapter devices as possible. This results in greater parallelism and the ability to run
greater numbers of nfsd daemons. On a system where disk I/O has been well-distributed, it is possible to
reach a point where CPU load becomes the limiting factor on the server’s performance.
Misuses of NFS That Affect Performance
Many of the misuses of NFS occur because people do not realize that the files that they are accessing are
at the other end of an expensive communication path. A few examples of this are as follows:
v An application running on one system doing random updates of an NFS-mounted inventory file,
supporting a real-time retail cash register application.
v A development environment in which a source code directory on each system was NFS-mounted on all
of the other systems in the environment, with developers logging onto arbitrary systems to do editing
and compiles. This practically guaranteed that all of the compiles would be obtaining their source code
from, and writing their output to, remote systems.
v Running the ld command on one system to transform .o files in an NFS-mounted directory into an a.out
file in the same directory.
v Applications that issue writes that are not page-aligned (for example 10 KB). Writes less than 4 KB in
size will always result in a pagein and in the case of NFS, this pagein goes over the network.
It can be argued that these are valid uses of the transparency provided by NFS. Perhaps this is so, but
these uses do cost processor time and LAN bandwidth and degrade response time. When a system
configuration involves NFS access as part of the standard pattern of operation, the configuration designers
should be prepared to defend the consequent costs with offsetting technical or business advantages, such
as:
v Placing all of the data or source code on a server, rather than on individual workstations, improves
source-code control and simplifies centralized backups.
v A number of different systems access the same data, making a dedicated server more efficient than one
or more systems combining client and server roles.
Another type of application that should not be run across NFS file systems is an application that does
hundreds of lockf() or flock() calls per second. On an NFS file system, all the lockf() or flock() calls (and
other file locking calls) must go through the rpc.lockd daemon. This can severely degrade system
performance because the lock daemon may not be able to handle thousands of lock requests per second.
Regardless of the client and server performance capacity, all operations involving NFS file locking will
probably seem unreasonably slow. There are several technical reasons for this, but they are all driven by
the fact that if a file is being locked, special considerations must be taken to ensure that the file is
synchronously handled on both the read and write sides. This means there can be no caching of any file
data at the client, including file attributes. All file operations go to a fully synchronous mode with no
caching. Suspect that an application is doing network file locking if it is operating over NFS and shows
unusually poor performance compared to other applications on the same client/server pair.
NFS Tuning Checklist
Following is a checklist that you can follow when tuning NFS:
1. Check to see if you are overrunning the server.
The general method is to see if slowing down the client will increase the performance. The following
methods can be tried independently:
v Try running with just one biod daemon on the mount.
Try running with just one biod daemon on the affected client. If performance increases, then
something is being overrun either in the network or on the server. Run the stopsrc -s biod
command to stop all the SRC biod daemons. It will leave one kernel process biod with which you
can still run. See if it runs faster with just the one biod process. If it has no effect, restart the biod
Chapter 11. Monitoring and Tuning NFS Use
305
daemons with the startsrc -s biod command. If it runs faster, attempt to determine where the
packets are being dropped when all daemons are running. Networks, network devices, slow server,
overloaded server, or a poorly tuned server could all cause this problem.
v Try reducing the read/write sizes to see if things speed up.
v If you are using UDP, try using a TCP mount instead.
2. Check for Oerrs.
This is probably the most common under-configuration error that affects NFS servers.
If there are any Oerrs, increase the transmit queues for the network device. This can be done with
the machine running, but the interface must be detached before it is changed (rmdev -l). You cannot
shut down the interface on a diskless machine, and you may not be at liberty to shut down the
interface if other work is going on. In this case, you can use the chdev command to put the changes
in ODM so they will be activated on the next boot. Start by doubling the current value for the queue
length, and repeat the process (adding an additional 30 each time) until no Oerrs are reported. On
SP2 systems where NFS is configured to run across the high-speed switch, Oerrs may occur on the
switch when there has been a switch fault and the switch was temporarily unavailable. An error is
counted towards Oerr for each attempt to send on the switch that fails. These errors cannot be
eliminated.
3. Look for any Errors.
Any counts that are very large indicate problems.
4. Check for NFS UDP/TCP buffer overruns.
When using UDP, buffer overruns on the NFS server are another frequent under-configuration for
NFS servers. TCP sockets can be similarly overrun on very busy machines. Tuning for both is similar,
so this section discusses only UDP.
Run the netstat -s command and examine the UDP statistics for the socket buffer overflows
statistic. If it is anything other than 0, you are probably overrunning the NFS UDP buffer. Be aware,
however, that this is the UDP socket buffer drop count for the entire machine, and it may or may not
be NFS packets that are being dropped. You can confirm that the counts are due to NFS by
correlating between packet drop counts on the client using the nfsstat -cr command to socket buffer
overruns on the server while executing an NFS write test.
Socket buffer overflows can happen on heavily stressed servers or on servers that are slow in
relation to the client. Up to 10 socket buffer overflows are probably not a problem. Hundreds of
overflows are. If this number continually goes up while you watch it, and NFS is having performance
problems, NFS needs tuning.
Two factors can tune NFS socket buffer overruns. First try increasing the number of nfsd daemons
that are being run on the server. If that does not solve the problem, you must adjust two kernel
variables, sb_max(socket buffer max) and nfs_socketsize (the size of the NFS server socket buffer).
Use the no command to increase sb_max. Use the nfso command to increase the nfs_socketsize
variable.
The sb_max parameter must be set larger than nfs_socketsize. It is hard to suggest new values.
The best values are the smallest ones that also make the netstat report 0 or just produce a few
socket buffer overruns.
Remember, in AIX Version 4, the socket size is set dynamically. Configurations using the no and nfso
command must be repeated every time the machine is booted. Add them in the /etc/rc.nfs file, right
before the nfsd daemons are started and after the biod daemons are started. The position is crucial.
5. Check for mbuf problems.
See if there are any requests for mbufs denied or delayed. If so, increase the number of mbufs
available to the network. For more information about tuning to eliminate mbuf problems, see Tuning
mbuf Pool Performance.
6. Check for very small interpacket delays.
306
Performance Management Guide
There have been rare cases where this has caused problems with machines. If there is a router or
other hardware between the server and client, you can check its documentation to see if the
interpacket delays can be configured. If so, try increasing the delay.
7. Check for media mismatches.
When packets are traversing two media with widely different speeds, the router might drop packets
when taking them off the high speed net and trying to get them out on the slower net. This has been
seen particularly when a router was trying to take packets from a server on FDDI and send them to a
client on Ethernet. It could not send out the packets fast enough on Ethernet to keep up with the
FDDI. The only solution is to try to slow down the volume of client requests, use smaller read/write
sizes, and limit the number of biod daemons that can be used on a single file system.
8. Check for MTU mismatches.
Run the netstat -i command and check the MTU on the client and server. If they are different, try
making them the same and see if the problem is eliminated. Also be aware that slow or wide area
networks between the machines, routers, or bridges. may further fragment the packets to traverse
these network segments. Attempt to determine the smallest MTU between source and destination,
and change the rsize/wsize on the NFS mount to some number lower than that lowest-commondenominator MTU.
9. Check for routing problems.
Use the traceroute command to look for unexpected routing hops or delays.
10. Check for errlog entries.
Run the errpt command and look for reports of network device or network media problems. Also look
for any disk errors that might be affecting NFS server performance on exported file systems.
Chapter 11. Monitoring and Tuning NFS Use
307
308
Performance Management Guide
Chapter 12. Monitoring and Tuning Java
This chapter provides insights and general guidelines for isolating bottlenecks and tuning performance in
Java applications.
What is Java?
Java is an object-oriented programming language developed by Sun Microsystems, Inc.. It is modeled
after C++, and was designed to be small, simple, and portable across platforms and operating systems at
the source and at the binary level. Java programs (applets and applications) can therefore run on any
machine that has installed the Java Virtual Machine (JVM).
Why Java?
Java is a key component of Network Computing Framework (NCF), IBM’s software road map for
e-business. Java has significant advantages over other languages and other environments that make it
suitable for just about any programming task.
The advantages of Java are as follows:
v Java is platform-independent.
One of the most significant advantages of Java is its ability to move easily from one computer system to
another. Crucial to any World Wide Web software is the ability to run the same program on many
different systems, Java succeeds at this by being platform-independent at both the source and binary
levels.
v Java is object-oriented.
Another advantage of Java is the ability to take advantage of the object-oriented methodology. This
allows you to create modular programs and reusable code.
v Java is easy to learn.
Java was designed to be easy to use and is therefore easier to write, compile, debug, and learn.
v Java is the solution to e-business.
Because of Java’s robustness, ease of use, cross-platform capabilities and security features, it has
become a language of choice for providing worldwide Internet solutions.
Java Performance Guidelines
The following are basic Java performance issues:
v Use StringBuffer instead of string concatenations, when doing excessive string manipulations to avoid
unnecessarily creating objects that eventually must undergo garbage collection.
v Avoid excessive writing to the Java console to reduce the cost of string manipulations, text formatting,
and output.
v Avoid the costs of object creation and manipulation by using primitive types for variables when
necessary.
v Cache frequently used objects to reduce the amount of garbage collection needed, and avoid the need
to repeatedly create the objects.
v Group native operations to reduce the number of Java Native Interface (JNI) calls when possible.
v Use synchronized methods only when necessary, to limit the multitasking in the JVM and operating
system.
v Avoid invoking the garbage collector unless necessary. If you must invoke it, do so only during idle time
or some noncritical phase.
v Use int instead of long whenever possible, because 32-bit operations are executed faster than 64-bit.
© Copyright IBM Corp. 1997, 2002
309
v Declare methods as final whenever possible. Final methods are handled better by the JVM.
v Use the key word static final when creating constants in order to reduce the number of times the
variables need to be initialized.
v Avoid unnecessary ″casts″ and ″instanceof″ references because casting in Java is not done at compile
time but at run time.
v Avoid the use of vectors whenever possible when an array will suffice.
v Add and delete items from the end of the vector for better performance.
v Compile Java files with the -O option.
v Avoid allocating objects within loops.
v Use buffer I/O and tune the buffer size.
v Use connection pools and cached-prepared statements for database access.
v Use connection pools to the database and reuse connections rather than repeatedly opening and
closing connections.
v Maximize thread lifetimes and minimize thread creation and destruction cycles.
v
v
v
v
v
Minimize contention for shared resources.
Minimize creation of short lived objects.
Avoid remote method calls.
Use callbacks to avoid blocking remote method calls.
Avoid creating an object that would only be used for accessing a method.
v Whenever possible, keep synchronized methods out of loops.
v Store string and char data as Unicode in the database.
v Reorder CLASSPATH so that the most used libraries occur first.
Monitoring Java
There are several tools to monitor and identify performance inhibitors in your Java application.
vmstat
This command provides information about various system resources. It reports statistics on kernel
threads in the run queue as well as in the wait queue, memory usage, paging space, disk I/O,
interrupts, system calls, context switches, and CPU activity.
iostat This command reports detailed disk I/O information.
topas This command reports CPU, network, disk I/O, Workload Manager and process activity.
tprof
This command can be used to profile the application to pinpoint any hot routines/methods which
often can be considered performance problems.
ps -mo THREAD
This command shows to which CPU a process or thread is bound.
Java profilers [-Xrunhprof, Xrunjpa64 (64–bit kernel), -Xrunjpa 32–bit kernel)]
Used to determine which routines or methods are the most heavily used.
java -verbose:gc
This option can be used to check the impact of garbage collection on your application. It reports
total time spent doing garbage collection, average time per garbage collection, average memory
collected per garbage collection, and average objects collected per garbage collection.
Tuning Java
The following are recommended AIX settings for your JAVA environment.
310
Performance Management Guide
AIXTHREAD_SCOPE=S
Starting with AIX 4.3.1, the default value for this variable is P. This signifies process-wide
contention scope (M:N). For Java applications, you should set this value to S, which signifies
system-wide contention scope (1:1).
AIXTHREAD_MUTEX_DEBUG=OFF
Maintains a list of active mutexes for use by the debugger.
AIXTHERAD_COND_DEBUG=OFF
Maintains a list of condition variables for use by the debugger.
AIXTHREAD_RWLOCK_DEBUG=OFF
The pthreads library maintains a list of active mutual exclusion locks, condition variables, and
read-write locks for use by the debugger. When a lock is initialized, it is added to the list if it is not
there already. This list is implemented as a linked list, so searching it to determine if a lock is
present or not has a performance implication when the list gets large. The problem is compounded
by the fact that the list is protected by a lock, which is held for the duration of the search
operation. Other calls to the pthread_mutex_init() subroutine have to wait while the search is
done. For optimal performance, you should set the value of this thread-debug options to OFF. Their
default is ON
SPINLOOPTIME=500
The spinloop time is the number of times that a process can spin on a busy lock before blocking.
This value is set to 40 by default. If the tprof output indicates high CPU usage for the check_lock
routine, and if locks are usually available within a short amount of time, you should increase the
spin time by setting the value to 500 or higher.
Also, the following settings are recommended for your Java environment:
ulimit -d unlimited
ulimit -m unlimited
ulimit -n unlimited
ulimit -s unlimited
Certain environment parameters and settings can be used to tune Java performance within the
operating system. In addition, many of the techniques for tuning system components, such as
CPU, memory, network, I/O, and so on, can serve to increase Java performance. To determine
which may be beneficial to your situation, refer to the specific sections in this book.
To obtain the best possible Java performance and scalability, you should use the latest available
versions for operating system and Java, as well as for your Just-In-Time (JIT) compiler .
Performance Implications for Garbage Collection
The most common performance problem associated with Java relates to the Garbage Collection
mechanism. If the size of the Java heap is too large, the heap will have to reside outside of main memory.
This would cause increased paging activity, which would affect Java performance. Also, a large heap can
take several seconds to fill up. This means that, although Garbage Collection would occur less frequently,
pause times associated to Garbage Collection will increase. To tune the Java Virtual Machine (JVM) heap,
use the java command with option -ms or -mx. Use the Garbage Collection statistics to help determine
optimal settings.
Chapter 12. Monitoring and Tuning Java
311
312
Performance Management Guide
Chapter 13. Analyzing Performance with the Trace Facility
The operating system’s trace facility is a powerful system-observation tool. The trace facility captures a
sequential flow of time-stamped system events, providing a fine level of detail on system activity. Events
are shown in time sequence and in the context of other events. Trace is a valuable tool for observing
system and application execution. Unlike other tools that only provide CPU utilization or I/O wait time,
trace expands that information to aid in understanding what events are happening, who is responsible,
when the events are taking place, how they are affecting the system and why.
The operating system is instrumented to provide general visibility to system execution. Users can extend
visibility into their applications by inserting additional events and providing formatting rules.
Care was taken in the design and implementation of this facility to make the collection of trace data
efficient, so that system performance and flow would be minimally altered by activating trace. Because of
this, the trace facility is extremely useful as a performance-analysis tool and as a problem-determination
tool.
The following sections provide more information on the trace facility:
v Understanding the Trace Facility
v Example of Trace Facility Use
v
v
v
v
Starting and Controlling Trace from the Command Line
Starting and Controlling Trace from a Program
Using the trcrpt Command to Format a Report
Adding New Trace Events
Understanding the Trace Facility
The trace facility is more flexible than traditional system-monitor services that access and present statistics
maintained by the system. It does not presuppose what statistics will be needed, instead, trace supplies a
stream of events and allows the user to decide what information to extract. With traditional monitor
services, data reduction (conversion of system events to statistics) is largely coupled to the system
instrumentation. For example, many systems maintain the minimum, maximum, and average elapsed time
observed for executions of task A and permit this information to be extracted.
The trace facility does not strongly couple data reduction to instrumentation, but provides a stream of trace
event records (usually abbreviated to events). It is not necessary to decide in advance what statistics will
be needed; data reduction is to a large degree separated from the instrumentation. The user may choose
to determine the minimum, maximum, and average time for task A from the flow of events. But it is also
possible to:
v Extract the average time for task A when called by process B
v Extract the average time for task A when conditions XYZ are met
v Calculate the standard deviation of run time for task A
v Decide that some other task, recognized by a stream of events, is more meaningful to summarize.
This flexibility is invaluable for diagnosing performance or functional problems.
In addition to providing detailed information about system activity, the trace facility allows application
programs to be instrumented and their trace events collected in addition to system events. The trace file
then contains a complete record of the application and system activity, in the correct sequence and with
precise time stamps.
© Copyright IBM Corp. 1997, 2002
313
Implementation
A trace hook is a specific event that is to be monitored. A unique number is assigned to that event called a
hook ID. The trace command monitors these hooks.
The trace command generates statistics on user processes and kernel subsystems. The binary
information is written to two alternate buffers in memory. The trace process then transfers the information
to the trace log file on disk. This file grows very rapidly. The trace program runs as a process which may
be monitored by the ps command. The trace command acts as a daemon, similar to accounting.
The following figure illustrates the implementation of the trace facility.
Figure 30. Implementation of the Trace Facility. This illustration shows the trace process. In this process, the user
process (kernel subsystems) sends trace hook calls to trace buffers labled A and B. From the buffers, they pass
through the trace driver and on to the trace log file of the user kernel.
Monitoring facilities use system resources. Ideally, the overhead should be low enough as to not
significantly affect system execution. When the trace program is active, the CPU overhead is less than 2
percent. When the trace data fills the buffers and must be written to the log, additional CPU is required for
file I/O. Usually this is less than 5 percent. Because the trace program claims and pins buffer space, if the
environment is memory-constrained, this might be significant. Be aware that the trace log and report files
can become very large.
Limiting the Amount of Trace Data Collected
The trace facility generates large volumes of data. This data cannot be captured for extended periods of
time without overflowing the storage device. There are two ways to use the trace facility efficiently:
v The trace facility can be turned on and off in multiple ways to capture system activity. It is practical to
capture in this way seconds to minutes of system activity for post processing. This is enough time to
characterize major application transactions or interesting sections of a long task.
v The trace facility can be configured to direct the event stream to standard output. This allows a real-time
process to connect to the event stream and provide data reduction as the events are recorded, thereby
creating long-term monitoring capability. A logical extension for specialized instrumentation is to direct
the data stream to an auxiliary device that can either store massive amounts of data or provide dynamic
data reduction. This technique is used by the performance tools tprof, pprof, netpmon, and filemon.
Starting and Controlling Trace
The trace facility provides three distinct modes of use:
314
Performance Management Guide
Subcommand Mode
Trace is started with a shell command (trace) and carries on a dialog with the user through
subcommands. The workload being traced must be provided by other processes, because the
original shell process is in use.
Command Mode
Trace is started with a shell command (trace -a) that includes a flag which specifies that the trace
facility is to run asynchronously. The original shell process is free to run ordinary commands,
interspersed with trace-control commands.
Application-Controlled Mode
Trace is started with the trcstart() subroutine and controlled by subroutine calls such as trcon()
and trcoff() from an application program.
Formatting Trace Data
A general-purpose trace-report facility is provided by the trcrpt command. The report facility provides little
data reduction, but converts the raw binary event stream to a readable ASCII listing. Data can be visually
extracted by a reader, or tools can be developed to further reduce the data.
The report facility displays text and data for each event according to rules provided in the trace format file.
The default trace format file is /etc/trcfmt, which contains a stanza for each event ID. The stanza for the
event provides the report facility with formatting rules for that event. This technique allows users to add
their own events to programs and insert corresponding event stanzas in the format file to specify how the
new events should be formatted.
Viewing Trace Data
When trace data is formatted, all data for a given event is usually placed on a single line. Additional lines
may contain explanatory information. Depending on the fields included, the formatted lines can easily
exceed 80 characters. It is best to view the reports on an output device that supports 132 columns.
Example of Trace Facility Use
The following takes you through an example of a typical trace.
Note: This example is more meaningful if the input file is not already cached in system memory. Choose
as the source file any file that is about 50 KB in size and has not been used recently.
Obtaining a Sample Trace File
Trace data accumulates rapidly. Bracket the data collection as closely around the area of interest as
possible. One technique for doing this is to issue several commands on the same command line. For
example:
# trace -a -k "20e,20f" -o trc_raw ; cp ../bin/track /tmp/junk ; trcstop
captures the execution of the cp command. We have used two features of the trace command. The -k
″20e,20f″ option suppresses the collection of events from the lockl() and unlockl() functions. These calls
are numerous on uniprocessor systems, but not on SMP systems, and add volume to the report without
giving us additional information. The -o trc_raw option causes the raw trace output file to be written in our
local directory.
Formatting the Sample Trace
We use the following form of the trcrpt command for our report:
# trcrpt -O "exec=on,pid=on" trc_raw > cp.rpt
This reports both the fully qualified name of the file that is run and the process ID that is assigned to it.
Chapter 13. Analyzing Performance with the Trace Facility
315
The report file shows us that there are numerous VMM page assign and delete events in the trace, like the
following sequence:
1B1 ksh
8526
0.003109888
0.162816
VMM page delete:
V.S=0000.150E ppage=1F7F
working_storage delete_in_progress process_private computational
1B0 ksh
8526
0.003141376
0.031488
VMM page assign:
V.S=0000.2F33 ppage=1F7F
working_storage delete_in_progress process_private computational
We are not interested in this level of VMM activity detail at the moment, so we reformat the trace as
follows:
# trcrpt -k "1b0,1b1" -O "exec=on,pid=on" trc_raw > cp.rpt2
The -k ″1b0,1b1″ option suppresses the unwanted VMM events in the formatted output. It saves us from
having to retrace the workload to suppress unwanted events. We could have used the -k function of the
trcrpt command instead of that of the trace command to suppress the lockl() and unlockl() events, if we
had believed that we might need to look at the lock activity at some point. If we had been interested in
only a small set of events, we could have specified -d ″hookid1,hookid2″ to produce a report with only
those events. Because the hook ID is the leftmost column of the report, you can quickly compile a list of
hooks to include or exclude. A comprehensive list of trace hook IDs is defined in the
/usr/include/sys/trchkid.h file.
Reading a Trace Report
The header of the trace report tells you when and where the trace was taken, as well as the command
that was used to produce it:
Thu Oct 28 13:34:05 1999
System: AIX texmex Node: 4
Machine: 000691854C00
Internet Protocol Address: 09359BBB 9.53.155.187
Buffering: Kernel Heap
trace -a -k 20e,20f -o trc_raw
The body of the report, if displayed in a small enough font, looks similar to the following:
ID
101
101
134
In
v
v
v
PROCESS NAME PID
ksh
8526
ksh
7214
cp
7214
ELAPSED_SEC
0.005833472
0.012820224
0.014451456
DELTA_MSEC
0.107008
0.031744
0.030464
APPL
SYSCALL KERNEL INTERRUPT
kfork LR = D0040AF8
execve LR = 10015390
exec: cmd=cp ../bin/track /tmp/junk pid=7214 tid=24713
cp.rpt2 you can see the following information:
The fork(), exec(), and page fault activities of the cp process.
The opening of the input file for reading and the creation of the /tmp/junk file
The successive read()/write() system calls to accomplish the copy.
v The process cp becoming blocked while waiting for I/O completion, and the wait process being
dispatched.
v How logical-volume requests are translated to physical-volume requests.
v The files are mapped rather than buffered in traditional kernel buffers, and the read accesses cause
page faults that must be resolved by the Virtual Memory Manager.
v The Virtual Memory Manager senses sequential access and begins to prefetch the file pages.
v The size of the prefetch becomes larger as sequential access continues.
v When possible, the disk device driver coalesces multiple file requests into one I/O request to the drive.
The trace output looks a little overwhelming at first. This is a good example to use as a learning aid. If you
can discern the activities described, you are well on your way to being able to use the trace facility to
diagnose system-performance problems.
316
Performance Management Guide
Filtering of the Trace Report
The full detail of the trace data may not be required. You can choose specific events of interest to be
shown. For example, it is sometimes useful to find the number of times a certain event occurred. To
answer the question ″How many opens occurred in the copy example?″ first find the event ID for the
open() system call. This can be done as follows:
# trcrpt -j | grep -i open
You should be able to see that event ID 15b is the OPEN SYSTEM CALL event. Now, process the data
from the copy example as follows:
# trcrpt -d 15b -O "exec=on" trc_raw
The report is written to standard output, and you can determine the number of open() subroutines that
occurred. If you want to see only the open() subroutines that were performed by the cp process, run the
report command again using the following:
# trcrpt -d 15b -p cp -O "exec=on" trc_raw
Starting and Controlling Trace from the Command Line
The trace facility is configured and data collection optionally started by the trace command, the detailed
syntax of which is described in the AIX 5L Version 5.2 Commands Reference.
After trace is configured by the trace command, there are controls to turn data collection on and off and to
stop the trace facility (the trcstop subcommand deconfigures trace and unpins buffers). You can invoke
the controls through: subcommands, commands, and subroutines. The subroutine interfaces are described
in Starting and Controlling Trace from a Program.
Controlling Trace in Subcommand Mode
If the trace routine is configured without the -a option, it runs in subcommand mode. Instead of the normal
shell prompt, a prompt of ″->″ displays. In this mode, the following subcommands are recognized:
trcon
Starts or resumes collection of event data
trcoff
Suspends collection of event data
q or quit
Stops collection of event data and terminates the trace routine
!command
Runs the specified shell command
?
Displays the available commands
For example:
# trace -f -m "Trace of events during mycmd"
-> !mycmd
-> q
#
Controlling Trace by Commands
If the trace routine is configured to run asynchronously (trace -a), trace can be controlled by the following
commands:
trcon
Starts or resumes collection of event data
trcoff
Suspends collection of event data
trcstop
Stops collection of event data and terminates the trace routine
Chapter 13. Analyzing Performance with the Trace Facility
317
For example:
#
#
#
#
trace -a -n -L 2000000 -T 1000000 -d -o trace.out
trcon
cp /a20kfile /b
trcstop
By specifying the -d (defer tracing until the trcon subcommand is entered) option, you can limit how much
tracing is done on the trace command itself. If the -d option is not specified, then tracing begins
immediately and can log events for the trace command initializing its own memory buffers. Typically, we
want to trace everything but the trace command itself.
By default, the kernel buffer size (-T option) can be at most one half of the log buffer size (-L option). If
you use the -f flag, the buffer sizes can be the same.
The -n option is useful if there are kernel extension system calls that need to be traced.
Starting and Controlling Trace from a Program
The trace facility can be started from a program, through a subroutine call. The subroutine is trcstart() and
is in the librts.a library. The syntax of the trcstart() subroutine is as follows:
int trcstart(char *args)
where args is the options list that you would have entered for the trace command. By default, the system
trace (channel 0) is started. If you want to start a generic trace, include a -g option in the args string. On
successful completion, the trcstart() subroutine returns the channel ID. For generic tracing, this channel ID
can be used to record to the private generic channel.
When compiling a program using this subroutine, the link to the librts.a library must be specifically
requested (use -l rts as a compile option).
Controlling Trace with Trace Subroutine Calls
The controls for the trace routine are available as subroutines from the librts.a library. The subroutines
return zero on successful completion. The subroutines are:
int trcon()
Begins or resumes collection of trace data.
int trcoff()
Suspends collection of trace data.
int trcstop()
Stops collection of trace data and terminates the trace routine.
Using the trcrpt Command to Format a Report
The trace report facility reads the trace log file, formats the trace entries, and writes a report. The trcrpt
command displays text and data for each event according to rules provided in the trace format file
(/etc/trcfmt). Stanzas in the format file provide formatting rules for events or hooks. Users adding hooks to
programs can insert corresponding event stanzas in the format file to print their trace data (see Adding
New Trace Events).
The trcrpt facility does not produce any summary reports, but you can use the awk command to create
simple summaries through further processing of the trcrpt output.
The detailed syntax of the trcrpt command is described in the AIX 5L Version 5.2 Commands Reference.
318
Performance Management Guide
Formatting a Report on the Same System
The trcrpt command formats reports of trace event data contained in the trace log file. You can specify the
events to be included (or omitted) in the report, as well as determine the presentation of the output with
this command.
You can use the System Management Interface Tool (SMIT) to run the trcrpt command by typing the
SMIT fast path:
# smitty trcrpt
To create a trace report to the newfile file, type:
# trcrpt -o newfile
Formatting a Report on a Different System
It is often desirable to run the trcrpt command on another system than the system where the trace is
collected. There may be various reasons for this, such as:
v The system being traced might not be available for you to run the trcrpt command, and the trace might
be collected by the system administrator or someone at the remote site.
v The system being traced is too busy for you to run the trcrpt command.
v The system being traced does not have enough file system space left to accommodate a very large
trcrpt file.
You can run the trace command on a system and run the trcrpt command on that trace file on a different
system. In order for this to work correctly, the output of the trcnm command is needed from the system
where the trace was run. Run the trcnm command and redirect the output into a file, as follows:
# trcnm > trace.nm
If you want to use the trace file for other performance tools such as tprof, pprof, netpmon, and filemon,
run the gennames Gennames_File command.
That file is then used with the -n flag of the trcrpt command, as follows:
# trcrpt -n trace.nm -o newfile
If -n is not specified, then the trcrpt command generates a symbol table from the system on which the
trcrpt command is run.
Additionally, a copy of the /etc/trcfmt file from the system being traced might be beneficial bacause that
system may have different or more trace format stanzas than the system where the trcrpt command is
being run. The trcrpt command can use the -t flag to specify the trace format file (by default it uses the
/etc/trcfmt file from the system where the trcrpt command is being run). For example:
# trcrpt -n trace.nm -t trcfmt_file -o newfile
Formatting a Report from trace -C Output
If trace was run with the -C flag, one or more trace output files are generated. For example, if the trace file
name was specified as trace.out and -C all was specified on a 4-way SMP, then a trace.out, trace.out-1,
trace.out-2, trace.out-3, and trace.out-4 file was generated. When you run the trcrpt command, specify
trcrpt -C all and trace.out as the file name, and all the files will be read, as follows:
# trcrpt -C all -r trace.out > trace.tr
This trace.tr file can then be used as input for other commands (it will include the trace data from each
CPU). The reason for the -C flag on trace is so that the trace can keep up with each CPU’s activities on
those systems which have many CPUs (more than 12, for example). Another reason is that the buffer size
for the trace buffers is per CPU when you use the -C all flag.
Chapter 13. Analyzing Performance with the Trace Facility
319
Adding New Trace Events
The operating system is shipped instrumented with key events. The user need only activate trace to
capture the flow of events from the operating system. Application developers may want to instrument their
application code during development for tuning purposes. This provides them with insight into how their
applications are interacting with the system.
To add a trace event, you must design the trace records generated by your program in accordance with
trace interface conventions. You then add trace-hook macros to the program at the appropriate locations.
Traces can then be taken through any of the standard ways of invoking and controlling trace (commands,
subcommands, or subroutine calls). To use the trcrpt program to format your traces, add stanzas
describing each new trace record and its formatting requirements to the trace format file.
Possible Forms of a Trace Event Record
A trace event can take several forms. An event consists of a hook word, optional data words, and an
optional time stamp, as shown in the following figure. A four-bit type is defined for each form that the event
record can take. The type field is imposed by the recording routine so that the report facility can always
skip from event to event when processing the data, even if the formatting rules in the trace format file are
incorrect or missing for that event.
Figure 31. Format of a Trace Event Record. This illustration is a table containing 7 rows. Cells in the first row are
labeled 12–bit hook ID, 4–bit Type and 16–bit Data Field. The next 6 rows are simply labeled Data Word 1 through
Data Word 5, and the last row is labeled 32–bit Time Stamp. A row heading for row 1 is Hook Word (required). The
next 5 rows are labeled D1 (optional), D2 (optional), D3 (optional), D4 (optional), and (optional). The last row is
labeled T (optional).
An event record should be as short as possible. Many system events use only the hook word and time
stamp. The data words should seldom be used because using them is less efficient and is intrusive. A long
format allows the user to record a variable length of data. In this long form, the 16-bit data field of the
hook word is converted to a length field that describes the length of the event record.
Trace Channels
The trace facility can accommodate up to eight simultaneous channels of trace-hook activity, which are
numbered 0-7. Channel 0 is always used for system events, but application events can also use it. The
other seven channels, called generic channels, can be used for tracing application-program activity.
When trace is started, channel 0 is used by default. A trace -n channel_number command starts trace to a
generic channel. Use of the generic channels has some limitations:
320
Performance Management Guide
v The interface to the generic channels costs more CPU time than the interface to channel 0 because of
the need to distinguish between channels and because generic channels record variable-length records.
v Events recorded on channel 0 and on the generic channels can be correlated only by time stamp, not
by sequence, so there may be situations in which it is not possible to determine which event occurred
first.
Macros for Recording Trace Events
Macros to record each possible type of event record are defined in the /usr/include/sys/trcmacros.h file.
The event IDs are defined in the /usr/include/sys/trchkid.h file. Include these two files in any program
that is recording trace events.
The macros to record events on channel 0 with a time stamp are a follows:
TRCHKL0T(hw)
TRCHKL1T(hw,D1)
TRCHKL2T(hw,D1,D2)
TRCHKL3T(hw,D1,D2,D3)
TRCHKL4T(hw,D1,D2,D3,D4)
TRCHKL5T(hw,D1,D2,D3,D4,D5)
Similarly, to record events on channel 0 without a time stamp, use:
TRCHKL0(hw)
TRCHKL1(hw,D1)
TRCHKL2(hw,D1,D2)
TRCHKL3(hw,D1,D2,D3)
TRCHKL4(hw,D1,D2,D3,D4)
TRCHKL5(hw,D1,D2,D3,D4,D5)
The type field of the trace event record is set to the value that corresponds to the macro used, regardless
of the value of those 4 bits in the hw parameter.
Only two macros record events to one of the generic channels (1-7). These are as follows:
TRCGEN(ch,hw,D1,len,buf)
TRCGENT(ch,hw,D1,len,buf)
These macros record in the event stream specified by the channel parameter (ch) a hook word (hw), a data
word (D1) and len bytes from the user’s data segment beginning at the location specified by buf.
Use of Event IDs
The event ID in a trace record identifies that record as belonging to a particular class of records. The
event ID is the basis on which the trace mechanism records or ignores trace hooks, as well as the basis
on which the trcrpt command includes or excludes trace records in the formatted report.
Event IDs are 12 bits (three hexadecimal digits) for a possible 4096 IDs. Event IDs that are reserved and
shipped with code are permanently assigned to avoid duplication. To allow users to define events in their
environments or during development, the range of event IDs from hex 010 through hex 0FF has been
reserved for temporary use. Users can freely use IDs in this range in their own environment (that is, any
set of systems within which the users are prepared to ensure that the same event ID is not used
ambiguously).
Note: It is important that users who make use of this event range do not let the code leave their
environment. If you ship code instrumented with temporary hook IDs to an environment in which
you do not control the use of IDs, you risk collision with other programs that already use the same
IDs in that environment.
Chapter 13. Analyzing Performance with the Trace Facility
321
Event IDs should be conserved because there are so few of them, but they can be extended by using the
16-bit Data Field. This yields a possible 65536 distinguishable events for every formal hook ID. The only
reason to have a unique ID is that an ID is the level at which collection and report filtering are available in
the trace facility.
A user-added event can be formatted by the trcrpt command if there is a stanza for the event in the
specified trace format file. The trace format file is an editable ASCII file (see Syntax for Stanzas in the
Trace Format File).
Examples of Coding and Formatting Events
The following example shows the use of trace events to time the execution of a program loop:
#include <sys/trcctl.h>
#include <sys/trcmacros.h>
#include <sys/trchkid.h>
char *ctl_file = "/dev/systrctl";
int ctlfd;
int i;
main()
{
printf("configuring trace collection \n");
if (trcstart("-ad")){
perror("trcstart");
exit(1);
}
printf("opening the trace device \n");
if((ctlfd = open(ctl_file,0))<0){
perror(ctl_file);
exit(1);
}
printf("turning trace on \n");
if(ioctl(ctlfd,TRCON,0)){
perror("TRCON");
exit(1);
}
for(i=1;i<11;i++){
TRCHKL1T(HKWD_USER1,i);
}
/* The code being measured goes here. The interval */
/* between occurrences of HKWD_USER1 in the trace */
/* file is the total time for one iteration.
*/
printf("turning trace off\n");
if(ioctl(ctlfd,TRCSTOP,0)){
perror("TRCOFF");
exit(1);
}
printf("stopping the trace daemon \n");
if (trcstop(0)){
perror("trcstop");
exit(1);
}
}
exit(0);
When you compile the sample program, you must link to the librts.a library as follows:
# xlc -O3 sample.c -o sample -l rts
322
Performance Management Guide
HKWD_USER1 is event ID 010 hexadecimal (you can verify this by examining the
/usr/include/sys/trchkid.h file). The report facility does not format the HKWD_USER1 event, unless rules
are provided in the trace format file. The following example of a stanza for HKWD_USER1 could be used:
# User event HKWD_USER1 Formatting Rules Stanza
# An example that will format the event usage of the sample program
010 1.0 L=APPL "USER EVENT - HKWD_USER1" O2.0
\n \
"The # of loop iterations =" U4
\n \
"The elapsed time of the last loop = " \
endtimer(0x010,0x010) starttimer(0x010,0x010)
When you enter the example stanza, do not modify the master format file /etc/trcfmt, but instead make a
copy and keep it in your own directory (assume you name it mytrcfmt). When you run the sample
program, the raw event data is captured in the default log file because no other log file was specified to
the trcstart() subroutine. You can filter the output report to get only your events. To do this, run the trcrpt
command as follows:
# trcrpt -d 010 -t mytrcfmt -O "exec=on" > sample.rpt
You can browse the sample.rpt file to see the result.
Syntax for Stanzas in the Trace Format File
The trace format file provides rules for presentation and display of the expected data for each event ID.
This allows new events to be formatted without changing the report facility. Rules for new events are
simply added to the format file. The syntax of the rules provides flexibility in the presentation of the data.
A trace format stanza can be as long as required to describe the rules for any particular event. The stanza
can be continued to the next line by terminating the present line with a ’\’ character. The fields are
described in the AIX 5L Version 5.2 Files Reference.
Comments in the /etc/trcfmt file describe other format and macro possibilities and describe how a user
can define additional macros.
Chapter 13. Analyzing Performance with the Trace Facility
323
324
Performance Management Guide
Chapter 14. Using Performance Diagnostic Tool (PDT)
PDT attempts to identify performance problems automatically by collecting and integrating a wide range of
performance, configuration, and availability data. The data is regularly evaluated to identify and anticipate
common performance problems.
PDT assesses the current state of a system and tracks changes in workload and performance. It attempts
to identify incipient problems and suggest solutions before the problems become critical.
For the most part, PDT functions with no required user input. PDT data collection and reporting are easily
enabled, and then no further administrator activity is required. Periodically, data is collected and recorded
for historical analysis, and a report is produced and mailed to the adm user. Normally, only the most
significant apparent problems are recorded on the report. If there are no significant problems, that fact is
reported. PDT can be customized to direct its report to a different user or to report apparent problems of a
lower severity level.
This chapter contains the following main sections:
v Structure of PDT
v Scope of PDT Analysis
v Analyzing the PDT Report
v Installing and Enabling PDT
v Customizing PDT
v Responding to PDT Report Messages
Structure of PDT
As shown in the following illustration, the PDT application consists of three components:
v The collection component comprises a set of programs that periodically collect and record data.
v The retention component periodically reviews the collected data and discards data that is obsolete.
v The reporting component periodically produces a diagnostic report from the current set of historical
data.
© Copyright IBM Corp. 1997, 2002
325
Figure 32. PDT Component Structure. This illustration is similar to an organizational chart with the three PDT
application components at the top. They are; collection control, retention control and reporting control. A data path is
drawn from collection control to collectors and then on to PDT History. A data path is drawn from retention control
directly to PDT History. Data travels from Reporting control to the reporter and a periodic report is sent by file or mail.
Data in PDT History can be discarded when it is obsolete or passed on to the reporter where periodic reports are sent
by file or mail. All data is originated by the cron.
PDT considers various aspects of a system’s configuration, availability, and delivered performance in
making its assessment. In particular, areas of configuration imbalance are sought out (such as
I/O-configuration balance, paging-configuration balance) as well as other configuration problems (for
example, disks not allocated to volume groups). A wide variety of trending assessments is made, including
file sizes, file-system sizes, paging-area usage, network delays, and workload-related delays.
Scope of PDT Analysis
PDT collects configuration, availability, workload, and performance data on a daily basis. This data is
maintained in a historical record. Approximately one month’s data is kept in this way. Also on a daily basis,
PDT generates a diagnostic report, which is mailed to the adm user.
In addition to mailing the report, PDT stores a copy in the /var/perf/tmp/PDT_REPORT file. Before the
new report is written, the previous report is renamed /var/perf/tmp/PDT_REPORT.last.
While many common system performance problems are of a specific nature (a system might have too little
memory), PDT also attempts to apply some general concepts of well-performing systems to its search for
problems. Some of these concepts, together with examples of their application to the operating system,
are as follows:
v Balanced Use of Resources
In general, if there are several resources of the same type, then a balanced use of those resources
produces better performance.
– Comparable numbers of physical volumes (disks) on each disk adapter
– Paging space distributed across multiple physical volumes
– Roughly equal measured load on different physical volumes
326
Performance Management Guide
v Operation within Bounds
Resources have limits to their use. Trends attempting to exceed those limits should be detected and
reported.
– A disk drive cannot be used more than 100 percent of the time.
– File and file-system sizes cannot exceed the allocated space.
v Identified Workload Trends
Trends can indicate a change in the nature of the workload, as well as increases in the amount of
resource used:
– Number of users logged on
– Total number of processes
– CPU-idle percentage
v Error-Free Operation
Hardware or software errors often produce performance problems:
– Check the hardware and software error logs.
– Report bad VMM pages.
v Changes Investigated
New workloads or processes that start to consume resources may be the first sign of a problem.
v Appropriate Setting of System Parameters
There are many parameters in a system, and they must be set correctly.
– Is maxuproc set too low?
– Are the memory-load-control-parameter settings too high or too low?
PDT normally uses less than 30 seconds of CPU time. Daily data collection takes several elapsed
minutes.
Analyzing the PDT Report
A PDT report consists of several sections and differs depending on the severity level chosen (see PDT
Severity Levels).
Header
The header section indicates the release number of PDT, the date the report was printed, the host from
which the data was collected, and the range of dates of the data that fed the analysis. The content of this
section is consistent for all severity levels.
An example of a report header follows:
Performance Diagnostic Facility 1.0
Report printed: Thu Feb 10 17:54:50 2000
Host name: itsosmp.itsc.austin.ibm.com
Range of analysis includes measurements
from: Hour 12 on Tuesday, February 8th, 2000
to: Hour 17 on Wednesday, February 9th, 2000
Notice: To disable/modify/enable collection or reporting
execute the pdt_config script as root
Chapter 14. Using Performance Diagnostic Tool (PDT)
327
Alerts
The Alerts section focuses on identified violations of applied concepts and thresholds. The following
subsystems may have problems when they appear in the Alerts section: file systems, I/O configuration,
paging configuration, I/O balance, paging space, virtual memory, real memory, processes, and network.
For severity 1 levels, alerts focus on file systems, physical volumes, paging, and memory. If severity 2 or 3
is selected, information on configuration and processes is added.
An example of an Alerts section follows:
------------------------ Alerts --------------------I/O CONFIGURATION
- Note: volume hdisk1 has 872 MB available for allocation
while volume hdisk0 has 148 MB available
- Physical volume hdisk2 is unavailable; (in no volume group)
PAGING CONFIGURATION
- Physical Volume hdisk2 (type: SCSI) has no paging space defined
- Paging space paging00 on volume group rootvg is fragmented
- Paging space paging01 on volume group uservg is fragmented
I/O BALANCE
- Phys. volume hdisk2 is not busy
volume hdisk2, mean util. = 0.00 %
PROCESSES
- First appearance of 20642 (cpubound) on top-3 cpu list
(cpu % = 24.10)
- First appearance of 20106 (eatmem) on top-3 memory list
(memory % = 8.00)
FILE SYSTEMS
- File system hd2 (/usr) is nearly full at 100 %
NETWORK
- Host ah6000e appears to be unreachable.
(ping loss % = 100) and has been for the past 4 days
The I/O configuration indicates that the data is not well-distributed through the disks and that hdisk2 is
not used at all. Add this disk to a volume group and define a paging space on it. The existing paging areas
are fragmented and should be reorganized; for example, with the reorgvg command. The I/O balance
section shows that hdisk2 is not busy, because hdisk2 has not been assigned to a volume group.
In most systems, /usr file system use is nearly 100 percent. Usually this is not a problem, but system
administrators should check if there is any application writing data to this file system.
Upward and Downward Trends
PDT employs a statistical technique to determine whether there is a trend in a series of measurements. If
a trend is detected, the slope of the trend is evaluated for its practical significance. An estimated date at
which a file system or the page space will be full is provided, based on an assumption of continued linear
growth. For upward trends, the following items are evaluated: files, file systems, hardware and software
errors, paging space, processes, and network. For downward trends, the following can be reported: files,
file systems, and processes.
An example of an Upward Trends and Downward Trends section follows:
---------------------- Upward Trends ---------------FILES
- File (or directory) /usr/adm/wtmp SIZE is increasing
now, 20 KB and increasing an avg. of 2163 bytes/day
328
Performance Management Guide
-
File (or directory) /var/adm/ras/ SIZE is increasing
now, 677 KB and increasing an avg. of 11909 bytes/day
FILE SYSTEMS
- File system hd9var (/var) is growing
now, 17.00 % full, and growing an avg. of 0.38 %/day
- File system lv00 (/usr/vice/cache) is growing
now, 51.00 % full, and growing an avg. of 4.64 %/day
At this rate, lv00 will be full in about 9 days
PAGE SPACE
- Page space hd6 USE is growing
now, 81.60 MB and growing an avg. of 2.69 MB/day
At this rate, hd6 will be full in about 29 days
ERRORS
- Software ERRORS; time to next error is 0.958 days
---------------------- Downward Trends -------------PROCESSES
- Process 13906 (maker4X.e) CPU use is declining
now 1.20 % and declining an avg. of 0.68 % per day
- Process 13906 (maker4X.e) MEMORY use is declining
now 13.00 and declining an avg. of 0.98 % per day
FILES
- File (or directory) /tmp/ SIZE is declining
FILE SYSTEMS
- File system hd3 (/tmp) is shrinking
The /usr/adm/wtmp is liable to grow unbounded. If it gets too large, login times can increase. In some
cases, the solution is to delete the file. In most cases, it is important to identify the user causing the
growth and work with that user to correct the problem.
The error log file is located in the directory /var/adm/ras. The ERRORS section shows that the number of
software errors is increasing, which is most likely the reason why the directory size increased. Check the
error log, verify which application is in error, and correct the problem.
The increase in the use of paging space might be due to a process with a memory leak. That process
should be identified and the application fixed. However, the paging space might not be well-dimensioned
and may need to be enlarged.
System Health
The System Health section provides an assessment of the average number of processes in each process
state on the system. Additionally, workload indicators are noted for any upward trends.
An example of a System Health section follows:
----------------------- System Health --------------SYSTEM HEALTH
- Current process state breakdown:
75.00 [ 100.0 %] : active
0.40 [ 0.5 %] : swapped
75.00 = TOTAL
[based on 1 measurement consisting of 10 2-second samples]
Summary
The severity level of the current report is listed, as well as an indication as to whether more details are
available at higher severity levels.
An example of a Summary section follows:
-------------------- Summary ------------------------This is a severity level 3 report
No further details available at severity levels > 3
Chapter 14. Using Performance Diagnostic Tool (PDT)
329
Any message (excluding header and summary information) occurring in the PDT report should be
investigated. The indicated problem should be corrected or an explanation for the condition obtained.
Possible responses to specific messages are covered in Responding to PDT Report Messages.
Installing and Enabling PDT
PDT is installed through the installp command as the bos.perf.diag_tool option of the operating system
version 4 BOS licensed program.
PDT must be enabled in order to begin collecting data and writing reports. Enable PDT by executing the
/usr/sbin/perf/diag_tool/pdt_config script. Only the root user is permitted to run this script. When
executed, the following message is displayed:
# /usr/sbin/perf/diag_tool/pdt_config
________________PDT customization menu__________________
1) show current PDT report recipient and severity level
2) modify/enable PDT reporting
3) disable
PDT reporting
4) modify/enable PDT collection
5) disable
PDT collection
6) de-install
PDT
7) exit pdt_config
Please enter a number:
When you respond with 4, default PDT collection and reporting is enabled. The crontab entry for user adm
is updated to add the PDT entries (see Modifying the Collection, Retention, and Reporting Times for more
detail). The entries execute a shell script called Driver_ in the /usr/sbin/perf/diag_tool directory. This
script is passed three different parameters, each representing a collection profile, at three different
collection times.
To terminate the pdt_config program, respond with 7.
To disable collection, respond with 5.
Customizing PDT
Certain aspects of PDT can be customized. For example, any user can be designated as the regular
recipient of PDT reports, and the retention period for data in PDT’s historical record can be modified. All
customization is performed either by modifying one of the PDT files in the directory
/var/perf/cfg/diag_tool/ or by executing the /usr/sbin/perf/diag_tool/pdt_config script.
It is recommended that no changes be made until after PDT has produced several reports, and a certain
familiarity with PDT has been acquired.
Changing the PDT Report Recipient and Severity Level
By default, PDT reports are generated with severity level 1 with only the most serious problems identified.
There are other severity levels (2 and 3) at which more detailed information is frequently available. Further,
whenever a PDT report is produced, it is mailed to the adm user. You can choose to have the report
mailed elsewhere or not mailed at all.
Both of these parameters are controlled with the /usr/sbin/perf/diag_tool/pdt_config script. The following
dialog changes the user and the severity level:
# /usr/sbin/perf/diag_tool/pdt_config
________________PDT customization menu__________________
1) show current
330
PDT report recipient and severity level
Performance Management Guide
2) modify/enable PDT reporting
3) disable
PDT reporting
4) modify/enable PDT collection
5) disable
PDT collection
6) de-install
PDT
7) exit pdt_config
Please enter a number: 1
current PDT report recipient and severity level
adm 1
________________PDT customization menu__________________
1) show current PDT report recipient and severity level
2) modify/enable PDT reporting
3) disable
PDT reporting
4) modify/enable PDT collection
5) disable
PDT collection
6) de-install
PDT
7) exit pdt_config
Please enter a number: 2
enter [email protected] for recipient of report : rsmith
enter severity level for report (1-3): 2
report recipient and severity level
rsmith 2
________________PDT customization menu__________________
1) show current PDT report recipient and severity level
2) modify/enable PDT reporting
3) disable
PDT reporting
4) modify/enable PDT collection
5) disable
PDT collection
6) de-install
PDT
7) exit pdt_config
Please enter a number: 1
current PDT report recipient and severity level
rsmith 2
________________PDT customization menu__________________
1) show current PDT report recipient and severity level
2) modify/enable PDT reporting
3) disable
PDT reporting
4) modify/enable PDT collection
5) disable
PDT collection
6) de-install
PDT
7) exit pdt_config
Please enter a number: 7
In the preceding example, the recipient is changed to user rsmith, and the severity is changed to 2. This
means that user rsmith will receive the PDT report, and that both severity 1 and 2 messages will be
included. Note the use of option 1 to determine the current PDT report recipient and report severity level.
The user and security level could also be changed directly in the /var/perf/cfg/diag_tool/.reporting.list
file.
To terminate reporting (but allow collection to continue), option 3 is selected, for example:
# /usr/sbin/perf/diag_tool
________________PDT customization menu__________________
Chapter 14. Using Performance Diagnostic Tool (PDT)
331
1) show current PDT report recipient and severity level
2) modify/enable PDT reporting
3) disable
PDT reporting
4) modify/enable PDT collection
5) disable
PDT collection
6) de-install
PDT
7) exit pdt_config
Please enter a number: 3
disable PDT reporting done
________________PDT customization menu__________________
1) show current PDT report recipient and severity level
2) modify/enable PDT reporting
3) disable
PDT reporting
4) modify/enable PDT collection
5) disable
PDT collection
6) de-install
PDT
7) exit pdt_config
Please enter a number: 1
reporting has been disabled (file .reporting.list not found).
PDT Severity Levels
The following lists indicate the possible problems associated with each severity level. Remember that
selecting severity n results in the reporting of all problems of severity less than or equal to n.
Severity 1 Problems
v Journaled File System (JFS) becomes unavailable
v JFS nearly full
v Physical volume not allocated to a volume group
v All paging spaces defined on one physical volume
v System appears to have too little memory for current workload
v Page space nearly full
v Possible problems in the settings of load control parameters
v VMM-detected bad memory frames
v Any host in .nodes becomes unreachable
Severity 2 Problems
v Imbalance in the I/O configuration (for example, disks per adapter)
v Imbalance in allocation of paging space on physical volumes with paging space
v Fragmentation of a paging space in a volume group
v
v
v
v
v
v
v
Significant imbalance in measured I/O load to physical volumes
New process is identified as a heavy memory or CPU consumer
A file in .files exhibits systematic growth (or decline) in size
A file system or page space exhibits systematic growth (or decline) in space utilization
A host in .nodes exhibits degradation in ping delays or packet loss percentage
A getty process consumes too much CPU time
A process with high CPU or memory consumption exhibits systematic growth (or decline) in resource
use
v maxuproc indicated as being possibly too low for a particular user ID
v A WORKLOAD TRACKING indicator shows an upward trend
332
Performance Management Guide
Severity 3 Messages
Severity 3 messages provide additional detail about problems identified at severity levels 1 and 2. This
includes the data-collection characteristics, such as number of samples, for severity 1 and 2 messages.
Obtaining a PDT Report on Demand
As an alternative to using the periodic report, any user can request a current report from the existing data
by executing /usr/sbin/perf/diag_tool/pdt_report SeverityNum. The report is produced with the given
severity (if none is provided, SeverityNum defaults to 1) and written to standard output. Generating a
report in this way does not cause any change to the /var/perf/tmp/PDT_REPORT or to
/var/perf/tmp/PDT_REPORT.last files.
Modifying the List of Files Monitored by PDT
PDT analyzes files and directories for systematic growth in size. It examines only those files and
directories listed in the file /var/perf/cfg/diag_tool/.files. The format of the .files file is one file or directory
name per line. The default content is as follows:
/usr/adm/wtmp
/var/spool/qdaemon/
/var/adm/ras/
/tmp/
You can use an editor to modify this file to track files and directories that are important to your system.
Modifying the List of Hosts That PDT Monitors
PDT tracks the average ping delay to hosts whose names are listed in the /var/perf/cfg/diag_tool/.nodes
file. This file is not shipped with PDT (which means that no host analysis is performed by default), but may
be created by the administrator. The format of the .nodes file is one host name per line in the file. For
example, to monitor nodes chuys and hulahut, the file .nodes would be as follows:
chuys
hulahut
Changing the Historical-Record Retention Period
Periodically, a retention shell script is run that discards entries in the PDT historical record that are older
than the designated retention period. The retention of all data is governed by the same retention policy.
This policy is described in the /var/perf/cfg/diag_tool/.retention.list file. The default .retention.list
content is as follows:
* * * 35
which causes all data to be retained no more than 35 days. The number 35 can be replaced by any
unsigned integer.
PDT uses the historical record to assess trends and identify system changes. Extending the retention
period increases the scope of this analysis, but at the cost of additional disk storage and PDT processing
time.
The PDT historical record is maintained in /var/perf/tmp/.SM. The retention script creates a copy of this
file in /var/perf/tmp/.SM.last prior to performing the retention operation. In addition, historical data that is
discarded is appended to /var/perf/tmp/.SM.discards.
The existence of /var/perf/tmp/.SM.last provides limited backup, but the administrator should ensure that
the /var/perf/tmp/.SM file is regularly backed up. If the file is lost, PDT continues to function, but without
the historical information. Over time, the historical record will grow again as new data is collected.
Chapter 14. Using Performance Diagnostic Tool (PDT)
333
Modifying the Collection, Retention, and Reporting Times
Collection, reporting and retention are driven by three entries in the user adm cron table. Collection occurs
on every weekday at 9 a.m (Driver_ daily). Reporting occurs every weekday at 10 a.m (Driver_ daily2).
The retention analysis is performed once a week, on Saturday evening at 9 p.m (Driver_ offweekly). The
following files are used:
/var/perf/cfg/diag_tool/.collection.control
Handles collection information
/var/perf/cfg/diag_tool/.retention.control
Handles retention information
/var/perf/cfg/diag_tool/.reporting.control
Handles reporting information
The cron entries (created by executing the /usr/sbin/perf/diag_tool/pdt_config script and selecting option
2) are shown below:
0 9 * * 1-5
0 10 * * 1-5
0 21 * * 6
/usr/sbin/perf/diag_tool/Driver_ daily
/usr/sbin/perf/diag_tool/Driver_ daily2
/usr/sbin/perf/diag_tool/Driver_ offweekly
The default times can be changed by altering the crontab for user adm.
Modifying the Thresholds
The file /var/perf/cfg/diag_tool/.thresholds contains the thresholds used in analysis and reporting. These
thresholds, listed below, have an effect on PDT report organization and content.
v DISK_STORAGE_BALANCE
The SCSI controllers having the largest and the smallest disk storage are identified. This is a static size,
not the amount allocated or free. If the difference (in MB) between these two controllers exceeds
DISK_STORAGE_BALANCE, a message is reported:
SCSI Controller %s has %.0lf MB more storage than %s
The default value for DISK_STORAGE_BALANCE is 800. Any integer value between 0 and 10000 is
valid.
v PAGING_SPACE_BALANCE
The paging spaces having the largest and the smallest areas are identified. If the difference (in MB)
between these two exceeds PAGING_SPACE_BALANCE, a message is reported. The default value is
4. Any integer value between 0 and 100 is accepted. This threshold is presently not used in analysis
and reporting.
v NUMBER_OF_BALANCE
The SCSI controllers having the largest and the least number of disks attached are identified. If the
difference between these two counts exceeds NUMBER_OF_BALANCE, a message is reported:
SCSI Controller %s has %.0lf more disks than %s
The default value is 1. It can be set to any integer value in the range of 0 to 10000.
The same type of test is performed on the number of paging areas on each physical volume:
Physical Volume %s has %.0lf paging areas, while Physical Volume %s has only %.0lf
v MIN_UTIL
Applies to process utilization. Changes in the top three CPU consumers are only reported if the new
process had a utilization in excess of MIN_UTIL.
First appearance of %s (%s) on top-3 cpu list
The same threshold applies to changes in the top-three memory consumers list:
334
Performance Management Guide
First appearance of %s (%s) on top-3 memory list
The default value is 3. Any integer value from 0 to 100 is valid.
v FS_UTIL_LIMIT
Applies to journaled file system utilization. If the file system has a percentage use above
FS_UTIL_LIMIT, a message is reported:
File system %s (%s) is nearly full at %.0lf %%
The same threshold is applied to paging spaces:
Paging space %s is nearly full at %.0lf %%
The default value is 90 percent. Any integer value between 0 and 100 is accepted.
Special attention should be given to /, /var, and /tmp file systems. The operating system uses these
areas for normal operation. If there remains no space in one of these, the behavior of the system is
unpredictable. Error messages are provided when the execution of commands fails, but to detect these
file system problems earlier, decrease FS_UTIL_LIMIT to 70 or 80 percent.
v MEMORY_FACTOR
The objective is to determine if the total amount of memory is adequately backed up by paging space. If
real memory is close to the amount of used paging space, then the system is likely paging and would
benefit from the addition of memory.
The formula is based on experience and actually compares MEMORY_FACTOR * memory with the
average used paging space.
The current default is 0.9. By decreasing this number, a warning is produced more frequently:
System has %.0lf MB memory; may be inadequate.
Increasing this number eliminates the message altogether. It can be set anywhere between 0.001 and
100.
v TREND_THRESHOLD
Used in all trending assessments. It is applied after a linear regression is performed on all available
historical data. This technique basically draws the best line among the points. The slope of the fitted line
must exceed the last_value * TREND_THRESHOLD.
File system %s (%s) is growing,
now, %.2lf %% full, and growing an avg. of %.2lf %%/day
The objective is to try to ensure that a trend, however strong its statistical significance, has some
practical significance.
For example, if we determine that a file system is growing at X MB a day, and the last_value for the file
system size is 100 MB, we require that X exceeds 100 MB * TREND_THRESHOLD to be reported as a
trend of practical significance. The default value is 0.01; so a growth rate of 1 MB per day would be
required for reporting. The threshold can be set anywhere between 0.00001 and 100000.
This threshold assessment applies to trends associated with:
– CPU use by a top-three process
– Memory use by a top-three process
–
–
–
–
–
–
Size of files indicated in the .files file
Journaled file systems
Paging spaces
Hardware and software errors
Workload indicators
Processes per user
Chapter 14. Using Performance Diagnostic Tool (PDT)
335
– Ping delay to nodes in the .nodes file
– Percentage of packet loss to nodes in the .nodes file
v EVENT_HORIZON
Used also in trending assessments. For example, in the case of file systems, if there is a significant
(both statistical and practical) trend, the time until the file system is 100 percent full is estimated. If this
time is within EVENT_HORIZON, a message is reported:
At this rate, %s will be full in about %.0lf days
The default value is 30, and it can be any integer value between 0 and 100000.
This threshold applies to trends associated with:
– File Systems (JFS)
– Page Spaces
PDT Error Reporting
Errors can occur within each of the different PDT components. In general, an error does not terminate
PDT. Instead, a message is output to the PDT standard error file, /var/perf/tmp/.stderr. That phase of
processing then terminates.
Users experiencing unexpected behavior, such as the PDT report not being produced as expected, should
examine the /var/perf/tmp/.stderr file.
Uninstalling PDT
It is not possible to uninstall PDT directly using the pdt_config command, but if option 6 is requested, a
message describes the steps necessary to remove PDT from the system:
# /usr/sbin/perf/diag_tool/pdt_config
________________PDT customization menu__________________
1) show current PDT report recipient and severity level
2) modify/enable PDT reporting
3) disable
PDT reporting
4) modify/enable PDT collection
5) disable
PDT collection
6) de-install
PDT
7) exit pdt_config
Please enter a number: 6
PDT is installed as package bos.perf.diag_tool in the bos lpp.
Use the installp facility to remove the package
Responding to PDT Report Messages
PDT identifies many types of problems. Responses to these indications depend on the individual
organization’s available resources and set of priorities. The following samples suggest some possibilities
(cmds stands for commands):
Problem:
JFS file system becomes unavailable
Response:
Investigate why file system is unavailable. The file system could have been removed.
Useful cmds:
lsfs (to determine file system status)
Problem:
JFS file system nearly full
336
Performance Management Guide
Response:
This problem could be caused by large or core files within the file system. Look for large files in
the file system, possibly caused by a runaway process. Attempt to identify the process or user that
generated those files. The system administrator should also verify if PDT report indicates a
long-term growth trend for this file system. Has this file system exhibited long-term growth trend
(examine the rest of the PDT report or past PDT reports)?
Useful cmds:
du, ls
Problem:
Physical volume not allocated to a volume group
Response:
If a physical volume is not allocated to a volume group, the operating system has no access to
this disk, and its space is being wasted. Use the lspv command to ensure that the disk is not
allocated to any volume group, and if not, use the extendvg command to add the disk to a volume
group.
Useful cmds:
lspv (to confirm that the volume is not allocated)
smitty (to manipulate volume groups)
Problem:
All paging spaces defined on one physical volume
Response:
The system has more than one physical volume, yet all paging space is defined on a single
volume. If the system experiences paging, this configuration will result in reduced performance. A
better I/O throughput could be achieved if the paging space is split equally among all physical
volumes. Only one paging space should be defined per physical volume because the system will
only have access to one at a time. Use SMIT to create, modify, activate, or deactivate the paging
areas.
Useful cmds:
smitty (to modify paging spaces)
Problem:
Apparently too little memory for current workload
Response:
If the system is paging heavily, more memory may be required on the system for good
performance. The vmstat or svmon commands provide further details about the paging activity.
See The vmstat Command, and The svmon Command, for more information on those commands.
Useful cmds:
lsps -a, vmstat, svmon
Problem:
Page space nearly full
Response:
The system’s paging space may not be well dimensioned and may need to be enlarged, unless
the problem is due to a process with a memory leak, in which case that process should be
identified and the application fixed. For systems up to 256 MB of memory, the paging space
should be twice the size of real memory. For memories larger than 256 MB, use the following
formula:
total paging space = 512 MB + (memory size - 256 MB) * 1.25
Useful cmds:
ps aucg (to examine process activity)
smitty (to modify page space characteristics)
Chapter 14. Using Performance Diagnostic Tool (PDT)
337
Problem:
Possible problems in the settings of load control parameters
Response:
The memory-load-control parameters are evaluated in relation to current paging activity. For
example, if thrashing is occurring and load control is not enabled, it may be appropriate to enable
load control. This situation might also be due to inappropriate load-control parameter settings. Use
the schedtune command to view or alter the configuration. Refer to Tuning VMM Memory Load
Control with the schedtune Command, for further details.
Useful cmds
schedtune
Problem:
VMM-detected bad memory frames
Response:
It might be necessary to have the memory analyzed. Compare the amount of installed memory
with the memory actually accessible. If the latter is less than the former, then bad memory has
been identified.
You can use /usr/sbin/perf/diag_tool/getvmparms to examine the value of numframes to
determine the actual number of valid 4 KB memory frames.
Useful cmds:
lscfg | grep mem (to obtain installed memory size in MB)
Problem:
Any host in .nodes becomes unreachable
Response:
Determine if problem is with current host (has a change in the /etc/hosts file been made?), with
the remote host (is it down?), or with the network (is the nameserver down?).
The problem may be due to name resolution. Either Domain Name Service (DNS) configuration
files or /etc/hosts should be checked depending on the type of name resolution being used at the
environment.
Use the ping command to check if the machine has access to other nodes in the same network.
The remote node might be down. If it cannot access any other node, cables and connections
should be verified, as well as the routing table of the current machine. Verify cables and
connections by executing the netstat -r command.
Useful cmds:
ping, netstat
Problem:
Imbalance in the I/O configuration (number of disks per adapter)
Response:
The number of disks per adapter should be equal whenever possible, to prevent one adapter from
being overloaded. A guideline is to have no more than four devices per adapter, especially if the
access to the disks is mostly sequential. Consider moving disks around so that an individual
adapter is not overloaded.
Useful cmds:
lscfg (to examine the current configuration)
iostat (to determine if the actual load on the adapters is out of balance)
Problem:
Imbalance in allocation of paging space on physical volumes with paging space
Response:
A substantial imbalance in the sizes of paging spaces can cause performance problems. The
338
Performance Management Guide
paging space should be equally distributed throughout the disks. Consider making paging spaces
the same size, except for a few extra megabytes on the primary paging space (hd6).
Useful cmds:
smitty pgsp
Problem:
Fragmentation of a paging space in a volume group
Response:
Paging performance is better if paging areas are contiguous on a physical volume. However, when
paging areas are enlarged, it is possible to create fragments that are scattered across the disk
surface. Use the reorgvg command to reorganize the paging spaces.
Useful cmds:
lspv -p hdiskn for each physical volume in the volume group. Look for more than one PP Range
with the same LVNAME and a TYPE of paging.
Problem:
Significant imbalance in measured I/O load to physical volumes
Response:
The data is most likely not well-distributed throughout the disks. Use the iostat command to obtain
information about the I/O activity of each disk (refer to Assessing Disk Performance with the iostat
Command). A disk should not be utilized more than 40 percent over a period of time.
If one physical volume seems to be getting little I/O activity, consider moving data from busier
physical volumes onto less busy volumes. In general, the more evenly the I/O is distributed, the
better the performance.
Distribute data throughout the disks in a manner that balances I/O. Use the filemon command to
obtain information about the most accessed files and file systems. This can be a good starting
point in reorganizing the data. Refer to Detailed I/O Analysis with the filemon Command, for more
information on the filemon command.
Useful cmds:
iostat -d 2 20 (to view the current distribution of I/O across physical volumes)
Problem:
New process is a heavy consumer of memory or CPU
Response:
Top CPU and memory consumers are regularly identified by PDT. If any of these processes have
not been detected before, they are highlighted in a problem report. Examine these processes for
unusual behavior. Note that PDT simply looks at the process ID. If a known heavy user terminates,
then is resumed (with a different process ID), it will be identified here as a new heavy user.
Useful cmds:
ps aucg (to view all processes and their activity)
Problem:
Any file in .files exhibits systematic growth (or decline) in size
Response:
Look at the current size. Consider the projected growth rate. What user or application is
generating the data? For example, the /var/adm/wtmp file is liable to grow unbounded. If it gets
too large, login times can increase. In some cases, the solution is to delete the file. In most cases,
it is important to identify the user causing the growth and work with that user to correct the
problem.
Useful cmds:
ls -al (to view file/directory sizes)
Problem:
Any file system or paging space exhibits systematic growth (or decline) in space used
Chapter 14. Using Performance Diagnostic Tool (PDT)
339
Response:
Consider the projected growth rate and expected time until exceeding the available space. Analyze
the problem by identifying which user or process is generating the data. It may be necessary to
enlarge the file system (or page space). On the other hand, the growth may be an undesirable
effect (for example, a process having a memory leak).
Useful cmds:
smitty (to manipulate file systems/page spaces)
ps aucg, svmon (to view process virtual memory activity)
filemon (to view file system activity)
Problem:
Degradation in ping response time or packet loss percentage for any host in .nodes
Response:
There is probably a performance problem in the host or in the network. Is the host in question
experiencing performance problems? Is the network having performance problems?
Useful cmds:
ping, rlogin, rsh (to time known workloads on remote host)
Problem:
A getty process that consumes too much CPU time
Response:
Getty processes that use more than just a few percent of the CPU may be in error. It is possible in
certain situations for these processes to consume system CPU, even though no users are actually
logged in. In general, the solution is to terminate the process.
Useful cmds:
ps aucg (to see how much CPU is being used)
Problem:
A process that is a top consumer of CPU or memory resources exhibits systematic growth
or decline in consumption
Response:
Known large consumers of CPU and memory resources are tracked over time to see if their
demands grow. As major consumers, a steady growth in their demand is of interest from several
perspectives. If the growth is normal, this represents useful capacity planning information. If the
growth is unexpected, then evaluate the workload for a change (or a chronic problem, such as a
memory leak). Use the vmstat and svmon commands while the process is running to gather more
information on its behavior.
Useful cmds:
ps aucg, vmstat, svmon
Problem:
maxuproc indicated as being possibly too low for a particular userid
Response:
It is likely that this user is reaching the maxuproc threshold.
maxuproc is a systemwide parameter that limits the number of processes that nonroot users are
allowed to have simultaneously active. If the limit is too low, the user’s work can be delayed or
terminated. On the other hand, the user might be accidentally creating more processes than
needed or appropriate. Further investigation is warranted in either case. Consult the user in order
to clearly understand what is happening.
Useful cmds:
lsattr -E -l sys0 | grep maxuproc to determine the current value of maxuproc (although it is also
340
Performance Management Guide
reported directly in the PDT message).
chdev -l sys0 -a maxuproc=100 to change maxuproc to 100 (for example). Root user authority
is required.
Problem:
A WORKLOAD TRACKING indicator shows an upward trend
Response:
The response depends on which workload indicator shows the trend:
loadavg
Refers to 15-minute load average. In general, it indicates that the level of contention in the
system is growing. Examine the rest of the PDT report for indicators of system bottlenecks
(for example, substantial page space use might indicate a memory shortage; I/O
imbalances might indicate that the I/O subsystem requires attention).
nusers
Shows that the number of logged-on users on the system is growing. This is important
from a capacity planning perspective. Is the growth expected? Can it be explained?
nprocesses
Indicates that the total number of processes on the system is growing. Are there users
reaching the maxuproc limitation? Perhaps there are ″runaway″ applications forking too
many processes.
STAT_A
Number of active processes. A trend here indicates processes are spending more time
waiting for the CPU.
STAT_W
Number of swapped processes. A trend here indicates that processes are contending
excessively for memory.
STAT_Z
Number of zombie processes. Zombies should not stay around for a long period of time. If
the number of zombies on a system is growing, this may be cause for concern.
STAT_I
Number of idle processes.
STAT_T
Number of processes stopped after receiving a signal. A trend here might indicate a
programming error.
STAT_x
Number of processes reported by the ps command as being in state x, where x is a state
not listed in the other STAT_* states. The interpretation of a trend depends on the
meaning of the character x. Refer to Using the ps Command, for more information on the
ps command.
cp
Time required to copy a 40 KB file. An upward trend in the time to do a file copy suggests
degradation in the I/O subsystem.
idle_pct_cpu0
Idle percentage for processor 0. An upward trend in the idle percentage might indicate
increased contention in non-CPU resources such as paging or I/O. Such an increase
suggests the CPU resource is not being well-utilized.
idle_pct_avg
Average idle percentage for all processors. An upward trend in the idle percentage might
indicate increased contention in non-CPU resources such as paging or I/O. Such an
increase suggests the CPU resource is not being well-utilized.
Chapter 14. Using Performance Diagnostic Tool (PDT)
341
342
Performance Management Guide
Chapter 15. Reporting Performance Problems
If you believe that you have found a possible performance problem in the operating system, you can use
tools and procedures for reporting the problem and supplying problem-analysis data. These tools are
intended to ensure that you get a prompt and accurate response, with a minimum of effort and time on
your part.
The main sections in this topic are:
v
v
v
v
Measuring the Baseline
What is a Performance Problem
Performance Problem Description
Reporting a Performance Problem
Measuring the Baseline
Performance problems are often reported immediately following some change to system hardware or
software. Unless there is a pre-change baseline measurement with which to compare post-change
performance, quantification of the problem is impossible.
Changes to any of the following can affect performance:
v Hardware configuration - Adding, removing, or changing configurations such as how the disks are
connected
v Operating system - Installing or updating a fileset, installing PTFs, and changing parameters
v Applications - Installing new versions and fixes
v Applications - Configuring or changing data placement
v Application tuning
v Tuning options in the operating system, RDBMS or an application
v Any changes
The best option is to measure the environment before and after each change. The alternative is running
the measurements at regular intervals (for example, once a month) and save the output. When a problem
is found, the previous capture can be used for comparison. It is worth collecting a series of outputs in
order to support the diagnosis of a possible performance problem.
To maximize performance diagnosis, collect data for various periods of the working day, week, or month
when performance is likely to be an issue. For example, you might have workload peaks as follows:
v In the middle of the mornings for online users
v During a late-night batch run
v During the end-of-month processing
v During major data loads
Use measurements to collect data for each of these peaks in workload, because a performance problem
might only cause problems during one of these periods and not during other times.
Note: Any measurement has an impact on the performance of the system being measured.
The AIX Performance PMR (perfpmr) data collection tools are the preferred method for gathering baseline
data. Access these tools via the web at ftp://ftp.software.ibm.com/aix/tools/perftools/perfpmr. Follow the
instructions in the README file in the directory that matches the AIX version you will be measuring to
obtain, install, and collect data on your system.
© Copyright IBM Corp. 1997, 2002
343
What is a Performance Problem
Support personnel need to determine when a reported problem is a functional problem or a performance
problem. When an application, a hardware system, or a network is not behaving correctly, this is referred
to as a functional problem. For example, an application or a system with a memory leak has a functional
problem.
Sometimes functional problems lead to performance problems; for example, when the functions are being
achieved, but the speed of the functions are slow. In these cases, rather than tune the system, it is more
important to determine the root cause of the problem and fix it. Another example would be when
communication is slowed because of networks or name servers that are down.
Performance Problem Description
Support personnel often receive problem reports stating that someone has a performance problem on the
system and providing some data analysis. This information is insufficient to accurately determine the
nature of a performance problem. The data might indicate 100 percent CPU utilization and a high run
queue, but that may have nothing to do with the cause of the performance problem.
For example, a system might have users logged in from remote terminals over a network that goes over
several routers. The users report that the system is slow. Data might indicate that the CPU is very heavily
utilized. But the real problem could be that the characters get displayed after long delays on their terminals
due to packets getting lost on the network (which could be caused by failing routers or overloaded
networks). This situation might have nothing to do with the CPU utilization on the machine. If on the other
hand, the complaint was that a batch job on the system was taking a long time to run, then CPU utilization
or I/O bandwidth might be related.
Always obtain as much detail as possible before you attempt to collect or analyze data, by asking the
following questions regarding the performance problem:
v Can the problem be demonstrated by running a specific command or reconstructing a sequence of
events? (for example: ls /slow/fs or ping xxxxx). If not, describe the least complex example of the
problem.
v Is the slow performance intermittent? Does it get slow, but then disappear for a while? Does it occur at
certain times of the day or in relation to some specific activity?
v Is everything slow or only some things?
v What aspect is slow? For example, time to echo a character, or elapsed time to complete a transaction,
or time to paint the screen?
v When did the problem start occurring? Was the situation the same ever since the system was first
installed or went into production? Did anything change on the system before the problem occurred
(such as adding more users or migrating additional data to the system)?
v If client/server, can the problem be demonstrated when run just locally on the server (network versus
server issue)?
v If network related, how are the network segments configured (including bandwidth such as 10 Mb/sec or
9600 baud)? Are there any routers between the client and server?
v What vendor applications are running on the system, and are those applications involved in the
performance issue?
v What is the impact of the performance problem on the users?
Reporting a Performance Problem
You should report operating system performance problems to IBM support. Use your normal software
problem-reporting channel. If you are not familiar with the correct problem-reporting channel for your
organization, check with your IBM representative.
344
Performance Management Guide
The AIX Performance PMR (perfpmr) data collection tools are the best way to collect performance data
when an AIX performance problem is suspected. Access these tools via the web at
ftp://ftp.software.ibm.com/aix/tools/perftools/perfpmr Follow the instructions in the README file in the
directory that matches the AIX version you will be measuring to obtain, install, and collect data on your
system. Instructions are also provided on how to send the data to IBM support for analysis once a PMR
has been opened.
When someone reports a performance problem, it is not enough just to gather data and then analyze it.
Without knowing the nature of the performance problem, you might waste a lot of time analyzing data
which may have nothing to do with the problem being reported.
Before you involve support personnel to report a problem, prepare in advance the information that you will
be asked to supply to facilitate the problem to be investigated. Your local support personnel will attempt to
quickly solve your performance problem directly with you.
Three further ways you can help to get the problem resolved faster are:
1. Provide a clear written statement of a simple specific instance of problem, but be sure to separate the
symptoms and facts from the theories, ideas and your own conclusions. PMRs that report ″the system
is slow″ require extensive investigation to determine what you mean by slow, how it is measured, and
what is acceptable performance.
2. Provide information about everything that has changed on the system in the weeks before the problem.
Missing something that changed can block a possible investigation path and will only delay finding a
resolution. If all the facts are available, the performance team can quickly eliminate the unrelated ones.
3. Use the correct machine to supply information. In very large sites it is easy to accidentally collect the
data on the wrong machine. This makes it very hard to investigate the problem.
When you report the problem, supply the following basic information:
v A problem description that can be used to search the problem-history database to see if a similar
problem has already been reported.
v What aspect of your analysis led you to conclude that the problem is due to a defect in the operating
system?
v What is the hardware and software configuration in which the problem is occurring?
– Is the problem confined to a single system, or does it affect multiple systems?
– What are the models, memory sizes, as well as number and size of disks on the affected systems?
– What kinds of LAN and other communications media are connected to the systems?
– Does the overall configuration include those for other operating systems?
v What are the characteristics of the program or workload that is experiencing the problem?
– Does an analysis with the time, iostat, and vmstat commands indicate that it is CPU-limited or
I/O-limited?
– Are the workloads being run on the affected systems: workstation, server, multiuser, or a
combination?
v What are the performance objectives that are not being met?
– Is the primary objective in terms of console or terminal response time, throughput, or real-time
responsiveness?
– Were the objectives derived from measurements on another system? If so, what was its
configuration?
If this is the first report of the problem, you will receive a PMR number for use in identifying any additional
data you supply and for future reference.
Include all of the following items when the supporting information and the perfpmr data for the PMR is first
gathered:
Chapter 15. Reporting Performance Problems
345
v A means of reproducing the problem
– If possible, a program or shell script that demonstrates the problem should be included.
– At a minimum, a detailed description of the conditions under which the problem occurs is needed.
v The application experiencing the problem:
– If the application is, or depends on, any software product, the exact version and release of that
product should be identified.
– If the source code of a user-written application cannot be released, the exact set of compiler
parameters used to create the executable program should be documented.
346
Performance Management Guide
Chapter 16. Application Tuning
Before spending a lot of effort to improve the performance of a program, use the techniques in this chapter
to help determine how much its performance can be improved and to find the areas of the program where
optimization and tuning will have the most benefit. For an extensive discussion of these techniques, see
Optimization and Tuning Guide for XL Fortran, XL C and XL C++. Also see Designing and Implementing
Efficient Programs for additional hints and tips.
In general, the optimization process involves several steps:
v Some tuning involves changing the source code, for example, by reordering statements and
expressions. This technique is known as hand tuning.
v For FORTRAN and C programs, optimizing preprocessors are available to tune and otherwise transform
source code before it is compiled. The output of these preprocessors is FORTRAN or C source code
that has been optimized.
v The FORTRAN or C++ compiler translates the source code into an intermediate language.
v A code generator translates the intermediate code into machine language. The code generator can
optimize the final executable code to speed it up, depending on the selected compiler options. You can
increase the amount of optimization performed in this step by hand-tuning or preprocessing first.
The speed increase is affected by two factors:
v The amount of optimization applied to individual parts of the program
v The frequency of use for those parts of the program at run time
Speeding up a single routine might speed up the program significantly if that routine performs the majority
of the work, on the other hand, it might not improve overall performance much if the routine is rarely called
and does not take long anyway. Keep this point in mind when evaluating the performance techniques and
data, so that you focus on the techniques that are most valuable in your work.
This chapter contains the following major sections:
v Profiling
v Compiler Optimization Techniques
v Optimizing Preprocessors for FORTRAN and C
v Code-Optimization Techniques
Profiling
You can use profiling tools to identify which portions of the program are executed most frequently or where
most of the time is spent. Profilers are typically used after a basic tool, such as the vmstat or iostat
commands, shows that a CPU bottleneck is causing the slow performance.
Before you begin locating hot spots in your program, you need a fully functional program and realistic data
values to feed it to.
Timing Commands
Use the timing commands discussed in Using the time Command to Measure CPU Use for testing and
debugging programs whose performance you are recoding and trying to improve. The output from the time
command is in minutes and seconds, as follows:
real
user
sys
0m26.72s
0m26.53s
0m0.03s
The output from the timex command is in seconds:
© Copyright IBM Corp. 1997, 2002
347
real 26.70
user 26.55
sys 0.02
Comparing the user+sys CPU time to the real time will give you an idea if your application is CPU-bound
or I/O-bound.
Note: Be careful when you do this on an SMP system (see time and timex Cautions).
The timex command is also available through the SMIT command on the Analysis Tools menu, found
under Performance and Resource Scheduling. The -p and -s options of the timex command allow data
from accounting (-p)