U SER’S M ANUAL
DO-0975-005
SmartRAID V User’s Manual
Copyright © 1998, 1999 Distributed Processing Technology Corporation
All Rights Reserved – Printed in U.S.A.
Storage Manager is a trademark of Distributed Processing Technology Corporation.
SmartCache, SmartRAID, RAIDstation and DPT are registered trademarks of Distributed Processing Technology Corporation.
Microsoft, MS-DOS, Windows NT, Windows, and the Windows logo are either registered trademarks or
trademarks of Microsoft Corporation in the United States and/or other countries.
NetWare is a trademark of Novell Corporation.
OpenServer, UnixWare and SCO are trademarks of The Santa Cruz Operation. YES, IT RUNS WITH
UNIXWARE mark – developer tested only. SCO makes no warranties with respect to this product.
I2O is a registered trademark of Intel Corporation.
All other trademarks are property of their respective owners.
ii
SmartRAID V User’s Manual
FCC Statement
Warning: Changes or modifications to this unit not expressly approved by the
party responsible for compliance could void the user’s authority to operate the
equipment.
This equipment has been tested and found to comply with the limits for a Class
B digital device, pursuant to Part 15 of the FCC Rules. These limits are designed
to provide reasonable protection against harmful interference in a residential
installation. This equipment generates, uses, and can radiate radio frequency
energy and, if not installed and used in accordance with the instructions, may
cause harmful interference to radio communications. However, there is no
guarantee that interference will not occur in a particular installation. If this
equipment does cause harmful interference to radio or television reception,
which can be determined by turning the equipment off and on, the user is
encouraged to try to correct the interference by one or more of the following
measures:
• Reorient or relocate the receiving antenna.
• Increase the separation between the equipment and receiver.
• Connect the equipment into an outlet on a circuit different from that to
which the receiver is connected.
• Consult the dealer or an experienced radio TV technician for help.
This device complies with Part 15 of the FCC Rules. Operation is subject to the
following two conditions: (1) this device may not cause harmful interference,
and (2) this device must accept any interference received, including interference
that may cause undesired operation.
NOTE: If the SX4055F Fibre Channel Expansion Module is used, the DPT
controller will not meet Class B limits. A DPT controller with an SX4055F module
does comply with the limits for a Class A digital device.
VCCI Statement
This is a Class B product based on the standard of the Voluntary Control
Council for Interference from Information Technology Equipment (VCCI).
If this is used near a radio or television receiver in a domestic
environment, it may cause radio interference. Install and use the
equipment according to the instruction manual.
iii
SmartRAID V User’s Manual
Distributed Processing Technology
Limited Product Warranty
PROOF OF PURCHASE MAY BE REQUIRED. YOU MAY REGISTER YOUR PURCHASE BY RETURNING THE
ENCLOSED WARRANTY CARD WITHIN 30 DAYS OF PURCHASE.
Distributed Processing Technology Corporation (“DPT”) warrants to the purchaser of this product that it will be free
from defects in material and workmanship for the period as set forth below:
PRODUCT TYPE
WARRANTY (Years from date of purchase)
Hard disks
5 years
RAID/SCSI Controllers
3 years
Cache, Memory, SCSI Expansion and RAID /Caching Modules
3 years
Storage Array Cabinets (Tower and Rackmount)
3 years
Battery Module
1 year
FIRMWARE IS NOT COVERED BY ANY WARRANTY. SOFTWARE IS PROVIDED “AS IS”. DPT DOES NOT WARRANT
THAT THE SOFTWARE WILL MEET YOUR REQUIREMENTS OR THAT ITS USE WILL BE ERROR-FREE OR
UNINTERRUPTED. DPT does warrant that the diskette(s) on which the program is furnished will be free from defects in
materials and workmanship under normal use for a period of 90 days from the date of purchase.
If Distributor or Distributor’s customer notifies DPT of a defect in the product within the applicable warranty period, DPT
shall, at its sole option, either (i) repair the defective product (ii) replace the defective product or (iii) refund the purchase
price paid by the Distributor. Replacement parts or products may be new or reconditioned. Product which fails to function
upon proper installation shall be deemed dead on arrival (“DOA”), subject to verification by DPT. DOA product shall be
replaced by DPT provided that DPT has been notified of the DOA within forty-five (45) days of the date of purchase by
Distributor or Distributor’s customer. Product returned for any warranty service will not be accepted by DPT unless (i) a
return material authorization (RMA) number has been received from DPT along with instructions for returning the product
prior to the return, (ii) returned product is in its original or equivalent packaging (to ensure product is not damaged in
shipment) and (iii) the product has been returned to DPT within 15 days of issuance of the RMA number. DPT will not be
responsible for damage to product during the return shipment to DPT. DPT reserves the right to (i) charge a $20 re-box fee
or (ii) reject any returned product not shipped in the original or equivalent packaging. Freight charges associated with
improperly packed product rejected by DPT will be the responsibility of the customer. At its option, DPT may require proof
of purchase. Distributor or Distributor’s customer shall be responsible for all costs of shipping products for warranty
service unless the product is DOA, in which event DPT shall be responsible for the cost of return shipment to the
customer. If DPT determines that the product is not defective within the terms of the warranty, customer shall pay the cost
of repair at DPT’s then prevailing rate. This limited warranty is contingent upon proper use of the product and does not
cover product which has been modified, or subjected to unusual physical or electrical stress, unauthorized service, or
failure to perform preventative maintenance. Replaced parts or products become the property of DPT.
DPT MAKES NO OTHER EXPRESS OR IMPLIED WARRANTY INCLUDING THE WARRANTIES OF MERCHANTABILITY
OR FITNESS FOR A PARTICULAR PURPOSE. IN NO EVENT WILL DPT BE LIABLE TO YOU FOR ANY DAMAGES,
INCLUDING LOST PROFITS, LOSS OF DATA OR OTHER INCIDENTAL OR CONSEQUENTIAL DAMAGES. DPT’s
LIABILITY FOR BREACH OF WARRANTY IS LIMITED EXCLUSIVELY TO REPAIR OR REPLACEMENT OF PRODUCTS.
Some states do not allow the exclusion of implied warranties or limitations of liability for incidental or consequential
damages, so the above exclusions may not apply to you. This warranty gives you specific legal rights. You may have other
rights, which vary from state to state.
Distributed Processing Technology
140 Candace Drive
Maitland, Florida • 32751 • USA
Phone (407) 830-5522 • Fax (407) 260-5366
World Wide Web: http://www.dpt.com
iv
SmartRAID V User’s Manual
Distributed Processing Technology
End-User Software License Agreement
You may (i) use the software on a single machine; (ii) make only one copy of the
software into any machine-readable or printed form for backup purposes; (iii)
merge the software into other software for your use on the single machine
(provided that any portion of the software merged into other software will
continue to be subject to the terms and conditions of this license); (iv) transfer
the software and this license to another party if the other party agrees to accept
the terms and conditions of this license; and (v) not make copies of any written
materials. If you transfer the software you must either transfer all copies of the
software and accompanying materials to the same party or destroy any copies
not transferred. You may not attempt to defeat any protection method
implemented by DPT to prevent unauthorized use of the software, and you may
not modify, adapt, translate, reverse engineer, decompile or disassemble either
software or written materials nor create derivative works based upon the
software or written materials. This license is effective until terminated by you
by destroying the software together with all copies. This license shall terminate
automatically if you fail to comply with any term or condition of this license.
This agreement is governed by the laws of the State of Florida.
U.S. Government Restricted Rights: The software and any manuals are
provided with restricted rights. If the software and manuals are acquired under
the terms of a DOD Contract, use, duplication or disclosure by the Government
is subject to restrictions as set forth in subparagraph (c)(I)(ii) of the Rights in
Technical Data and Computer Software Clause at 252.227-7013. If acquired under
the terms of a civilian agency contract, use, reproduction or disclosure is subject
to 52.227-19(a) through (d).
v
SmartRAID V User’s Manual
Release Notes
To review release notes and documentation errata associated with the
SmartRAID V products, select Help–View README Information in Storage
Manager or view the READ.ME file in the DPT install directory for the
latest information.
Check the DPT web site for availability of an updated version of this
manual (SmartRAID V User’s Manual) in Adobe Acrobat PDF format.
The current SmartRAID V Quick Install Guide is also available for
download from the DPT web site in PDF format. The Quick Install Guide
contains text in English, French, German, Italian, Polish and Spanish.
Technical Support
Our technical support staff is available Monday – Friday between 6 am and
6 pm Eastern time. Between 6 pm and 8 pm, you can leave a callback
message. You can contact us by fax or e-mail 24 hours a day.
Telephone:
Fax:
Internet:
407-830-5522 (Press 6 and follow the prompts.)
407-830-4793
http://www.dpt.com/techsup/supporthelp.html
DPT offers priority Technical Support as a fee-based option. If you choose
this option your call is given priority over all other support calls.
900-555-4378
at the rate of $1.35 per minute
407-830-5522
Press 6 and select the Priority Technical Support option.
The rate is $30 for the first hour and $1.00 per minute after
the first hour.
Product information and the latest versions of DPT drivers and utilities
can be obtained at no charge from the DPT FTP site (ftp.dpt.com) or from
the Technical Support section of our World Wide Web site 24 hours a day.
vi
SmartRAID V User’s Manual
Table of Contents
FCC Statement ................................................................................................................. iii
VCCI Statement ................................................................................................................ iii
Limited Product Warranty ................................................................................................. iv
End-User Software License Agreement ............................................................................. v
Release Notes .................................................................................................................. vi
Technical Support ............................................................................................................. vi
Read This First! ..................................................................................... 1-1
About the Documentation ..............................................................................................
User’s Manual ............................................................................................................
Storage Manager Help System ..................................................................................
Installation Roadmap .....................................................................................................
What’s In the Box? ........................................................................................................
System Requirements ...................................................................................................
Safety Information .........................................................................................................
Working with Electricity .................................................................................................
Preventing Electrostatic Discharge ................................................................................
1-2
1-2
1-3
1-3
1-5
1-5
1-6
1-7
1-8
About Your New Controller .................................................................. 2-1
Introduction ................................................................................................................... 2-2
Product List ................................................................................................................... 2-3
Product Feature Overview ............................................................................................. 2-4
SmartRAID V Millennium Features ............................................................................ 2-5
SmartRAID V Decade and Century Features ............................................................ 2-7
SX405x Bus Expansion Module Features .................................................................. 2-8
RA4050 RAID Accelerator Features .......................................................................... 2-9
BB4050 Battery Module Features ............................................................................ 2-10
SmartRAID V Controller LEDs .................................................................................... 2-11
LED Display During Power-Up ................................................................................. 2-11
LED Display During Controller Idle .......................................................................... 2-11
LED Display During Controller Active ....................................................................... 2-12
Cache Status LEDs .................................................................................................. 2-12
BB4050 LED Indicators ............................................................................................ 2-13
Audible Alarm .............................................................................................................. 2-13
Configuration and Installation............................................................. 3-1
Installation Overview .....................................................................................................
Configuration .................................................................................................................
Narrow and Wide SCSI ..............................................................................................
Configuring Cables .....................................................................................................
Configuring SCSI Termination ....................................................................................
3-2
3-3
3-3
3-3
3-4
vii
SmartRAID V User’s Manual
Configuring Device IDs .............................................................................................. 3-6
SCSI ....................................................................................................................... 3-6
Fibre Channel ......................................................................................................... 3-6
Installation ..................................................................................................................... 3-7
Installing the RA4050 RAID Accelerator .................................................................... 3-8
Installing an SX405x Bus Expansion Module ............................................................. 3-9
Installing SIMMs ....................................................................................................... 3-11
Installing DIMMs ...................................................................................................... 3-13
Installing the BB4050 Battery Backup Module ......................................................... 3-14
Installing the Controller ............................................................................................ 3-15
Determining the Booting Controller ....................................................................... 3-16
Controller IRQ and Address .................................................................................. 3-16
NVRAM Reset ...................................................................................................... 3-17
Using Microsoft Cluster Server ................................................................................... 3-18
System Requirements .............................................................................................. 3-19
How SmartRAID V Works with MSCS ...................................................................... 3-20
Storage Manager on ROM ................................................................... 4-1
Overview ....................................................................................................................... 4-2
Keyboard Reference ...................................................................................................... 4-3
Menu Reference ............................................................................................................ 4-4
Icon Reference .............................................................................................................. 4-4
Screen Layout ............................................................................................................... 4-5
The Menu Bar ............................................................................................................ 4-5
The Left Pane – Tree View ......................................................................................... 4-6
The Right Pane – Information View ............................................................................ 4-6
Running SMOR ............................................................................................................. 4-8
Information/Configuration Windows ............................................................................... 4-9
DPT I2O BIOS Settings .............................................................................................. 4-9
Controller Information Windows ............................................................................... 4-11
Controller Information Tab ..................................................................................... 4-11
HBA Configuration Tab .......................................................................................... 4-13
Device Information ................................................................................................... 4-15
Array and Array Group Information .......................................................................... 4-17
Setting the Configuration ............................................................................................. 4-19
Array Operations ......................................................................................................... 4-20
Creating an Array ..................................................................................................... 4-20
Creating a Parity Group ........................................................................................... 4-22
Deleting an Array ..................................................................................................... 4-23
Hot Spares ............................................................................................................... 4-23
Rebuilding a Failed Array ......................................................................................... 4-24
Formatting a Drive ....................................................................................................... 4-24
Managing Controller Firmware .................................................................................... 4-25
Upgrading Firmware .................................................................................................... 4-25
viii
SmartRAID V User’s Manual
Saving Controller Firmware to a Disk Image ............................................................... 4-27
Creating a Firmware Disk Image .............................................................................. 4-28
Restoring Firmware from a Disk Image ................................................................... 4-28
Software Installation ............................................................................ 5-1
Windows NT 4.0 ............................................................................................................ 5-2
Installing a New System ............................................................................................. 5-3
Upgrading an Existing System ................................................................................... 5-5
Adding to an Existing System .................................................................................... 5-5
Adding to an Existing System
(Microsoft OSM) ......................................................................................................... 5-6
Windows 95/98 .............................................................................................................. 5-7
Installing Windows 98 (Full Install Version) ................................................................ 5-7
Installing Windows 95
(OEM SR2 – Version 950b) ........................................................................................ 5-9
Installing Windows 95
(Retail Upgrade – Version 950a) .............................................................................. 5-11
Adding SmartRAID V to an Existing Windows 95/98 Configuration ......................... 5-11
Upgrading to Windows 98 from Windows 95
(Upgrade Version) .................................................................................................... 5-12
Installing Storage Manager for Windows ..................................................................... 5-13
SCO UNIX ................................................................................................................... 5-14
Installing SCO UNIX 3.2V4.2 or OpenServer 5 ........................................................ 5-14
Adding to an Existing SCO UNIX 3.2V4.2 or OpenServer 5 System ....................... 5-15
SCO UnixWare 7 ......................................................................................................... 5-15
Novell NetWare ........................................................................................................... 5-17
NetWare 4.11 – Boot Controller ............................................................................... 5-18
NetWare 4.11 – Secondary Controller ..................................................................... 5-19
NetWare 4.2 – Boot Controller ................................................................................. 5-20
NetWare 4.2 – Secondary Controller ....................................................................... 5-21
NetWare 5.0 – Boot Controller ................................................................................. 5-22
NetWare 5.0 – Secondary Controller ....................................................................... 5-23
Installing Remote Communication Support .............................................................. 5-24
Red Hat Linux .............................................................................................................. 5-24
Red Hat Linux 5.2 and 6.0 ....................................................................................... 5-25
Red Hat Linux 6.1 .................................................................................................... 5-25
Installing Storage Manager for Linux ........................................................................ 5-26
Storage Manager .................................................................................. 6-1
Introduction ...................................................................................................................
System Requirements ...................................................................................................
Running Storage Manager ............................................................................................
Using Storage Manager Locally .................................................................................
6-2
6-2
6-3
6-3
ix
SmartRAID V User’s Manual
Using Storage Manager Remotely ............................................................................. 6-3
Storage Manager on ROM ......................................................................................... 6-4
Physical Configuration View .......................................................................................... 6-4
Logical Configuration View ............................................................................................ 6-6
Logical Device Address ................................................................................................. 6-7
Status ............................................................................................................................ 6-8
Information Windows ................................................................................................... 6-10
Host Bus Adapter Information Window .................................................................... 6-10
Battery Backup Configuration ............................................................................... 6-11
HBA Configuration ................................................................................................ 6-13
Flash Configuration ............................................................................................... 6-15
Controller Caching for Windows NT ...................................................................... 6-16
Device Information Window ...................................................................................... 6-17
Caching Configuration ................................................................................................. 6-18
Array Groups ............................................................................................................... 6-19
Creating an Array Group .......................................................................................... 6-22
Naming an Array Group ........................................................................................... 6-24
Dynamic Array Expansion ........................................................................................ 6-24
Requirements and Restrictions ............................................................................. 6-25
Expanding an Array Group ................................................................................... 6-26
Reconfiguring Windows NT After Array Expansion ............................................... 6-27
Deleting an Array Group .......................................................................................... 6-28
Saving the Subsystem Configuration .......................................................................... 6-28
Events ......................................................................................................................... 6-29
Event Logging .......................................................................................................... 6-30
Event Notification ..................................................................................................... 6-31
Pager Event Messaging ........................................................................................... 6-33
Broadcasters ............................................................................................................... 6-34
Broadcaster for NetWare .......................................................................................... 6-34
Installing the Broadcaster ..................................................................................... 6-34
Stopping/Restarting the Broadcaster .................................................................... 6-34
Viewing Events ..................................................................................................... 6-34
Broadcaster for SCO UNIX Systems ....................................................................... 6-35
Installing the Broadcaster ..................................................................................... 6-35
Stopping/Restarting the Broadcaster .................................................................... 6-35
Viewing Events ..................................................................................................... 6-35
Broadcaster for Windows NT ................................................................................... 6-36
Installing the Broadcaster ..................................................................................... 6-36
Stopping/Restarting the Broadcaster .................................................................... 6-36
Viewing Events ..................................................................................................... 6-36
Formatting Drives ........................................................................................................ 6-38
Drive Failures .............................................................................................................. 6-39
Audible Alarm ........................................................................................................... 6-39
Rebuilding a Degraded Array ................................................................................... 6-40
x
SmartRAID V User’s Manual
Hot Spares ...............................................................................................................
Verify ...........................................................................................................................
Background Task Priority .............................................................................................
I/O Statistics ................................................................................................................
Controller I/O Statistics ............................................................................................
Device I/O Statistics .................................................................................................
Remote Communication ..............................................................................................
Communicating Across a Network ...........................................................................
Installation and Configuration ...................................................................................
Editing the Configuration File ................................................................................
Setting up the Server ............................................................................................
Connecting Servers and Workstations ..................................................................
Connecting to Remote Systems ...............................................................................
Making a Manual Connection ...............................................................................
Using the Phonebook ............................................................................................
Using a Custom Connection .................................................................................
6-41
6-42
6-43
6-44
6-44
6-45
6-48
6-49
6-49
6-50
6-51
6-52
6-53
6-54
6-55
6-56
Theory of Operation ............................................................................. 7-1
Intelligent RAID Controllers ........................................................................................... 7-2
Intelligent I/O .............................................................................................................. 7-3
Benefits ...................................................................................................................... 7-3
I2O and Intelligent RAID Controllers .......................................................................... 7-4
Pipelined Parallel Processing (P3) ................................................................................ 7-5
Disk Caching ................................................................................................................. 7-6
Software and Hardware Caching ............................................................................... 7-7
Elevator Sorting .......................................................................................................... 7-8
Caching ...................................................................................................................... 7-9
Write-back and Write-through Algorithm ................................................................. 7-9
Flush Strategy ....................................................................................................... 7-10
Predictive Caching Algorithms ................................................................................. 7-10
Virtual Cache ........................................................................................................ 7-10
Read Ahead .......................................................................................................... 7-10
Pre-Fetch .............................................................................................................. 7-11
Predictive Caching in Multi-user Systems ................................................................ 7-11
Caching Host Reads ................................................................................................ 7-12
Optimizing Operating System Cache Size ............................................................... 7-12
Optimizing Controller Cache Size ............................................................................ 7-12
RAID ............................................................................................................................ 7-13
Disk Striping ............................................................................................................. 7-13
Definition of RAID Levels ......................................................................................... 7-14
Dual-Level RAID ...................................................................................................... 7-18
Creating Data Redundancy ...................................................................................... 7-19
Handling I/O Errors .................................................................................................. 7-20
Degraded Mode ....................................................................................................... 7-21
Rebuilding a Failed Hard Drive ................................................................................. 7-21
xi
SmartRAID V User’s Manual
SmartRAID V Architecture ........................................................................................... 7-22
ASPI Protocol ........................................................................................................... 7-22
I2O Messaging Protocol ........................................................................................... 7-22
Command Processing .......................................................................................... 7-23
Command Overlapping ......................................................................................... 7-23
Command Queuing ............................................................................................... 7-24
Auto Request Sense ............................................................................................. 7-24
Byte/Word Alignment ............................................................................................ 7-24
Scatter/Gather ...................................................................................................... 7-24
PCI Data Transfer and Bus Mastering ......................................................................... 7-25
On-Board Microprocessor ........................................................................................... 7-26
SCSI Interface ............................................................................................................. 7-26
SCSI Device IDs ...................................................................................................... 7-26
Logical Unit Numbers (LUN) .................................................................................... 7-27
Bridge Controllers .................................................................................................... 7-27
Commands and Messages ..................................................................................... 7-27
Disconnect/Reconnect ............................................................................................. 7-27
Transfer Padding and Residue Reporting ................................................................ 7-28
Single-Ended and Differential SCSI ......................................................................... 7-28
Multi-Mode SCSI ...................................................................................................... 7-29
Wide SCSI ............................................................................................................... 7-29
SCSI Transfer Rate .................................................................................................. 7-29
Cabling Single-Ended SCSI ..................................................................................... 7-30
Cabling Ultra2 (LVD) SCSI ....................................................................................... 7-32
SCSI Termination ..................................................................................................... 7-33
TERMPWR .............................................................................................................. 7-34
Fibre Channel Arbitrated Loop (FC-AL) ....................................................................... 7-35
DPT I2O BIOS ............................................................................................................. 7-36
Assembly Drawings ............................................................................. A-1
PM3755F ....................................................................................................................... A-2
PM3755U2B .................................................................................................................. A-3
PM3754U2 .................................................................................................................... A-4
PM1554U2, PM2554U2 and PM2654U2 ....................................................................... A-5
SX4054U2-1 Bus Expansion Module ............................................................................ A-6
SX4054U2-2 Bus Expansion Module ............................................................................ A-7
SX4055U2-1 Bus Expansion Module ............................................................................ A-8
SX4055U2-2 Bus Expansion Module ............................................................................ A-9
SX4055F FC-AL Bus Expansion Module .................................................................... A-10
RA4050 RAID Accelerator ........................................................................................... A-11
BB4050 Battery Module .............................................................................................. A-12
Troubleshooting ................................................................................... B-1
xii
SmartRAID V User’s Manual
Specifications ....................................................................................... C-1
DC Power Requirements ............................................................................................... C-2
Environmental Specifications ........................................................................................ C-3
Memory ......................................................................................................................... C-3
Capacities .................................................................................................................. C-3
SIMMs ........................................................................................................................ C-3
DIMMs ........................................................................................................................ C-4
Battery ........................................................................................................................... C-4
SNMP ..................................................................................................... D-1
Introduction ................................................................................................................... D-2
Simple Network Management Protocol ...................................................................... D-2
Management Information Base .................................................................................. D-2
Management Console ................................................................................................ D-3
DPT MIB Information .................................................................................................. D-3
What’s Included ............................................................................................................. D-6
System Requirements ................................................................................................... D-6
Hardware Support ...................................................................................................... D-6
Management Consoles .............................................................................................. D-6
Operating System Support ......................................................................................... D-7
Installation of DPT SNMP Software ............................................................................... D-7
Installing DPT SNMP Support .................................................................................... D-7
Installing SNMP for Microsoft Windows 95 ................................................................. D-8
Installing SNMP for Microsoft Windows 98 ................................................................. D-8
Operation ...................................................................................................................... D-9
SNMP Software Overview ............................................................................................. D-9
DPT SNMP Sub-Agent ............................................................................................. D-10
DPT SNMP Trap Broadcaster Module ...................................................................... D-11
Glossary ............................................................................................... G-1
Index ....................................................................................................... I-1
xiii
SmartRAID V User’s Manual
xiv
CHAPTER 1:
Read This First!
Before you begin installing your new
DPT controller, please take the time to
read this chapter. This chapter is an
important guide to the rest of the
documentation and provides a summary
of the installation process.
The term RAID controller used
throughout this document applies to all
SmartRAID V products.
SmartRAID V
Documentation
A chapter-by-chapter tour of the
documentation set. This includes
the User’s Manual you are
reading, a Quick Install Guide
and the Storage Manager Help
Files.
Installation Roadmap
The SmartRAID V installation
process at a glance, with crossreferences to specific chapters in
this manual.
What’s in the Box
A list of parts and accessories
you received with your new
controller.
System Requirements
What type of host system you
need to successfully install and
use a DPT SmartRAID V
controller.
Safety Information
Notices regarding electrical
safety and electrostatic
discharge protection.
1-1
1-2
Read This First!
About the Documentation
The complete documentation set for the SmartRAID V product line
consists of three parts:
• The User’s Manual (this book), which contains information that helps
you to configure and install your SmartRAID V product and using the
Storage Manager software. This document also contains information
about using the DPT storage subsystem utility – Storage Manager on
ROM; background information about the peripheral bus, SCSI, RAID,
and caching; system specifications; troubleshooting tips; a glossary of
terms; and other topics of interest to SmartRAID V users.
• The Quick Install Guide, a brief version of installation steps designed
for the experienced user. This document contains instructions in
English, French, German, Italian, Polish and Spanish.
• The Storage Manager Help System, which contains information about
using the Storage Manager software, using SCSI devices and creating
disk arrays.
NOTE
An updated version of this manual (in Adobe Acrobat PDF
format) may be available for download on the DPT web site
(www.dpt.com). SmartRAID V Quick Install Guides are also
available for download from the DPT web site.
User’s Manual
The User’s Manual contains seven chapters and 4 appendices.
Chapter 1, Read This First – This chapter provides an overview of
the rest of the documentation, and a roadmap of the installation
process.
Chapter 2, About Your New SmartRAID V Controller – This chapter
describes the features of the various SmartRAID V controllers, and
the add-on modules that you can use with your controller.
Chapter 3, Configuration and Installation – This chapter provides
instructions about setting up device IDs and termination,
assembling the main board, plug-in modules and SIMM or DIMM
memory modules, selecting the proper cables, and installing the
controller into your PC.
Chapter 4, Storage Manager on ROM – This chapter provides
instructions for using the Storage Manager on ROM (SMOR) utility.
SMOR is a DPT storage management utility that is part of the
controller ROM and is available during boot. You can use SMOR to
configure your controller and create disk arrays before installing an
operating system on your computer.
Read This First!
1-3
Chapter 5, Software Installation – This chapter provides
instructions for installing DPT drivers and Storage Manager for
all supported operating systems. After the hardware is configured
and the disk arrays created, you can install the operating system,
SmartRAID V drivers and the full, OS-specific version of Storage
Manager.
Chapter 6, Storage Manager – This chapter provides information
about using DPT’s Storage Manager software. Storage Manager
is a utility that lets you access the full feature set of your
SmartRAID V controller and manage your storage subsystem.
Chapter 7, Theory of Operations – This chapter provides an
overview of the technology used in the SmartRAID V series of
controllers: caching, RAID, DPT controller architecture, I2O,
PCI, SCSI, Fibre Channel and other related topics.
Appendix A, Assembly Drawings – This appendix provides outline
drawings of the SmartRAID V circuit boards. These drawings assist
you in locating the various components on the boards.
Appendix B, Troubleshooting – This appendix is a list of common
problems and suggested solutions.
Appendix C, Specifications – This appendix provides electrical and
environmental specifications for the SmartRAID V product line.
Appendix D, SNMP – This appendix describes the DPT
implementation of the Simple Network Management Protocol
feature for SmartRAID V hardware.
Glossary – The Glossary provides definitions of terms and
acronyms used throughout the documentation.
Storage Manager Help System
The Storage Manager online help information contains both topical and
pop-up helps for Storage Manager, DPT products and RAID concepts.
Installation Roadmap
The figure on the following page provides an overview of the installation
process, and refers to specific chapters in the documentation that provide
information about each step.
1-4
Read This First!
Using RAIDstation?
YES
NO
Configure Cabinet
Refer to: Storage Cabinets and Devices manual
Fibre Channel?
NO
Set SCSI ID
Configure Termination
See Chapter 3, Configuration and Installation
YES
YES
Add-on Modules?
NO
Plug in RA4050 / SX405x
Plug in cache memory
Plug in BB4050
See Chapter 3, Configuration and Installation
Install the controller
Attach cables
See Chapter 3, Configuration and Installation
Power on system
Run SMOR (Storage Manager on ROM)
Set termination option
Verify hardware
Configure disk array
See Chapter 4, Storage Manager on ROM
YES
New System?
NO
Install operating system
Install drivers
See Chapter 5, Software Installation
Install drivers
See Chapter 5, Software Installation
Install Storage Manager
Read This First!
1-5
What’s In the Box?
All SmartRAID V controller kits include a controller, utility diskettes
(containing drivers and the Storage Manager software for supported
operating systems), and documentation. The documentation set consists
of this User’s Manual, a Quick Install Guide, and the Storage Manager
online help system.
NOTE
SmartRAID V Ultra2 SCSI controllers include the following
additional items:
• One internal Ultra/Ultra2 4-device Wide SCSI cable.
• A multimode, Ultra2 terminator (installed on the end of the
included cable).
System Requirements
The DPT Storage Manager software and device drivers require
approximately 2MB of disk space. See Chapter 5, “Software Installation”
for information about supported operating systems.
All SmartRAID V controllers are PCI 2.2 compliant and are designed to
operate in host systems that comply with revision 2.2 of the PCI
specification.
SmartRAID V Decade and Century controllers are multifunction PCI
devices. The host system must be able to properly configure multifunction
PCI devices, where one of the devices is a bridge.
SmartRAID V Millennium controllers include a PCI bridge. The host
system must be able to properly configure PCI bridges and any devices
located behind the bridge.
Refer to the file READ.ME in the DPT install directory for more
information about computer systems and motherboards that DPT has
tested for compatibility with SmartRAID V controllers. You can also select
Help–View README Information in Storage Manager to see this
information.
1-6
Read This First!
Safety Information
Throughout this manual are various notices that indicate procedures or
practices that can result in loss of data, damage to equipment or personal
injury. Be sure to read the following sections for additional information
regarding electrical shock hazards and preventing damage from
electrostatic discharge.
The following symbol will accompany IMPORTANT and CAUTION notices
that can result in damage to equipment or loss of data.
!
These notices can be identified as shown by the following examples:
NOTE
This type of notice is used to emphasize important
information or procedures that should be followed but if
ignored would not result in permanent damage to equipment
or software.
IMPORTANT
This type of notice is used to emphasize a procedure that, if not followed,
can cause malfunction of the equipment or software or result in loss of data.
CAUTION
This type of notice is used to indicate hazards or unsafe practices
which could result in personal injury or product-property damage.
!
Read This First!
1-7
Working with Electricity
Any device that uses electricity must be treated with caution. Follow these
guidelines to ensure general safety.
• Keep the chassis area clear and dust-free during and after installation.
• Do not perform any action that creates a potential hazard to people or
makes the equipment unsafe.
• Before working on the system, unplug the power cord.
• Disconnect all power before doing the following:
– Installing or removing a chassis
– Working near power supplies
• Do not work alone when potentially hazardous conditions exist.
• Never assume that power has been disconnected from a circuit.
Always check.
• Look carefully for possible hazards in your work area, such as moist
floors, ungrounded power extension cables, or missing safety grounds.
Notices in this manual with the following symbol indicate a potential
electric shock hazard to personnel or equipment.
1-8
Read This First!
Preventing Electrostatic Discharge
CAUTION
Electrostatic discharge (ESD) can damage electronic
components and equipment. ESD occurs when electronic
components are improperly handled and can result in
complete or intermittent failures. Always follow ESDprevention procedures when removing and replacing
components.
Use the following guidelines to prevent ESD damage:
• Always use an ESD wrist or ankle strap and ensure that it makes skin
contact.
• Connect the equipment end of the strap to an unpainted metal chassis
surface.
• If no wrist strap is available, ground yourself by touching the metal
chassis.
• When installing a component, use any available ejector levers or
captive installation screws to properly seat the bus connectors in the
backplane or card slot. These devices prevent accidental removal,
provide proper grounding for the system, and help to ensure that bus
connectors are properly seated.
• When removing a component, use any available ejector levers or
captive installation screws to release the bus connectors from the
backplane or card slot.
• Handle adapter cards by available handles or edges only. Avoid
touching the printed circuit boards or connectors.
• Place a removed component board-side-up on an antistatic surface or
in an approved antistatic container.
• If you plan to return the component to DPT, immediately place it in a
static shielding container.
• Avoid contact between printed circuit boards and clothing. The wrist
strap only protects components from ESD voltages on the body; ESD
voltages on clothing can still cause damage.
CHAPTER 2:
About Your New Controller
This chapter presents the features of DPT
SmartRAID V SCSI and Fibre Channel
controllers, along with the Bus Expansion,
RAID Accelerator and Battery Backup
modules.
Feature Overview
A look at the features common to
SmartRAID V hardware.
Millennium
Specific features of our high
performance RAID controller.
Decade and Century
What makes our low cost RAID
controllers so powerful, and how
you can upgrade them to meet your
growing needs.
SX405x and RA4050
Expansion Modules
SX405x modules add additional
SCSI or Fibre Channel busses. An
RA4050 module upgrades your
Decade or Century controller to
hardware RAID and a maximum
64MB of cache.
BB4050 Battery Module
The BB4050 provides up to 72
hours of battery backup safety
for your data cache when using
a PM3755U2B controller.
LEDs and Audible Alarm
Information about LED indicators
and the audible alarm feature.
2-1
2-2
About Your New SmartRAID V Controller
Introduction
DPT SmartRAID V products incorporate the latest intelligent controller
technology to deliver optimum performance for desktop systems, file
servers or multi-user host systems.
NOTE
ECC-protected cache is available when you use DPT SM4050 or
DM4050/DM4060 memory modules.
PM3755U2 controllers require DPT DM4050 or DM4060 memory
modules. Non-DPT DIMM modules are not supported.
€ Millennium controllers support RAID 0, 1, 5 and multilevel RAID
(0+1 and 0+5). Millennium controllers support a maximum of 256MB
of onboard cache.
• Century controllers support RAID 0 and RAID 1. You can upgrade to
hardware RAID 5 support and up to 64MB of onboard cache by adding
an RA4050 RAID Accelerator module.
Century PM2654U2 controllers support RAID 0, RAID 1 and RAID 5
with a pre-installed RA4050 RAID Accelerator. You can have a
maximum of 64MB of cache on the RA4050 module.
• Decade controllers support RAID 0, 1 and 5 and include 4MB of
onboard cache.
You can upgrade a Decade controller by adding an RA4050 RAID
Accelerator module to support hardware XOR for RAID 5 and a
maximum of 64MB of onboard cache.
• SX405x Bus Expansion Modules add extra SCSI channels or a Fibre
Channel port to SmartRAID V controllers. You can use a Fibre
Channel expansion module to provide both SCSI and Fibre Channel
connections on a SmartRAID V SCSI controller.
• The RA4050 RAID Accelerator provides up to 64 MB of cache and
adds hardware RAID 5 capability to a Century controller or adds
hardware XOR and up to 64MB of cache to a Decade controller.
• The BB4050 Battery Module provides battery backup for the onboard
cache on a Millennium PM3755U2B controller.
About Your New SmartRAID V Controller
2-3
Product List
The following SmartRAID V products and accessories are documented in
this manual:
Model
PM3755F
PM3755U2B
PM3754U2
Description
64-bit PCI to Fibre Channel RAID Controller
64-bit PCI to Ultra2 Wide SCSI RAID Controller
(w/battery backup capability)
32-bit PCI to Ultra2 Wide SCSI RAID Controller
PM2654U2-R1
PM2554U2
PM1554U2
32-bit PCI to Ultra2 Wide SCSI RAID Controller
32-bit PCI to Ultra2 Wide SCSI RAID Controller
32-bit PCI to Ultra2 Wide SCSI RAID Controller
SX4055F
64-bit Fibre Channel Single Bus Expansion Module
SX4055U2-1
64-bit Ultra2 Wide SCSI Single Bus Expansion Module
SX4055U2-2
64-bit Ultra2 Wide SCSI Dual Bus Expansion Module
RA4050
RAID Accelerator Module
BB4050
Battery Backup Module
SM4050-16
SM4050-64
16MB ECC SIMM
64MB ECC SIMM
DM4050-162
DM4050-642
DM4060-162
DM4060-642
16MB ECC DIMM
64MB ECC DIMM
16MB ECC DIMM
64MB ECC DIMM
1. PM2654U2 controllers are shipped with an RA4050 RAID
Accelerator preinstalled.
2. PM3755U2B controllers use DIMMs for cache memory.
2-4
About Your New SmartRAID V Controller
Product Feature Overview
All SmartRAID V controllers include:
• Support for I2O OSMs provided by major operating system vendors;
DPT provides drivers for some operating systems. See Chapter 5,
“Software Installation” for a list of drivers supplied by DPT.
• Certifications for major operating systems, including Novell NetWare
and Windows NT.
• Support for a variety of SCSI and Fibre Channel devices, including
hard disk, tape, CD-Recordable, CD-ROM, Magneto-Optical drives,
jukeboxes and scanners.
• Local and remote configuration, array status and I/O monitoring using
DPT Storage Manager software.
• Operating system independent configuration and RAID creation using
the Storage Manager on ROM (SMOR) utility.
• Support for SCSI-1, SCSI-2 and SCSI-3 with active termination.
• ASPI protocol support for third-party applications and utilities.
• Flash ROM for easy upgrades of controller firmware, I2O BIOS and
SMOR.
• Event logging and broadcasting, including messaging for
alphanumeric pagers.
• Dynamic Array Expansion, which lets you increase your storage
capacity under Windows NT by adding one or more drives to a RAID 0
or RAID 5 array while your system remains online. See Dynamic
Array Expansion in Chapter 6, “Storage Manager” for information on
using this feature.
• Predictive caching which analyzes disk read requests made by the
host to determine whether they are part of a pattern. If a pattern is
detected, the controller uses the pattern to predict which data the host
is likely to request in the near future, then reads this data into the
cache.
• Intelligent Hot Spare. A hot spare drive automatically replaces a failed
drive. When multiple hot spares are available on a controller, the
intelligent hot spare algorithm picks the best one based on capacity
and bus location. RAID 1 and RAID 5 arrays are rebuilt automatically
using the new drive.
About Your New SmartRAID V Controller
2-5
SmartRAID V Millennium Features
SmartRAID V Millennium products are high-performance RAID
controllers for host computers with a PCI expansion bus.
Key features of the Millennium include:
• High-performance RISC processor (80 MIPs)
• Hardware RAID 0, 1, and 5
• Support for striping multiple arrays as a single logical drive
(RAID 0+1 and RAID 0+5)
• Hardware XOR
• Up to 256MB of cache per controller
• 32-bit PCI and 64-bit PCI bus models
Conforms to revision 2.2 of the PCI specification. PCI clock speeds up
to 33 MHz are supported.
• One Ultra2 SCSI (PM375xU2) or Fibre Channel (PM3755F) bus
• Intelligent Hot Spare capability
• ECC protection for the data cache (using DPT ECC-capable memory
modules)
• Real-time monitoring of controller temperature and voltage
• Optional battery backup capability for DM4050/DM4060 controller
cache
The following Millennium controllers are available:
Type
Host Bus
Host/Internal Transfer Rate
I/O Transfer Rate1
Type
Host Bus
Host/Internal Transfer Rate
I/O Transfer Rate1
Battery Backup
1. Maximum per channel burst rate
2. Using a single 16MB DIMM
PM3754U2
Ultra2 Wide
32-bit PCI
132 MB/sec
80 MB/sec
PM3755U2B
Ultra2 Wide
64-bit PCI
264 MB/sec
80 MB/sec
72 hours2
PM3755F
Fibre Channel
64-bit PCI
264 MB/sec
100 MB/sec
2-6
About Your New SmartRAID V Controller
You can use an SX405x Bus Expansion Module to expand a Millennium
controller with one or two additional SCSI channels or with one Fibre
Channel Arbitrated Loop (FC-AL) port. Millennium Fibre Channel
controllers can be expanded with one additional FC-AL port. By using an
SX4055F Bus Expansion Module, Fibre Channel and SCSI buses can
coexist on the same SCSI controller (PM375xU2 models only). SX405x Bus
Expansion Modules are daughter cards that attach to the SmartRAID V
controller and do not require an additional card slot in the host system.
The PM3755U2B controller can be upgraded to include a battery-backed
cache by adding a BB4050 battery module.
• For controlling DPT RAIDstation storage cabinets or other
manufacturers’ cabinets, Millennium SCSI controllers provide a
Very High Density Cable Interconnect (VHDCI) 0.8 mm external
connector.
• To simultaneously control SCSI devices mounted in the host system,
Millennium SCSI controllers provide an internal, high-density 68-pin
connector.
• The Millennium Fibre Channel controller provides a single High
Speed Serial Data Connector (HSSDC) for external connections. Dual
loop redundancy is available by adding an SX4055F Expansion Module.
The memory sockets on a Millennium controller accept up to 256MB of
cache for maximum performance. For maximum data protection, use the
DPT ECC SIMMs or DIMMs listed below:
Model
NOTE
Description
SM4050-16
16 MB ECC SIMM
SM4050-64
64 MB ECC SIMM
DM4050-16
16 MB ECC DIMM
DM4050-64
64 MB ECC DIMM
DM4060-16
16 MB ECC DIMM
DM4060-64
64 MB ECC DIMM
The PM3755U2B controller only accepts DM4050 or DM4060
modules. Do not use non-DPT DIMMs for this controller.
The PM3754U2 controller can use either standard 32- or 36-bit1 singlesided, low-profile 60ns (or faster) EDO SIMMs. However, standard EDO
SIMMs do not provide ECC data protection. The PM3754U2 data cache is
protected by ECC only when DPT SM4050 memory modules are installed.
The green ECC Enabled LED is lit when DPT memory is installed. Refer
to Appendix A, “Assembly Drawings” for the location of this LED.
1
The parity bit is not used on 36-bit EDO SIMMs. However, the SIMMs will function in the controller as
non-parity memory.
About Your New SmartRAID V Controller
2-7
SmartRAID V Decade and Century Features
SmartRAID V Decade and Century controllers are designed to provide
high performance solutions for workstations and servers in environments
where cost is a factor.
The following Decade and Century controllers are available:
PM1554U2 / PM2554U2 / PM2654U2
Ultra2 Wide
Type
32-bit PCI
Host Bus
Host/Internal Transfer Rate
I/O Transfer Rate
1
132 MB/sec
80 MB/sec
1. Maximum transfer rate per channel
The Decade (PM1554U2 ) controller features include:
• Onboard I/O processor (31 MIPs)
• Hardware RAID 0, 1, 5, 0+1 and 0+5
• Firmware XOR
• 4MB of onboard cache RAM
• Intelligent Hot Spare capability
• Conforms to revision 2.2 of the PCI specification
NOTE
If an RA4050 module is attached to a PM1554U2 controller, the
controller’s onboard cache is disabled and only the cache on
the RA4050 cache is used (64MB maximum). The RA4050 also
adds hardware XOR capability which enhances RAID 5
performance.
Century controller features include:
• Onboard I/O processor
PM2554U2 (31 MIPs)
PM2654U2-R (40 MIPs)
• Hardware RAID 0, 1 and 0+1
• RAID 5 and 0+5 when using an RA4050 RAID Accelerator
• Up to 64MB of cache when using an RA4050 RAID Accelerator
• Intelligent Hot Spare capability
• Conforms to revision 2.2 of the PCI specification.
2-8
About Your New SmartRAID V Controller
The Century controller provides a single Ultra2 SCSI channel and
supports up to two additional channels with an SX405x Bus Expansion
Module. The SX405x is available in three models, providing either one or
two additional SCSI channels or one Fibre Channel Arbitrated Loop
(FC-AL) port. By using an SX4055F Bus Expansion Module both Ultra2
SCSI and Fibre Channel can coexist on the same controller.
Both the RA4050 RAID Accelerator and the SX405x Bus Expansion
Module are daughter cards that attach to the controller and do not require
an additional card slot in the host system. PM2654U2 controllers are
shipped with RA4050 and/or SX405x expansion modules already installed.
SX405x Bus Expansion Module Features
SX405x Bus Expansion Modules provide additional channels for
SmartRAID V controllers. The SX4055F Bus Expansion Module, can be
used to add a FC-AL port to a SmartRAID V Ultra2 SCSI controller or a
second Fibre Channel bus to a SmartRAID V Fibre Channel controller.
NOTE
SX405xU2 modules are not compatible with SmartRAID Fibre
Channel controllers. The SX4055U2 Bus Expansion Module is
not compatible with SmartRAID V controllers that have a 32-bit
connector.
The following DPT Bus Expansion Modules are available:
SX4054U2-1 (-2)
SX4055U2-1 (-2)
SX4055F
Type
Ultra2 Wide
Ultra2 Wide
Fibre Channel
Channels
Single (Dual)
Single (Dual)
Single
1
I/O Transfer Rate
80 MB/sec
80 MB/sec
100 MB/sec
Width
32-bit
64-bit
64-bit
Internal PCI Transfer Rate
132 MB/sec
264 MB/sec
264 MB/sec
1. Maximum burst rate per channel
About Your New SmartRAID V Controller
2-9
SX405x modules attach to the SmartRAID V controller and do not require
an additional slot in the host system:
• A single channel SX405xU2-1 Ultra2 SCSI Bus Expansion Module adds
one SCSI channel and provides one external Very High Density Cable
Interconnect (VHDCI) 0.8 mm connection and one internal highdensity 68 pin-connection.
• A dual channel SX405xU2-2 Ultra2 SCSI Bus Expansion Module adds
two SCSI channels and provides two external VHDCI 0.8 mm
connections and two internal high-density 68-pin connections.
• A Fibre Channel SX4055F Bus Expansion Module adds a single FC-AL
bus with a High Speed Serial Data Connector (HSSDC). Using this
module with a Fibre Channel controller provides dual-loop
redundancy.
RA4050 RAID Accelerator Features
The RA4050 RAID Accelerator module adds up to 64 MB of cache
hardware RAID 5 and 0+5 to a Century controller or additional cache and
hardware XOR to a Decade controller. The RA4050 connects to the end of
the controller and does not require an additional slot in the host system.
The RA4050 hardware cache can provide a substantial performance
improvement in systems with heavy disk I/O load, such as CAD
workstations, and in network servers running operating systems such as
Windows NT, Novell NetWare or UNIX.
The RA4050 supports up to 64 MB of cache using four SIMM sockets. Each
socket accepts either a 16MB or 64MB standard 32- or 36-bit2 single-sided,
low-profile 60ns (or faster) EDO SIMM, or one of the DPT ECC SIMMs
listed below:
Model
Description
SM4050-16
16 MB ECC SIMM
SM4050-64
64 MB ECC SIMM
DPT SM4050 SIMMs are recommended for systems where data integrity
is critical. Although, standard EDO memory modules can be used to
provide RAM for the cache, SM4050 SIMMs protect the cache with ECC
capability. The green ECC Enabled LED is lit when DPT memory is
installed. Refer to Appendix A, “Assembly Drawings” for the location of
this LED.
2
The parity bit is not used on 36-bit EDO SIMMs, however, the SIMMs will function in the controller as
non-parity memory.
2-10
About Your New SmartRAID V Controller
BB4050 Battery Module Features
The BB4050 Battery Module adds battery backup for the data cache on a
SmartRAID V PM3755U2B controller using DPT DM4050 or DM4060
memory modules.
The following features of the BB4050 provide additional data security for
high-reliability servers:
• Nickel Metal Hydride battery pack provides a backup time of 72 hours
(using a single 16MB memory module). Actual backup time depends
on the amount of cache used and is calculated by the controller
firmware.
• On board intelligence for power crossover and charge management.
• The battery status and backup capacity can be monitored using
Storage Manager.
• Fast recharge (five hours).
• Battery life of 300 charge/discharge cycles.
• A user initiated discharge/recharge cycle is available to prevent loss of
battery capacity through voltage depression by performing periodic
deep-discharging and recharging of the battery.
• Operating temperature range of 10ºC to 40ºC.
!
CAUTION
!
Danger of explosion if battery is incorrectly replaced. Replace only
with the same or equivalent type recommended by the
manufacturer. Dispose of used batteries according to the
manufacturer’s instructions.
About Your New SmartRAID V Controller
2-11
SmartRAID V Controller LEDs
SmartRAID V controllers provide nine LEDs that let you visually monitor
controller activity. Several different controller states are indicated by the
LED as outlined in the following sections. Two additional LEDs indicate
the status of the cache memory on the controller. See Appendix A,
“Assembly Drawings” for the location of the LEDs on your controller.
LED Display During Power-Up
During its power-up sequence, the controller passes through the following
states in sequence, as indicated by the LEDs:
NOTE
Memory Mapped I/O
LEDs 6 and 7 and LEDs 5 and 8 flash alternately
while the controller waits for the host computer to
initialize the address translation unit (ATU) on the
PCI bus.
Bridge
LEDs 5 and 7 and LEDs 6 and 8 flash alternately
while the controller waits for the host computer to
initialize its PCI-to-PCI bridge.
Fast Idle
The controller enters a fast idle pattern while
waiting for the I2O initialization commands after
the address translation unit (ATU) and PCI bridge
have been enabled.
The ATU, PCI bridge and cache module (if installed) must be
configured and enabled before the I2O initialization can be
completed. If any one of these components are not configured
or become disabled during the Fast Idle phase, the pattern will
return to the respective wait pattern for the affected
component.
LED Display During Controller Idle
When no commands are in progress and all bus activity has ceased, the
controller enters the Idle state. This is indicated by a rotating pattern in
LEDs 1 through 8.
2-12
About Your New SmartRAID V Controller
LED Display During Controller Active
When the power-up sequence is complete, viewing the LEDs can help you
determine the operating status of the controller. These LED patterns are
also useful for troubleshooting. Refer to Appendix B, “Troubleshooting” for
more information on interpreting LED patterns.
LED
Function
1
Heartbeat. Indicates that controller interrupts are enabled and
that the controller is alive. During controller activity, this LED
flashes four times a second.
2
Indicates the occurrence of a non-maskable interrupt (NMI) to the
I/O processor on the controller.
3
Indicates the controller’s internal operating system is in its idle
loop.
4
Indicates that the controller’s internal operating system is
processing an interrupt.
5
Reserved for future use.
6
Indicates that the cache controller is using DMA to perform a data
transfer.
7
Indicates the controller is generating parity information for a RAID
5 array (hardware XOR).
8
Indicates that there is a command on the SCSI or Fibre bus.
IRQ
Lit only when the controller activates an interrupt on the host PCI
bus.
Cache Status LEDs
Two LEDs on the SmartRAID V controller indicate the status of the
onboard cache RAM. Refer to Appendix A, “Assembly Drawings” for the
location of the LEDs on your controller.
• The green ECC Enabled LED is lit when all installed SIMMs or DIMMs
are DPT ECC memory modules. This indicates that the controller data
cache is ECC protected.
• The red ECC Error LED, is lit when a correctable or non-correctable
error has been detected in one of the ECC SIMMs or DIMMs. The
LED will stay lit, even after the error has been corrected, until the
controller is powered-down. Cache failure information is recorded in
the controller error log and can be viewed using the Event Log
window in Storage Manager.
About Your New SmartRAID V Controller
2-13
BB4050 LED Indicators
The BB4050 module has two LED indicators labeled CHARGE and
TRICKLE. These LEDs indicate the following conditions:
Status
TRICKLE LED
CHARGE LED
Momentary flash
On
Fast Charge
Off
On
Trickle Charging
On
Off
Charge inhibit
Off
Off
Flashing
On
On
On
1
Initial Powerup
Discharge
Battery not installed
1. If battery voltage and the ambient temperature are
acceptable, the Trickle LED will not remain lit.
Use the Battery Configuration dialog in Storage Manager to view the
current status of the BB4050 module.
Audible Alarm
DPT controllers with caching capability (Millennium models or Decade/
Century models using an RA4050 module) have an audible alarm. The
failure of a drive which is a member of an array attached to the controller
causes the audible alarm to sound. The alarm stops automatically (after
the initial system scan) when you start Storage Manager or SMOR.
2-14
About Your New SmartRAID V Controller
CHAPTER 3:
Configuration and Installation
Your new DPT SmartRAID V controller is
the center of a complete system consisting
of the controller, disk drives and other
peripherals, and the connecting cables.
This chapter discusses configuring these
components.
Configure Devices
This chapter includes information on using
SmartRAID V controllers in a Microsoft
Cluster Server (MSCS) configuration.
Attach Modules
Prepare your SCSI peripherals for
use by setting their SCSI IDs and
termination. Fibre Channel disks
and peripherals will configure
themselves automatically during
boot.
Assemble the various parts of your
DPT controller, which can include
the base controller, a RAID
Accelerator, Bus Expansion Module
and one or more memory modules.
The PM3755U2B controller can also
use the battery backup module.
Install in your System
Mount the assembled controller in
your computer system.
Configure the Software
Use our SMOR utility to set your
system parameters, install the
operating system and you are ready
to use your system.
Microsoft Cluster Server
How to use a SmartRAID V
controller in a Microsoft Cluster
Server configuration.
3-1
3-2
Adapter Configuration and Installation
Installation Overview
The process of installing a SmartRAID V controller consists of the
following steps:
1.
If you are using a RAIDstation storage cabinet, configure it as
described in the RAIDstation User’s Manual.
2.
Configure device IDs, cables and termination for SCSI devices in
the host system. Fibre Channel controllers and devices will
automatically configure themselves during boot.
3.
If you are using a RAID Accelerator, adding a Bus Expansion
Module, expanding your cache memory or using a battery module,
plug the appropriate modules onto the controller.
NOTE
SmartRAID V Millennium controllers or Century and Decade
controllers with an RA4050 module must have cache installed.
DPT recommends that SmartRAID V controllers have at least
16MB of cache installed.
4.
Install the controller and storage devices in the appropriate
enclosures. Attach all cables between the controller and the storage
devices.
5.
Run Storage Manager on ROM (SMOR) by pressing Ctrl+D during
system boot to configure the controller’s SCSI termination and
verify proper hardware configuration. You can also use SMOR to
configure your storage subsystem and disk arrays. (Refer to
Chapter 4, “Storage Manager on ROM”.)
6.
If you are setting up the computer system for the first time, install
the operating system on one of the controller’s disk drives or
arrays. During this process, you should also install any driver
updates for your DPT controller. Refer to Chapter 5, “Software
Installation” for specific instructions.
7.
Install any required operating system drivers and Storage
Manager. Refer to Chapter 5, “Software Installation” for additional
information.
Storage Manager for Microsoft Windows is on the DPT Windows
NT, Windows 95/98 diskette. Storage Manager for SCO UNIX (Motif
version) is on the SCO UNIX diskette.
Adapter Configuration and Installation
3-3
Configuration
DPT Fibre Channel controllers and attached Fibre Channel devices are
plug-and-play and will configure themselves automatically during boot.
SCSI devices in the system must be configured prior to use. This
configuration includes enabling SCSI termination for the devices and
setting the SCSI IDs for each device.
Narrow and Wide SCSI
The SCSI devices you will be installing can be either Narrow (8-bit) or
Wide (16-bit) SCSI devices. Wide SCSI disk drives allow data to be
transferred at twice the rate of older 8-bit devices. However, some SCSI
devices such as tape and CD-ROM drives still use an 8-bit interface. This
is not a problem because the Wide SCSI bus is backward compatible with
Narrow SCSI devices, allowing both types of SCSI devices to be used on
the same controller.
NOTE
You can mix Narrow and Wide devices on a single 16-bit SCSI
cable. However, any Narrow devices must be attached to the
bus ahead of any Wide devices. This ensures that the 16-bit
signals are correctly terminated.
Configuring Cables
SmartRAID V SCSI controllers contain a single Wide SCSI bus with one
internal and one external SCSI connector. Each SmartRAID V SCSI
controller kit contains a Wide, 68-conductor SCSI cable. This cable is for
connections internal to the computer cabinet only.
SmartRAID V Fibre Channel controllers contain one Fibre Channel port
with a single Fibre Channel connector. The SmartRAID V Fibre Channel
controller requires a High Speed Serial Data Connector (HSSDC) cable. If
you do not have an HSSDC cable, this cable is available from DPT or a
cable supplier of your choice.
NOTE
Internal and external, Narrow and Wide SCSI cables, Fibre
Channel cables, connector adapters and terminators can be
purchased directly from DPT (you can use our online web
store at www.dpt.com) or from a supplier of your choice.
3-4
Adapter Configuration and Installation
Configuring SCSI Termination
The devices on each physical end of a SCSI cable must be terminated.
Depending upon how you configure your system, you will either terminate
two SCSI devices, or the SCSI controller and one peripheral SCSI device.
NOTE
The SCSI cable supplied with your DPT controller has a
factory-installed active terminator on the end of the cable.
If you are using multiple SCSI busses on a single controller,
each separate bus must be terminated.
SCSI termination for SmartRAID V controllers is configured through the
SMOR utility or from the Configure Host Bus Adapter window in Storage
Manager. The controller has four possible termination settings:
Auto
The default setting. This can be used for all
cabling conditions except where a Narrow (8-bit)
cable is attached to the controller. For this case
use High Only.
Disabled
Turns off controller termination unconditionally.
Enabled
Turns on controller termination unconditionally.
High Only
Terminates only the additional signals that are
used on Wide SCSI devices. This allows Wide and
Narrow cables to be simultaneously attached to
the controller.
By using a 68-pin to 50-pin SCSI cable adapter, an 8-bit SCSI device can be
attached to a Wide SCSI cable along with Wide SCSI devices. However, the
device at the end of the cable must be a Wide SCSI device so that all SCSI
signals are terminated. For internal and external cables where one cable
is an 8-bit (Narrow) SCSI cable, set the controller termination to High
Only.
3-5
Adapter Configuration and Installation
The following illustrations show various SCSI cabling examples.
Terminate your SCSI devices as shown in the examples, ensuring that only
the devices at the ends of the cables are terminated.
Internal
Narrow SCSI
Internal
Wide SCSI
T
T
External
Narrow SCSI
External
Wide SCSI
T
T
SmartRAID V
Termination = High Only
SmartRAID V
Termination = Auto
Internal
Narrow SCSI
T
Internal
Wide SCSI
T
External
Wide SCSI
T
T
Internal
Wide SCSI
External
Wide SCSI
T
T
T
SmartRAID V
Termination = Auto
External
Narrow SCSI
T
SmartRAID V
Termination = High Only
T = Terminated
3-6
Adapter Configuration and Installation
Configuring Device IDs
The SCSI specification allows up to 7 SCSI devices (and a controller) to be
connected to a single 8-bit SCSI bus. A Wide SCSI bus can support up to 15
devices (and a controller). Up to 126 devices can be attached to a single
Fibre Channel loop.
SCSI
All SCSI devices, including the controller, must be assigned a unique SCSI
ID. SCSI IDs, which are typically set using jumpers or switches on
peripheral devices, can be assigned any number from 0 to 7 for 8-bit SCSI
devices or 0 to 15 for Wide SCSI devices. For more information on SCSI
IDs, see Chapter 7, “Theory of Operation.”
Set the SCSI ID of each SCSI device attached to the controller to a unique
ID number between 0 and 6. The SmartRAID V controller is set to ID 7 by
default (most SCSI controllers use ID 7.) Wide SCSI devices can also use
SCSI IDs 8 through 15. SCSI IDs can be duplicated on the same controller
if the devices using the same ID are not attached to the same bus.
If necessary, the SmartRAID controller ID can be changed to any ID 0 – 7
by using SMOR or from the Configure Host Bus Adapter window in
Storage Manager.
NOTE
Changing the controller ID is not recommended. You should
leave the SmartRAID V controller at SCSI ID 7.
Fibre Channel
DPT Fibre Channel controllers and their attached Fibre Channel devices
are configured automatically during boot. Device IDs are assigned within
the range 0 – 126 and cannot be changed manually.
Adapter Configuration and Installation
3-7
Installation
NOTE
SmartRAID V Millennium controllers and Century or Decade
controllers with an RA4050 module should have at least 16MB
of cache installed.
Several optional modules are available to upgrade the performance of
your SmartRAID V controller:
• RA4050 RAID Accelerator – adds RAID 5, hardware XOR and cache
memory capacity to the Decade and Century controllers.
• SX4054U2 Bus Expansion Module – adds SCSI channels to a
SmartRAID V Decade or Century controller.
• SX4055U2 Bus Expansion Module – adds SCSI channels to a 64-bit
SmartRAID V controller.
• SX4055F Bus Expansion Module – adds one Fibre Channel port to a
SmartRAID V controller.
• SM4050-16 (-64) ECC SIMM – adds 16 or 64 MB of cache memory to
SmartRAID V Millennium non-battery backup controllers or to
SmartRAID V controllers using the RA4050 RAID Accelerator.
• DM4050-16 (-64) or DM4060-16 (-64) ECC DIMM – adds 16 or 64 MB
of cache memory to SmartRAID V PM3755U2B controllers.
• BB4050 – adds battery-backed cache capability to a PM3755U2B
controller. The battery-backed cache must be DPT DM4050 or DM4060
memory modules. Non-DPT DIMMs are not compatible with the
BB4050.
3-8
Adapter Configuration and Installation
Installing the RA4050 RAID Accelerator
SmartRAID V Century controllers can be upgraded to support RAID 5,
0+5 and data caching by installing the RA4050 RAID Accelerator module.
A Decade (PM1554U2) controller has built-in RAID 5 and 0+5 support
without an RA4050 module. Adding an RA4050 module to a Decade
controller provides up to 64MB of cache and hardware XOR performance
for RAID 5 operation.
NOTE
Century or Decade controllers with an RA4050 module must
have at least one memory module installed on the RA4050.
To install the RA4050 module:
1.
Install the SIMMs in the SIMM sockets of the module. Refer to the
section, “Installing SIMMs” in this chapter.
2.
Attach the RA4050 module to the controller as shown in the
diagram below. The plastic docking guides should lock into place.
To remove the module, squeeze the plastic docking guides together
and separate the module from the controller.
3.
To confirm that the RA4050 and SIMMs have been properly
installed, start SMOR and select the controller. The Information
window should indicate the RA4050 is attached and display the
amount of cache memory.
Adapter Configuration and Installation
3-9
Installing an SX405x Bus Expansion Module
SmartRAID V controllers accept the following Bus Expansion Modules:
Model
Description
SX4054U2-1
Single Ultra2 SCSI Bus, 32-bit Expansion
Module. Adds one SCSI bus with one internal
and one external connector.
SX4054U2-2
Dual Ultra2 SCSI Bus, 32-bit Expansion Module.
Adds two SCSI buses with two internal and two
external connectors.
SX4055U2-1
Single Ultra2 SCSI Bus, 64-bit Expansion
Module. Adds one SCSI bus with one internal
and one external connector.
SX4055U2-2
Dual Ultra2 SCSI Bus, 64-bit Expansion Module.
Adds two SCSI buses with two internal and two
external connectors.
SX4055F
Single Fibre Channel-Arbitrated Loop, 64-bit
Expansion Module. Adds one external Fibre
Channel port (HSSDC).
SX4054 Bus Expansion Modules are intended for use with Decade or
Century controllers. For a 64-bit Millennium controller, use an SX4055x
Bus Expansion Module.
Up to 15 devices may be attached to each SCSI bus. Up to 125 devices can
be attached to the SX4055F Fibre Channel Bus Expansion Module.
Additional information about Ultra2 SCSI and Fibre Channel can be found
in Chapter 7, “Theory of Operations.”
NOTE
If the SX4054/SX4055 has been properly installed, it will be
listed by SMOR when the controller is selected. See Chapter 4,
“Storage Manger on ROM” for additional information.
3-10
Adapter Configuration and Installation
To install an SX4054/SX4055 module:
1.
Remove the module mounting bracket from the end of the module.
Keep the bracket screws.
2.
Attach the module to the controller, connecting P5 on the SX405x to
J7 on the controller. Refer to Appendix A, “Assembly Drawings” for
connector locations.
NOTE
SX4055 modules have two connectors (J7 and J8) that must
align correctly when installing the expansion module on a
Millennium 64-bit controller.
3.
Install the new module mounting bracket (with additional holes for
the added external connectors) and secure with the screws from
Step 1 above.
4.
Install the card in an available PCI slot and attach the mounting
bracket to the slot opening in the host system.
Adapter Configuration and Installation
3-11
Installing SIMMs
SmartRAID V PM3754U2 and PM3755F Millennium controllers support up
to 256MB of cache using four onboard SIMM sockets. One to four memory
modules of the same type and capacity can be installed.
SmartRAID V Decade and Century controllers can use up to 64MB of
cache provided by the SIMM sockets on an RA4050 module. One to four
SIMMs of the same type and capacity can be installed.
IMPORTANT
All installed SIMMs must be the same type and capacity.
Do not mix SM4050 ECC SIMMs and non-DPT SIMMs.
Do not use DPT SmartCorrect SM4000 SIMMs.
NOTE
PM1554U2 controllers have 4MB of onboard cache.
RA4050 maximum cache memory can be a single 64MB SIMM,
or four 16MB SIMMs. You can use less than 64MB of memory;
however, you must install at least one 16MB SIMM.
ECC protected cache requires DPT SM4050 SIMMs. The ECC
Enabled LED on the controller is lit when DPT SIMMs are
installed.
3-12
Adapter Configuration and Installation
To install SIMM modules:
1.
Install the SIMMs in the sockets as shown starting with socket 1.
Refer to Appendix A, “Assembly Drawings” for socket locations. Be
sure the SIMM is fully seated and locked in the socket.
Although standard 32- or 36-bit 16MB or 64MB low-profile singlesided 60ns EDO SIMMs can be used, DPT recommends that you
use DPT SM4050-16 or -64 ECC SIMMs for maximum data integrity.
2.
To confirm that the SIMMs have been properly installed, use SMOR
to view the amount of cache memory reported when the
SmartRAID V controller is selected.
Adapter Configuration and Installation
3-13
Installing DIMMs
IMPORTANT
All installed DIMMs must be the same type and capacity.
At least one DIMM must be installed.
Use only DPT DM4050 or DM4060 DIMMs.
Do not use DPT SmartCorrect DM4000 DIMMs.
The SmartRAID V Millennium model PM3755U2B has four onboard
DIMM sockets for cache memory. The green ECC Enabled LED on the
controller is lit when ECC-capable DIMMs are installed.
To install DIMM modules:
1.
Install the DIMMs in the sockets as shown starting with socket 1.
Refer to Appendix A, “Assembly Drawings” for socket locations.
2.
To confirm that the DIMMs are properly installed, start SMOR and
select the SmartRAID V controller to display the amount of cache
memory reported.
3-14
Adapter Configuration and Installation
Installing the BB4050 Battery Backup Module
The BB4050 Battery Backup Module is used only on a PM3755U2B
Millennium controller.
To install the BB4050 Module:
1.
Before installing the BB4050 module, be sure the battery pack
electrical connector is plugged in the socket on the BB4050 board.
2.
Align the Battery Backup Module and the Millennium controller as
shown in the illustration.
3.
Install the three fasteners through the appropriate holes in the
Millennium controller and into the matching posts in the Battery
Module. Tighten the fasteners.
IMPORTANT
Do not overtighten the fasteners; it can damage the circuit board.
Adapter Configuration and Installation
3-15
Installing the Controller
1.
Connect the computer’s disk activity LED cable to connector P6 on
the controller. See Appendix A, “Assembly Drawings” for the
location of this connector.
Pins 1 and 3 of P6 are connected to +5V and pins 2 and 4 are
connected to GND. Ensure that the positive lead of the LED cable
(usually a red wire or marked with a red stripe) is attached to pin 1
or 3 and the negative lead (usually a black wire) is attached to pin 2
or 4.
2.
If you are using the internal SCSI cable, connect this cable to the
controller.
3.
Install the controller in an available 32- or 64-bit PCI bus slot and
secure the controller bracket to the host system cabinet with the
screw provided.
In a system with multiple controllers, the controller that has the
lowest BIOS ROM address (typically, the lowest numbered PCI slot)
will become the booting controller. Refer to the “Determining the
Booting Controller” section for more information.
4.
NOTE
Connect any external cables to the controller.
If you have disk drives attached to a SCSI controller with a
Symbios chipset, use SMOR to set the Bootable Devices option
to Disable.
3-16
Adapter Configuration and Installation
Determining the Booting Controller
The controller whose BIOS has the lowest address will take control over
the boot process and that controller will become the booting controller.
Therefore, in a system with multiple controllers, you must ensure that the
DPT I2O BIOS occupies the lowest address if you want the DPT controller
to be the booting controller.
• In systems with multiple DPT SmartRAID V controllers, the DPT
controller in the lowest PCI slot number will be assigned the lowest
BIOS address, and will be the booting controller.
• In systems with both SmartRAID IV and SmartRAID V controllers
BIOS addresses are assigned in a system dependent manner. In this
case, the booting controller can be determined by observing the
system messages during the boot process.
If your system contains multiple DPT controllers, you can confirm which
controller has been chosen as the booting controller by running SMOR.
This utility will display a list of all DPT controllers, starting with the
booting controller.
Controller IRQ and Address
During the host system boot process, the host system BIOS should
automatically configure the DPT I 2O BIOS interrupt level (IRQ) and
memory location for all DPT PCI controllers in the system. If problems
occur, refer to Appendix B, “Troubleshooting” for additional help.
Adapter Configuration and Installation
3-17
NVRAM Reset
SmartRAID V controllers retain their setup parameters even when
powered off. These parameters are stored on the controller in an area of
nonvolatile memory (NVRAM). There is a possibility that, through
improper configuration, the controller can be put into a state where it
hangs the system during boot. If this happens, the parameters stored in
the NVRAM can be restored to their default settings by the following
procedure:
1.
Turn off power to the system.
2.
Place a shorting jumper across pins 1 and 2 of P4 on the controller.
Refer to Appendix A, “Assembly Drawings for the location of P4 on
your controller.
3.
Power on the system and wait until LEDs 3, 5, 7 and 8 on the
controller begin flashing.
4.
Turn off power to the system and remove the jumper.
5.
Restart the host system. If the system restarts normally, the
controller can now be configured using SMOR or Storage Manager.
If the system fails to boot, refer to Appendix B, “Troubleshooting” for
additional information.
3-18
Adapter Configuration and Installation
Using Microsoft Cluster Server
The information presented in this section is intended to help you configure
SmartRAID V controllers and a storage subsystem for host systems that
use Microsoft Cluster Server (MSCS) to provide fault-tolerant network
server resources. For specific information about installing and configuring
MSCS support on your Windows NT server, refer to the Microsoft Cluster
Server Administrator’s Guide.
MSCS lets you create a server cluster, that is two servers sharing a
common storage subsystem. Both servers can provide resources to
clients. However, they cannot both provide the same resources. If one
server should fail, the other server can take over the resources of the
failed server. This is called failover. The MSCS software on both servers
manages the switchover if one server should experience a system failure.
The storage subsystem controller in each server must also be able to
respond to the switchover and transfer the responsibility for I/O requests
from the failed server to the remaining server.
Clients do not see the individual servers in a cluster. They are presented
with a virtual server that contains the specific resources they require.
MSCS provides the following advantages for network servers:
• Availability
MSCS automatically detects the failure of an application or server and
restarts its resources on the other server. Users will experience only a
brief interruption in service.
In some cases it might be necessary for users to log on again when the
standby server assumes control. Refer to the Microsoft Cluster Server
Administrator’s Guide for additional information.
• Manageability
MSCS lets administrators view the status of cluster resources and
move the server workload onto different servers within the cluster as
necessary. This allows an administrator to perform manual load
balancing or rolling updates on the servers without taking data
storage or applications offline.
Adapter Configuration and Installation
3-19
System Requirements
DPT SmartRAID V controllers include support for MSCS host
configurations. To use Microsoft Cluster Server with a SmartRAID V
controller you need to do the following:
• Use host systems that conform to the Microsoft Hardware
Compatibility List for MSCS.
• Install Windows NT Advanced Server Enterprise Edition 4.0 (with
Service Pack 5 or later) or Windows 2000 Advanced Server on both
server systems.
Windows NT Workstation 4.0 does not support MSCS.
• Install and configure Microsoft Cluster Server software on both host
systems.
• Connect the servers using a TCP/IP network. This connection is used
by MSCS software to monitor the server operation and detect failures.
A static IP address is required for each server in the cluster. You
cannot use DHCP address assignment with MSCS unless you use the
DHCP server to assign the static IP addresses for MSCS resources.
You can use WINS or DNS to provide computer name to IP address
conversion. Refer to the Microsoft Cluster Server Administrator’s
Guide for additional information.
• Create a server cluster with two host systems that have SmartRAID V
controllers that are connected to a common storage subsystem. The
shared storage must use the NTFS file system.
NOTE
Only one server can control the storage subsystem under
MSCS. Any changes to the storage subsystem must be made
from the console for the server that currently owns the
subsystem.
Do not install the server operating system on the storage that
uses the shared SCSI bus.
• Enable Microsoft Cluster Server support on both SmartRAID V
controllers. See Chapter 4, “Storage Manager on ROM” for
information about activating SmartRAID V support for MSCS.
NOTE
Because the storage subsystem and controllers are on a
common SCSI bus, both controllers must have a unique SCSI
ID. To ensure they receive the highest priority for SCSI bus
access, use SCSI ID 7 (the default) for one controller and SCSI
ID 6 for the other controller. You can use SMOR to change the
SCSI ID assigned to a controller.
3-20
Adapter Configuration and Installation
How SmartRAID V Works with MSCS
When MSCS is operating on a server cluster, the servers continuously
send a heartbeat to each other across the private network connection.
When a server does not receive this signal for a preset interval, it
automatically begins the failover process of providing the resources of the
failed server and takes control of the storage subsystem.
When the transfer is complete, the failed server remains offline until the
problem that caused the failure is corrected. When the failed server is
restored, MSCS can initiate a failback operation so that the second server
releases its control of the other server’s resources and the previously
failed server begins servicing client requests for its assigned resources.
A logical array that is deleted can continue to appear as available in a
Storage Manager remote session from the other server in the cluster or
from a remote client. The cluster servers must be restarted to ensure that
the deleted array does not continue to be reported as available.
NOTE
Both SmartRAID V controllers must be connected to the shared
storage subsystem using special cables that include passthrough termination. These cables can be purchased from DPT.
Public Network
Private Network
Shared SCSI bus*
Server 1
Server 2
*Cables must use pass-through termination.
Shared Storage
CHAPTER 4:
Storage Manager on ROM
Storage Manager on ROM (SMOR) is a
BIOS-based setup utility that lets you
configure your DPT SmartRAID V controller
without starting an operating system and
using Storage Manager. You can also use
SMOR to perform basic array configuration.
SMOR makes the initial setup of your
SmartRAID V controller and RAID storage
easier and faster.
Run SMOR
Run SMOR by pressing Ctrl+D
during system boot.
Configure your Controller
Use SMOR to set the hardware
options on your controller. You can
view the configuration of the
storage subsystems, low-level
format disk drives, and configure
the DPT I2O BIOS on your
SmartRAID V controller.
Create Disk Arrays
Use SMOR to create disk arrays;
install your operating system and
your SmartRAID V system is ready
to use.
4-1
SMOR
4-2
Overview
To configure your hardware and create disk arrays when Storage
Manager is not available, run Storage Manager on ROM (SMOR) during
the system boot. This is especially useful for a new system where you need
to create disk arrays before you install the operating system.
After your storage subsystem is configured, install your operating system
according to the appropriate procedure in Chapter 5, “Software
Installation”. For access to all of the advanced features of your controller,
install the version of Storage Manager specific to your operating system.
Follow these steps to run SMOR and configure a new system:
1.
Press Ctrl+D when the DPT BIOS message appears during boot to
start SMOR.
2.
Inspect the hardware configuration as shown by SMOR.
a. Verify that all peripheral devices and controllers are shown. If any
devices are missing from the display, exit SMOR and check your
hardware connections.
b. View the Information window for each controller to verify that all
installed expansion and memory modules are shown. Note the
IRQ and Address displayed in the Configuration window. The IRQ
and Address values may be required during installation of your
operating system.
3.
Create disk arrays (see Array Operations in this chapter). Array
Groups can be created or modified at any time after system
installation. However, if the boot device will be an array, that array
must be created before the operating system is installed.
4.
Exit SMOR when you are finished with the configuration tasks.
Arrays that were created or modified start building at this time. For
large arrays, this process may take several hours. You can perform
other activities on the system while the build operation continues.
The array groups being built can be accessed while the build is in
progress. However, access time will be slower until the build is complete.
If you have exited SMOR and you want to monitor the progress of the build
operation, you can view the Array Group Information window for the array
in Storage Manger. See Chapter 6, “Storage Manager” for additional
information.
SMOR
Keyboard Reference
The navigation keys work in a similar fashion to those of Windows
Explorer. Use these keys to move around in SMOR:
Menu Control
Alt + (Menu Highlight)
Select the corresponding menu or menu
item.
Enter
Initiate an action.
Up/Down Arrows
Move between menu choices.
Esc
Return to Tree View.
Left Pane - Tree View Control
Up/Down Arrows
Move between elements within the tree.
Left/Right Arrows
Scroll the tree left and right.
+
Expand the current branch, showing the
devices attached to it. The element must
be preceded by +.
–
Collapse the current branch. The element
must be preceded by –.
Tab
Move to the right pane — Information View.
Right Pane – Information View Control
Alt + (tab highlight)
Select and go to the corresponding Tab
Page within the Information View.
Tab
Move to the next field in the window.
Shift-Tab
Move to the previous field in the window.
Space
Select or deselect an item (check-boxes or radio
buttons).
Up/Down Arrows
Change a combo-box (indicated by ) value.
Esc
Return to Tree View.
4-3
SMOR
4-4
Menu Reference
The SMOR Main Menu changes dynamically depending on what is
selected in the tree view. The following is a complete list of possible
selections:
FILE
Read System Config
Set System Config
Exit
Rescan the system – unsaved configuration
changes are lost.
Save and enable configuration changes.
Quit SMOR.
RAID
Create…
Delete
Rebuild
Stop Build
Create a new array.
Delete the currently selected array.
Rebuild a RAID 1, 5, 0+1 or 0+5 array.
Stop building or rebuilding an array.
ACTION
Make Hotspare
Remove Hotspare
Format Drive
Flash HBA
Save Firmware
Designate the currently selected drive as a
Hot Spare.
Make the currently selected Hot Spare
available for use by the operating system.
Low-level format a currently selected drive.
Update the firmware, I2O BIOS or SMOR image in
controller ROM.
Save the controller BIOS, firmware and SMOR
images to a bootable diskette.
HELP
About…
Help
Display information about SMOR.
Display basic help information.
Icon Reference
SMOR uses the following icons in Tree View:
Controller
Disk
CD-ROM
Array
Tape
Other Device
SMOR
4-5
Screen Layout
The SMOR interface works like the Windows Explorer tree-structured
interface. The screen is divided into three major components: a menu bar
across the top of the screen and two display panes below the menu bar.
Information & Configuration Tab Pages
Menu Bar
Tree View
The Menu Bar
To open a specific menu, press Alt+(highlighted letter of the menu
name). After a menu is open, you can select a specific menu item by
pressing the key for the letter highlighted on the menu item. For example,
to select the Read System Configuration item on the File menu, press
Alt+F to open the File menu, then press R to select Read System
Configuration.
NOTE
Depending on your configuration, not all menu items may be
available. Unavailable menu items are shown in a lowcontrast color.
4-6
SMOR
The Left Pane – Tree View
The Tree View, displayed in the left pane, is the central control for SMOR.
This view displays a tree structure that represents the organization of the
DPT storage subsystem. This includes DPT controllers, storage devices
and arrays detected by SMOR. By moving the highlight with the up and
down arrows, you can select items you want to view or configure. As items
are highlighted, the associated information on the item is displayed in the
Information View (display pane on the right side).
If an item in the Tree View is preceded by +, pressing the Enter or + key
expands the tree, showing the devices associated with or attached to that
item. If an item is preceded by –, pressing the Enter or – key will collapse
that portion of the tree, hiding the devices under that item.
If the text for an item is larger than the width of the Tree View pane, you
can scroll the pane horizontally by using the left and right arrow keys.
The Right Pane – Information View
To the right of the component tree is the Information View pane. This view
displays information related to the currently selected item in the tree. The
specific information displayed in the Information View varies depending
on the item selected. When there are separate types of information
available for the selected item, the Information View is separated into Tab
Pages. Tab Pages are generally information or configuration parameters
that are related to the selected item.
To select a specific Tab Page within the Information View, press Alt+(the
highlighted letter on the tab). For example, to change to a controller’s
Configuration tab, press Alt+C. You can also press the Tab key when in the
Tree View to move over to the currently displayed Information View tab
page. On a tab page within the Information View, you can move between
the items with the Tab or Shift+Tab keys. See the Keyboard Reference
section in this chapter for additional details on changing between the Tree
and Information Views, using the menu bar, and navigating within the
SMOR interface.
SMOR
4-7
Within the Information View, you select an item to configure by using the
Tab or Shift+Tab keys to move the highlight to the item. Items that cannot
be selected are shown in black. The way in which you change an item
depends on the type of control associated with it. Check-boxes are toggled
by using the Spacebar. List-box items (for example, SCSI Transfer Rate)
are changed using the up and down arrow keys. List-box items can be
recognized by the downward pointing arrow at the right of the item.
To leave the Information pane and return to the Tree View, press Esc. If
you have changed the configuration, you are prompted: Save changes?.
Press Tab to select Yes or No and press Enter.
NOTE
The items shown in the Information View, and the settings for
those items, will vary depending on the particular type of
controller, device or array selected in the Tree View.
SMOR
4-8
Running SMOR
Start SMOR by pressing Ctrl+D when the DPT I 2O BIOS message appears
on the screen during the boot sequence.
The letters that appear during the initial device scan process are:
C
CD-ROM
D
Disk drive
E
Scanner, SAF-TE or intelligent RAID enclosure, or a second
controller (MSCS).
H
Hot spare drives
T
Tape device
0, 1, 5 Physical arrays identified by the RAID level
The position of a letter corresponds to the SCSI ID assigned to that device.
SMOR
4-9
Information/Configuration Windows
When you highlight an item within the Tree View, the corresponding
Information View is displayed.
DPT I2O BIOS Settings
The example above is the default Information View when SMOR starts.
The settings in this view affect the DPT I2O BIOS and all the DPT
controllers in your system. To view or change these settings, highlight
DPT Configuration in the tree.
Controller Parameter
Default
Optional Settings
Bootable Devices
Normal
Disable
Enable Bootable CD-ROMs
Disabled
Enabled
Write Through
Write Back
1 second
1, 10, 20, 30
seconds
EBDA Relocation
Disabled
Enabled
Enable Extended Int13
Enabled
Disabled
DOS/Windows 3.1 Cache Setting
Scan Delay
Smart ROM Options:
SMOR
4-10
Bootable Devices
These settings let you modify the system boot process for host systems
with multiple peripheral controllers in cases where the DPT I 2O BIOS
does not provide effective or appropriate default operation.
The default setting causes the system to follow the order described in the
Determining the Booting Controller section of Chapter 3. If you select
Disabled, DPT controllers are not used as boot devices.
If you select Enable Bootable CD-ROMs, the DPT controller attempts to
detect a bootable CD-ROM that uses the El-Torito format.
NOTE
This option is disabled by default, because some bootable
CD-ROMs contain device-specific boot code that will not work
with DPT controllers.
DOS/Windows 3.1 Cache Setting
This parameter determines how the controller responds to Int13 write
commands under DOS, Windows 3.1 and certain operating system
installation programs. The default is Write Through to avoid problems that
can occur during operating system installation if Write Back caching is
enabled. After the operating system is installed, you can change to Write
Back caching for improved performance.
Change this setting back to Write Through during future operating system
installs or upgrades to avoid problems.
NOTE
The cache setting has no effect on controller cache operation
under Windows NT, UNIX or NetWare.
Scan Delay
Some SCSI devices require a time interval between power on and SCSI
bus reset and scan or they do not respond correctly. If devices are not
displayed in the Tree View after power on, set the delay to a longer
interval.
EBDA Relocation
This setting determines the way that SmartRAID V controllers handle
Extended BIOS Data Area (EBDA) relocation. You can enable this feature
to help avoid conflicts with other adapter cards if the controller is installed
in a host system with other adapters that follow standard EBDA relocation
rules.
Enable Extended Int13
This setting enables extended logical block addressing (LBA) for disk
devices. Logical block addressing enables operating system access to
disks larger than 8.6GB. You should not change this setting.
SMOR
4-11
Controller Information Windows
To view or change the configuration of a DPT RAID Controller, highlight
the controller in the Tree View. Two tabs are available: Information and
Configuration.
Controller Information Tab
The Information Tab for a Controller displays general information
reported by that controller. Some of the fields have special conditions:
Model
Serial #
Cache
DPT controller model number
Controller serial number
Amount of cache memory installed
Revision
FW Type
ECC
Controller firmware version
Firmware type
Yes (if DPT ECC memory is used)
SCSI Bus:
Width
Bus width (8-bits or 16-bits)
Busses
Number of busses on the controller
SCSI ID
SCSI ID assigned to the controller
Transfer
Maximum possible bus transfer rate
SMOR
4-12
Host Bus:
Bus Type
Always PCI for SmartRAIDV
Transfer
Host PCI bus transfer rate.
132 MB/sec for 32-bit PCI bus
264 MB/sec for 64-bit PCI bus
Attached Modules:
Identifies the expansion modules and
cache memory modules installed.
Attached Modules can be reported as:
SX405xU2-1, SX405xU2-2
SX405xF
RA4050
BB4050
SCSI Expansion Module*
Fibre Expansion Module*
RAID Accelerator
Battery Backup Module
*Bus Expansion Modules are always reported as either SX405xU2 or
SX405xF. SMOR is not able to determine which specific model is
attached.
Memory modules are reported as:
Bank n: nnMB ECC
Starting with Bank 0. ECC is displayed only when ECC memory is
used.
SMOR
4-13
HBA Configuration Tab
To modify the hardware parameters for the highlighted SmartRAID
controller, press Alt+C to display the Configuration tab.
Select Default to reset the parameters on this tab to their default values.
NOTE
Depending on the specific hardware, some items displayed in
the Configuration tab may be read-only.
The settings are subdivided according to which parameters relate to the
individual peripheral bus and which parameters relate to the entire
controller.
Controller Parameter
Default
Optional Settings
N/A
0, 1, 2
7
0 – 6 (SCSI)
0 – 126 (Fibre)*
TERMPWR
Enabled
Disabled
Termination:
Auto
On, Off, High Only
SCSI Xfer Rate (in MHz)
Auto
80, 40, 20, 10, 8, 5,
Asynchronous
Bus
ID
* Fibre Channel IDs are display only. They cannot be changed.
Bus
Each peripheral bus on a controller is assigned a number starting at 0.
SMOR
4-14
SCSI ID
SmartRAID V controllers are configured by default at SCSI ID 7. This
value should not be changed unless required for special configurations.
NOTE
Fibre Channel IDs can be 0 – 126. However, these IDs are
configured dynamically and cannot be changed using SMOR
or Storage Manager.
TERMPWR
By default SmartRAID V controllers supply termination power for other
SCSI devices through the TERMPWR line on the SCSI cable. If you want
to change this setting, refer to the TERMPWR section in Chapter 7,
“Theory of Operation” for more information.
Termination
This setting controls the SCSI termination for the controller and bus. The
default value (Auto) should not be changed, unless both internal and
external cables are attached to the controller or you are using an 8-bit
(Narrow) cable. Refer to the Configuring SCSI Termination section in
Chapter 3, “Configuration and Installation” for information on setting this
parameter.
NOTE
Termination is not used for Fibre Channel controllers.
SCSI Xfer Rate
DPT controllers automatically negotiate with each SCSI device at powerup or reset to set the maximum SCSI transfer rate. This parameter limits
the SCSI transfer rate to the value selected. This setting should not be
changed except when you are troubleshooting SCSI bus errors.
NOTE
If setting this parameter to 5MHz eliminates SCSI bus data
errors, this is usually an indication that the SCSI bus is too
long or that the bus is not terminated correctly.
PCI Xfer Rate
Mem Address
IRQ
These are display only fields for information purposes. The values in the
Mem Address and IRQ fields may be needed when you configure your
operating system.
Cluster Server Enable
This option enables or disables support for Microsoft Cluster Server
configurations using the selected controller. Do not enable this option
unless you are creating a cluster server configuration.
SMOR
4-15
Device Information
Individual devices are listed in the Tree View under the controller to which
they are connected. Highlight a device to view its configuration
information.
The Device Information view displays general device information and
hardware configuration. This view is divided into three parts:
Identification, Capabilities and Status. Progress shows a numeric
percentage of completion for active tasks.
Identification
The Identification section displays a general description of the highlighted
device:
Description
The manufacturer name and model number as reported by
the device, followed by the icon for the device.
Revision
The device firmware revision.
Address
The address of the device, in the form HBA n, Channel n,
ID n, LUN n.
Capacity
The capacity of the device in megabytes. For removable
media disk devices, the capacity reported is for the
currently inserted media, or No Media Inserted if no media
is inserted. Tape drives do not report media insertions.
Removable
Read Only
As reported by the device
Block Size
Block size reported by the device.
Negotiated
Bus speed negotiated between the device and controller.
Transfer Rate
Maximum transfer rate for the negotiated bus speed and
data path.
SMOR
4-16
SCSI Capabilities
The capabilities section is a list of the capabilities defined in the SCSI
specifications. A checkmark indicates that the device supports that SCSI
function.
Status
The following status conditions can be indicated for attached devices:
Dead
The device failed to respond to controller
commands. If the device becomes available it
will only change status after the system
configuration is read or the host is restarted.
Failed
A drive failure has occurred.
Impacted
An operation is being performed that results
in performance degradation in response to
server I/O requests.
Missing
The drive is physically missing or will not
respond to commands on the device bus.
Optimal
The device is fully functional.
Uninitialized
The drive is operational but has not been
initialized as a member of an array.
Verify
A Verify operation is being performed on the
array. I/O performance will be affected.
Warning
This status indicates that failure may be
imminent on a device with S.M.A.R.T. failure
prediction.
SMOR
4-17
Array and Array Group Information
SmartRAID V controllers implement RAID 0, 1 or 5 disk arrays in
hardware. RAID 0 arrays are composed of any combination of individual
drives. RAID 1 arrays are always composed of two drives. RAID 5 arrays
contain three or more drives.
After the arrays have been created, one or more arrays of the same RAID
level can be combined into a Parity Group (see the Creating a Parity
Group section). Arrays are striped into Parity Groups by the controller
firmware. All the drives in an array or Parity Group must be attached to
the same controller, and appear to the host as a single Logical Storage
Unit (LSU).
The Information window for any array may be viewed by highlighting that
array.
NOTE
Arrays do not start building until a Set System Configuration
action has been performed.
The Array Information window displays general array information and
hardware configuration. It is divided into three parts: identification, SCSI
Capabilities and Status.
SMOR
4-18
Identification
The identification section displays a general description of the highlighted
array:
Description
The RAID level used for the array.
Revision
The firmware revision of the controller to which the
array is attached.
Address
The address of the array, in the form HBA n, Channel
n, ID n, LUN n. Arrays are always assigned the
lowest logical address of any device in the array. This
field displays as much information as is necessary to
unambiguously define the address of the array.
Capacity
The usable capacity of the array in megabytes. The
available capacity of an array depends on the RAID
level of that array.
Removable
As reported by the devices in the array.
Block Size
Displays the sector (block) size for the selected device
in bytes. For disk drives, this value should be 512. If
the value is not 512, use SMOR to perform a low-level
format and create 512 byte sectors.
Stripe Size
Displays the stripe size used to create the array.
SCSI Capabilities
This section displays the SCSI capabilities of the array, as reported to the
operating system. The capabilities reported depend on the devices that
were used to create the array.
Status
This field displays the current status of the array. A progress indicator (a
numeric percentage of completion) can also appear if the array is Building
or Rebuilding. See the following table for array status definitions.
SMOR
Building
The array is being built.
Created
The array or device is defined but not
initialized.
Dead
A write-back cache to the array command
failed. This is an unrecoverable failure.
Degraded
A single drive in the array has failed; array
performance is affected.
Impacted
A verification is being performed on the array;
I/O performance is affected.
Missing
A drive is physically missing or will not
respond to commands on the device bus.
Optimal
The array is fully functional.
Pending
The array has been created and the build is
queued on the controller, but is not yet
started.
Rebuilding
Data is being rebuilt onto a drive in the array.
Warning
This status indicates that a drive in a fault
tolerant array has failed and the next drive
failure will cause loss of the array.
4-19
Setting the Configuration
There are two configuration options on the File menu:
Read System Config
Causes SMOR to rescan to detect any changes in
hardware configuration or status. Any changes
that have been made and not saved are lost. This
operation is run automatically when SMOR is
started.
Set System Config
Causes SMOR to save changes that have been
made to the storage subsystem configuration in
the controller memory. If any array groups or
parity groups have been created or modified, this
operation causes the controller to initiate a build
operation on the new groups.
SMOR
4-20
Array Operations
You can use SMOR to create or manage disk arrays. The following
sections describe how to use SMOR to create arrays and parity groups,
delete arrays, assign hot spare drives, and rebuild an array.
Refer to the Chapter 7, “Theory of Operation” for a complete discussion of
RAID levels and disk arrays.
Creating an Array
To create an array:
1.
Select RAID–Create.
2.
When the RAID Type dialog appears, select the RAID level you
want to use. The Stripe size is selected automatically; however you
can select a different stripe size value by highlighting the field and
using the up and down arrow keys to change the stripe size.
NOTE
Although you can change the stripe size, DPT recommends
using the default value. This value has been selected for
optimum performance based on the type of disk array you
want to create.
When you are ready to proceed, select OK.
The Eligible Devices list will appear. Devices on this screen can be
either individual disk drives or previously created array groups.
Array groups appear in the list when you select RAID 0 and eligible
array groups exist.
3.
Select the devices you want to include in the array:
a. To add devices to the array, highlight the device and press the
Spacebar. A checkmark appears next to the device to indicate that
it has been selected.
You might need to scroll the display down to view all eligible
devices.
SMOR
4-21
b. To remove a previously selected device from the array, highlight
the device and press the Spacebar.
4.
When you are finished selecting drives for the new array, select
Done.
5.
If you are creating a RAID 1 array, the RAID-1 Build Option dialog
appears.
RAID 1 arrays are built by copying the existing data from one
device to the other. Select the direction for the copy, then select OK.
6.
Select File–Set System Config to start the build process. The array
you have created will begin building at this time. If you have created
multiple arrays, they are built serially in the order they were
defined.
Alternatively, you can exit SMOR. You will be prompted to save the
configuration changes. If you choose to save the configuration and
have defined arrays, the build process will begin.
For large redundant arrays, the build process can take several hours to
complete. You can exit SMOR and perform other activities on the system
while the build continues. An array being built can be accessed during the
build process.
If you exit SMOR and you want to monitor the progress of the build
operation, you can use the Array Group Information window for the array
in Storage Manger. See Chapter 6, “Storage Manager” for additional
information.
SMOR
4-22
Creating a Parity Group
Creating a parity group (RAID 0+1 or 0+5) is similar to creating a normal
RAID 1 or RAID 5 array group. Do the following to create a RAID 0+1 or
RAID 0+5 parity group:
1.
Create and build your array groups as described in the preceding
section (Creating an Array).
Do not initiate the build process on any arrays that you intend to
use in a parity group.
2.
After your initial array groups are created, select RAID–Create
again.
3.
Select RAID 0 for the RAID type and click on OK.
4.
Select two or more arrays of the same type from the Eligible Devices
list and click on Done.
NOTE
You cannot combine arrays that use different RAID levels.
5.
Select File–Set System Config to begin the build process for the
parity group.
The Tree View will display the parity group LSU as:
(x,x,x,x) FW RAID-0
with the array groups that were striped together listed where disk drives
would normally be listed. Selecting an array group component will branch
to the disk drives for that array group. The LSU address will be the lowest
address of the array logical addresses that make up the parity group.
SMOR
4-23
Deleting an Array
To delete an array, follow the steps below:
1.
In the left pane, highlight the array that you want to remove, then
select RAID–Delete.
2.
You will be warned that all data will be lost. Choose Yes or No.
The array configuration for the devices is not deleted until you
select File–Set System Configuration or exit SMOR and choose to
save your changes.
Hot Spares
Any disk drive not assigned to an array or in use by the operating system
can be designated as a Hot Spare. Hot Spares automatically replace failed
drives in RAID 1 or RAID 5 arrays and are not accessible by the operating
system for other use. A Hot Spare can protect drives of equal or lesser
capacity attached to any peripheral bus on the same controller.
To assign a drive as a Hot Spare:
1.
Highlight the drive you want to use in the left pane.
2.
Select Action–Make Hotspare.
Selecting Action–Remove Hotspare will cause the selected Hot Spare drive
to be reassigned as a normal disk drive which is accessible by the
operating system.
Refer to Hot Spares in Chapter 6, “Storage Manager” for additional
information.
SMOR
4-24
Rebuilding a Failed Array
If a drive in a RAID 1 or RAID 5 array fails and the drive is not protected
by a Hot Spare, use the following procedure to replace the failed drive.
1.
Remove and replace the failed drive according the procedures in
your hardware documentation.
2.
When the failed drive has been replaced, choose RAID–Rebuild
Array to start the rebuild process.
The status of the array will change to Rebuilding (view the Information
window for that array). When the rebuild is complete, the array status will
change to Optimal.
NOTE
A DPT RAIDstation enclosure will automatically detect the
replacement of a failed drive and initiate a rebuild operation
as soon as the new drive is online.
Formatting a Drive
Formatting SCSI hard drives is not normally required. However, if you
have a drive that was previously formatted with a sector size other than
512 bytes you must change the format to 512-byte sectors before you can
use the drive with a SmartRAID V controller. You can use SMOR to start a
low-level format operation and create a 512 byte/sector format.
IMPORTANT
Do not remove power from the drive until the format operation is completed.
Doing so can cause some drives to be left in an indeterminate state that will
require return to the manufacturer for repair or replacement.
Low-level formatting large capacity drives can take considerable time. Do
the following to start a low-level format for a disk drive:
1.
Highlight the drive to be formatted.
2.
Select Action–Format Drive.
3.
Select OK and confirm.
To determine if the format has completed, view the Information
window for that drive.
SMOR
4-25
Managing Controller Firmware
SMOR includes several features that let you upgrade, save and restore the
firmware on your SmartRAID V controller. These options are listed on the
Action menu when a controller is selected in the Tree View.
The Flash HBA feature lets you install the latest DPT controller firmware,
I2O BIOS and SMOR utility software. The Save Firmware feature lets you
save the current controller firmware, I2O BIOS and SMOR software along
with the NVRAM settings to a floppy disk. This disk can be used to restore
the controller or copy the firmware, I2O BIOS, SMOR and NVRAM to
another controller.
Upgrading Firmware
The SmartRAID V controller firmware, DPT I2O BIOS and SMOR utility
are contained in flash ROM and can be upgraded by using SMOR. Each
component must be upgraded as a separate operation.
• Firmware image upgrades are contained in a xxxxxxxx.FWI file,
where the 8-character file name consists of the 4-digit controller model
and a 4-digit release number.
• DPT I 2O BIOS images are contained in a file named I2OBIOS.xxx
(where, xxx is the version number).
• SMOR updates are contained in a file named SMORyyyy.xxx. Where,
yyyy is the build number and the file extension (xxx) is the version
number.
• The default NVRAM settings are contained in the file DPTV1_0.NVR.
DPT periodically releases updated firmware, I2O BIOS and SMOR
software. You can obtain the latest files by contacting DPT Technical
Support or from our FTP site at ftp.dpt.com.
NOTE
It is much easier to perform the following procedure if the
image files are placed in the root directory of a diskette.
SMOR
4-26
Do the following to upgrade controller NVRAM, firmware, I2O BIOS, or
SMOR utility:
1.
Place the diskette containing the image files in the diskette drive of
the system where the controller is installed.
2.
In the Tree View, select the controller that you want to update.
3.
Select Action–Flash HBA. The source file browser will appear.
4.
Select the diskette drive in the source file browser window and
press Enter to expand the drive listing.
5.
Highlight the image file you want to use and select OK to update the
controller flash ROM.
After the controller reads the image, it will display the version
number of the component you selected.
Select Yes to confirm.
6.
A progress indicator will appear showing the progress of the
different stages of the flash operation.
If the operation does not complete successfully, refer to the procedures for
recovering from an incomplete or failed flash ROM upgrade in Appendix
B, “Troubleshooting”.
SMOR
4-27
Saving Controller Firmware to a Disk Image
This feature provides you with a way to save the current ROM software
and NVRAM settings for your controller to a bootable diskette. The
resulting diskette is a bootable disk and contains a copy of the three
controller software components: firmware, I2O BIOS and SMOR. The
utility also saves the current and default NVRAM settings.
The diskette will contain the following files:
SMORbbbb.vvv
Where bbbb is the build number and vvv is the version.
I2OBIOS.vvv
Where vvv is the version.
ffffvvbb.FWI
Where ffff is the adapter firmware type (ffff can be "3754" or "2554"). vv
is the version and bb is the build.
CURRENT.NVR
The current NVRAM settings for this controller.
DEFAULT.NVR
or
vvvvvvvv.NVR
Default NVRAM settings. vvvvvvvv is the specific version of the default
NVRAM settings if previously named.
SMORSbbb.vvv
This file is used to guarantee that the diskette will boot if there is no
current SMOR image available on the controller. (Where bbb is the
build number and vvv is the version of the in-memory copy of SMOR.)
The build and version may or may not be identical to the SMOR image
copied from the flash memory of the controller.
SMOR
4-28
Creating a Firmware Disk Image
To create the firmware backup diskette:
1.
Press Ctrl+D when the DPT I2O BIOS banner appears during
system startup to start SMOR.
2.
Move the highlight to the controller that you want to backup.
3.
Insert a blank, formatted diskette into the host diskette drive.
NOTE
This operation overwrites the entire diskette. Any existing
data is erased.
4.
Press Alt+A to select the Action drop-down menu.
5.
Select Save Firmware. Press Tab to highlight Yes when prompted.
Press Enter to begin the flash update.
The operation starts immediately. SMOR displays a progress bar and
numerical percentage of completion. When the operation is complete you
can exit SMOR and start the system normally.
Restoring Firmware from a Disk Image
To use the firmware backup:
1.
Insert the diskette with the backup images into your diskette drive.
2.
Power on or restart your system. The system will boot from the
diskette and start SMOR automatically.
3.
Highlight the controller that you want to update.
4.
Press Alt+A to open the Action menu and select Flash HBA.
5.
Highlight the diskette drive and press Enter. SMOR will display a
list of valid files on the diskette.
6.
Highlight the software image you want to use and press Tab to
highlight OK.
Press Enter to start the flash update operation. SMOR will read the
file and verify that it is a valid flash image.
7.
Press Tab to highlight Yes and press Enter to begin the flash update.
8.
Press Enter when you see the confirmation prompt
9.
When the flash update is complete you must restart the system to
enable the new software.
CHAPTER 5:
Software Installation
Your new SmartRAID V controller uses I 2O
drivers that may be supplied by DPT or your
operating system vendor. This chapter
discusses driver and operating system
installation when using SmartRAID V
controllers.
After you install the operating system, you
can install Storage Manager. Storage
Manager is available for:
Windows NT 4.0
A DPT I2O driver is supplied with
your controller.
Windows 95/98
A DPT I2O driver is supplied with
your controller.
SCO UNIX 3.2V4.2, ODT 3.0,
and OpenServer 5
Microsoft Windows 95/98
SmartRAID V controllers include a
SCO BTLD diskette with a DPT I2O
driver.
SCO UNIX 3.2v4.2
SCO UnixWare 7
SCO OpenServer 5
An I2O driver for UnixWare 7 is
available from SCO.
Microsoft Windows NT 4.0
SCO UnixWare 7
Red Hat Linux 6.x
NOTE
If a device driver for your operating
system is not supplied on the DPT
diskettes or is not available from the
DPT web site, contact the
manufacturer of your operating system
to obtain an I2O OSM for your system.
NetWare 4.11, 4.2 and 5.0
A DPT-qualified I2O driver from
Novell (for NetWare 4.2 and 5.0) is
supplied with your controller. The
driver for NetWare 4.11 is available
for download from the DPT web
site.
Red Hat Linux 5.2, 6.0 and 6.1
Diskette images are available from
DPT (www.dpt.com). These drivers
support Linux kernel versions
2.0.36 and 2.2.5.
BSD/OS 4.1 and FreeBSD
Diskette images are available from
DPT (www.dpt.com).
5-1
Software Installation
5-2
Windows NT 4.0
The DPT SmartRAID V driver for Windows NT 4.0 provides support for
both disk drives and other SCSI devices, such as CD-ROMs and scanners.
NOTES
• Windows NT must be updated with Service Pack 4 or later.
• If you choose the Microsoft I2O OSM in place of the DPT class or
miniport driver, you will not be able to use Storage Manager.
• The Windows NT 4.0 OSM is not compatible with Windows 2000. If
you have a SmartRAID V controller installed under Windows 2000,
use the DPT miniport driver instead of an OSM.
• If you are installing Windows NT from an IDE CD-ROM, press F6
as soon as the NT Detect screen appears during the installation
process. This will cause the installation program to prompt for
installation of a mass storage driver earlier in the install process
and ensure that the disk drives attached to the controller are
available when required.
• The DPT class driver is installed and enabled automatically. The
class driver provides enhanced performance and reduced CPU
utilization during disk accesses. This driver will load before the
Windows NT disk class driver and acquire all disks attached to
DPT controllers. If you want to use the DPT miniport driver, do the
following after you complete the initial installation:
Open the Control Panel.
Double click the Devices icon.
Disable the entry for DPT DISK.
Restart the system.
Software Installation
5-3
There are three driver installation scenarios for Windows NT 4.0:
• Installing the SmartRAID V controller into a new system, then
installing Windows NT 4.0
• Upgrading an existing Windows NT 4.0 system
• Adding a SmartRAID V controller into an existing Windows NT 4.0
system, to be used in addition to a currently installed SmartRAID IV
or SmartCache IV controller
Each of these cases requires a slightly different procedure. Be sure you
select the correct procedure and follow the procedure carefully. If you
need additional assistance, contact DPT Technical Support.
Installing a New System
This procedure describes how to install a SmartRAID V controller and
Windows NT 4.0 on a new system. If Windows NT is to coexist with another
operating system, that operating system must be installed before you
install Windows NT. During the Windows NT installation, you will need one
blank high-density diskette to create an Emergency Repair Diskette.
To install a SmartRAID V controller and Windows NT 4.0 into a new
computer system, follow the steps below:
1.
Install the SmartRAID V controllers and storage devices according
to the instructions in Chapter 3, “Configuration and Installation”.
2.
Boot your system and run SMOR to verify your configuration.
Create your disk arrays now. Each array will appear to the
operating system as a single logical drive.
3.
Ensure that the Windows NT CD-ROM is in the CD-ROM drive.
Boot the system from the Windows NT setup diskette 1.
4.
Insert the additional setup diskettes when prompted.
5.
Press S to skip Setup’s mass storage detection. Press S to Specify
Additional Devices. Select Other, then insert the DPT diskette for
Windows NT.
6.
Select DPT I2O SCSI Adapter Driver for NT 4.0 and greater.
7.
Press Enter and continue with the installation as described in the
Windows NT documentation.
Software Installation
5-4
NOTE
If a SmartRAID IV or SmartCache IV controller is installed,
Windows NT mass storage detection routines will load an
older DPT driver for that device from the Windows NT diskette
or CD-ROM. After Windows NT installation completes, update
the driver using the version contained on the DPT diskette
that was supplied with your SmartRAID IV or SmartCache IV
controller.
Alternatively, if you have a controller from another manufacturer or other
hardware that requires the use of Windows NT mass storage detection,
follow the steps below:
1.
Install the SmartRAID V controllers and storage devices according
to the instructions in Chapter 3, “Configuration and Installation”.
2.
Boot your system and run SMOR to verify your configuration.
Create your disk arrays now. Each array will appear to the
operating system as a single logical drive.
3.
Ensure that the Windows NT CD-ROM is in the CD-ROM drive.
Boot the system from the Windows NT Setup Disk 1.
4.
Insert additional Windows NT setup diskettes when prompted.
5.
Press Enter to perform Setup’s mass storage detection.
6.
After Windows NT detects any controllers for which it has built-in
drivers, you are prompted to either press S to specify additional
controllers or press Enter to continue Windows NT installation
without specifying additional controllers.
7.
Press S to Specify Additional Devices. Select Other, then insert the
DPT Windows NT diskette.
8.
Select DPT I2O SCSI Adapter Driver for NT 4.0 and greater.
9.
Press Enter and continue with the installation according to the
instructions in the Windows NT documentation.
Software Installation
5-5
Upgrading an Existing System
NOTE
If you have already installed a SmartRAID V controller into
your system and want to change drivers, refer to the Adding to
an Existing System section instead of this procedure.
To upgrade an existing Windows NT 4.0 system to a SmartRAID V
controller, follow the steps below:
1.
From the Control Panel, select the SCSI Adapters icon. Then select
the Drivers tab and click Add.
2.
Insert the DPT Windows NT diskette, then click Have Disk.
3.
Select DPT I2O SCSI Adapter Driver for NT 4.0 and greater.
4.
After the driver is loaded, shut down the system. Replace the
existing controller with the DPT SmartRAID V controller.
5.
Boot the system.
6.
If you are removing the old DPT controller, you should also remove
the old driver.
Adding to an Existing System
To add a DPT SmartRAID V controller to an existing Windows NT 4.0
system, do the following:
1.
From Control Panel, select the SCSI Adapters icon. Then select the
Drivers tab and click Add.
2.
Insert the DPT I2O drivers diskette, then click Have Disk.
3.
Select DPT I2O SCSI Adapter Driver for NT 4.0 and greater.
4.
After the driver is loaded, shut down the system. Install the new
DPT SmartRAID V controller into the system, leaving the existing
controller installed.
5.
Boot the system.
Software Installation
5-6
Adding to an Existing System
(Microsoft OSM)
NOTE
If you use the Microsoft I 2O OSM in place of the DPT miniport
driver, you will not be able to run Storage Manager.
To add a DPT SmartRAID V controller to an existing Windows NT 4.0
system using the Microsoft Windows NT OSM, do the following:
1.
Obtain a copy of the self-extracting file for the Microsoft I 2O drivers
and extract the drivers to a directory or diskette.
2.
From Control Panel, select the SCSI Adapters icon. Then select the
Drivers tab and click Add.
3.
When the Install Driver dialog box appears, select Have Disk.
4.
Insert the Microsoft Windows NT OSM I 2O driver diskette or type
the path to the directory to which you unzipped the driver files.
5.
Select Microsoft I20 Drivers.
6.
After the driver is loaded, shut down the system. Add the new DPT
SmartRAID V controller to the system, leaving the existing
controller installed.
7.
Boot the system.
The Microsoft I 2O OSM driver supports only disk drives. If you want
CD-ROM, tape, scanner, or other SCSI device support on your
SmartRAID V controller under Windows NT, you must to use the DPT
Windows NT 4.0 miniport driver.
Additionally, the Microsoft I 2O OSM driver for Windows NT does not
inherently support bootable controllers. If you want to boot the
SmartRAID V controller with the Microsoft I 2O OSM, do the following:
1.
Install Windows NT 4.0 with the DPT Windows NT 4.0 miniport
driver.
2.
After Windows NT is installed, remove the DPT Windows NT 4.0
miniport driver from the SCSI Adapters list.
3.
Install the Microsoft Windows NT OSM I2O driver. The I 2O BIOS
ROM on the SmartRAID V controller will manage the Windows NT
4.0 boot process up to the point where the Microsoft I 2O OSM loads.
Software Installation
5-7
Windows 95/98
The DPT SmartRAID V driver for Windows 95/98 provides support for both
disk drives and other devices, such as CD-ROMs and scanners.
There are several driver installation scenarios under Windows 95/98:
• Installing the SmartRAID V controller into a new system, then
installing Windows 98
• Installing a SmartRAID V controller into a new system, then installing
Windows 95
• Upgrading an existing Windows 95 system to Windows 98
• Adding a SmartRAID V controller to an existing Windows 95/98 system
• Upgrading Windows 3.1 to Windows 95 and installing a
SmartRAID V controller
Each of these cases requires a slightly different procedure.
Installing Windows 98 (Full Install Version)
This section describes the steps to install Windows 98 (Full Installation
Version) on one disk drive or array group using a SmartRAID V host
controller. Ensure that your hard drives, arrays, CD-ROM drive, and any
other SCSI device attached to the host controller are recognized by SMOR
before proceeding.
Do the following to install Windows 98 (Full Installation Version):
1.
Boot the system from a bootable DOS diskette.
2.
Run FDISK and create a Primary DOS Partition of an appropriate
size.
3.
Mark the Primary DOS Partition Active.
4.
Reboot the system from the DOS diskette.
5.
Perform a system format (FORMAT C: /S).
6.
Run the SETUPCD.EXE program on the DPT Windows NT diskette.
This program creates an AUTOEXEC.BAT file and CONFIG.SYS file
with support for your CD-ROM drive.
7.
Ensure that the MSCDEX.EXE program is in your boot path or the
root directory of the C: drive.
Software Installation
5-8
8.
Remove the boot diskette from the floppy disk drive and reboot the
system from the C: drive.
9.
Insert the Windows 98 Full Installation CD in the CD-ROM drive.
10. Enter the drive letter assigned to the CD-ROM and run SETUP (for
example, D:\SETUP).
11. Follow the Windows 98 installation prompts.
12. During one of the device detection boot processes, Windows 98 will
identify PCI Card and/or Memory Controller devices. Click Next and
display the list of drivers from a specific location.
NOTE
If you are not prompted for a PCI Card or Memory Controller
Driver or the Device Manager shows PCI Card with a yellow
exclamation mark in the Other Devices list, you might need a
patch for your motherboard chipset to recognize multifunction
bridge controllers.
Contact your motherboard vendor and apply the patch. After
the patch is installed and the system is restarted, Windows 98
will repeat the device discovery process. If you are installing
the operating system, go back to step 12 and continue the
procedure. If you are adding a SmartRAID V to an existing
configuration, go to step 3 on page 5-12 and continue with the
procedure.
13. Select Have Disk and insert the DPT Windows 95/98 diskette when
prompted. The system will then reboot back into Windows 98.
NOTE
If more than one DPT SmartRAID V controller is installed in
the computer, repeat steps 12 and 13 for each controller.
Device Manager (in Control Panel/System) should now show a DPT I2O
SCSI Host Adapter in the SCSI Controllers section.
You can use a text editor to remove the following entries from your
CONFIG.SYS file:
DEVICE=I2ODDL.SYS
DEVICE=DPTCDROM.SYS
You can also use a text editor to remove the MSCDEX entry from your
AUTOEXEC.BAT file. When you complete the installation procedure these
CONFIG.SYS and AUTOEXEC.BAT file entries are no longer needed.
Software Installation
5-9
Installing Windows 95
(OEM SR2 – Version 950b)
This section describes the steps to install Windows 95 (OEM SR2 Version)
on one disk drive or array group using a SmartRAID V controller. Ensure
that your disk drives, arrays, CD-ROM drive, and any other SCSI devices
attached to the controller are recognized by SMOR before proceeding.
Do the following to install Windows 95 (OEM SR2 Version) with a
SmartRAID V controller:
1.
Boot DOS from a bootable DOS diskette.
2.
Run FDISK and create a Primary DOS Partition of an appropriate
size.
3.
Mark the Primary DOS Partition Active.
4.
Reboot the system from the DOS diskette.
5.
Perform a system format (FORMAT C: /S).
6.
Run the SETUPCD.EXE program on the DPT diskette. This
program will create an AUTOEXEC.BAT and CONFIG.SYS to
support your CD-ROM drive.
7.
Ensure that the MSCDEX.EXE program is in your boot path or the
root directory of the C: drive.
8.
Reboot the system from the C: drive.
9.
Insert the Windows 95 OEM SR2 CD in the CD-ROM drive.
10. Enter the drive letter assigned to the CD-ROM and run SETUP (for
example, D:\SETUP).
11. Follow the Windows 95 installation prompts.
12. When Windows 95 is fully installed, start Device Manager (in
Control Panel – System) and look for a PCI Card entry under the
Other Devices category. There may be multiple entries if you have
multiple SmartRAID V host controllers installed. There may also be
an entry for PCI Memory Controller.
13. Select PCI Card and choose Update Driver from the Driver tab.
NOTE
If you are not prompted for a PCI Card or Memory Controller
Driver or if the Device Manager shows PCI Card with a yellow
exclamation mark in the Other Devices list, you might need a
patch for your motherboard chipset to recognize multifunction
bridge controllers.
5-10
Software Installation
Contact your motherboard vendor and apply the patch. After
the patch is installed and the system is restarted, Windows 95
will repeat the device discovery process. If you are installing
the operating system, go back to step 12 and continue the
procedure. If you are adding a SmartRAID V to an existing
configuration, go to step 3 on page 5-12 and continue the
procedure.
14. Insert the DPT Disk 1 diskette into the floppy drive and let Windows
search for the drivers.
15. Windows should identify the PCI Card as DPT I2O SCSI Host
Adapter. Select Finish.
At this point, you might need to force Windows 95 to copy the driver
from A: rather than the CD-ROM drive.
16. Select Cancel for testing the Device ROM.
17. If the Other Devices category shows a PCI Memory Controller,
select it and choose Update Driver from the Driver tab.
18. Let Windows search for the drivers on the DPT Windows 95/98
diskette.
19. Windows should now identify the PCI Memory Controller as DPT
I2O Memory Controller. Select Finish.
At this point, you might need to force Windows 95 to copy the driver
from A: rather than the CD-ROM drive.
20. If there are multiple SmartRAID V controllers, repeat steps 12
through 19 for each SmartRAID V controller in the computer. The
system should ask you to reboot when this action is completed.
21. Device Manager should now show a DPT I2O SCSI Host Adapter in
the SCSI Controllers section.
You can use a text editor to remove the following entries from your
CONFIG.SYS file:
DEVICE=I2ODDL.SYS
DEVICE=DPTCDROM.SYS
You can also use a text editor to remove the MSCDEX entry from your
AUTOEXEC.BAT file. When you complete the installation procedure these
CONFIG.SYS and AUTOEXEC.BAT file entries are no longer needed.
Software Installation
5-11
Installing Windows 95
(Retail Upgrade – Version 950a)
SmartRAID V controllers are not supported by Windows 3.1, therefore an
upgrade process from Windows 3.1 to Windows 95 does not apply. However,
the Retail Upgrade Version (950a) can be installed like a new operating
system if you have your original Windows 3.1 diskettes.
Follow the installation instructions in the Installing Windows 95 (OEM SR2
950b) section of this chapter. During the first phase of the installation the
Windows 95 installation program will prompt you to insert your original
Windows 3.1 disk to verify your eligibility for the upgrade. After the
upgrade is authenticated , the process continues as a normal installation
on a new system.
Adding SmartRAID V to an Existing Windows 95/
98 Configuration
This section describes the steps to install a SmartRAID V controller to an
existing Windows 95/98 configuration. Ensure that your hard drives,
arrays, CD-ROM drive, and any other SCSI device attached to the
controller are recognized by SMOR before proceeding.
Do the following to install a SmartRAID V controller in an existing
Windows 95/98 system:
NOTE
1.
Install the SmartRAID V controller in the system and start
Windows 95/98. If Windows prompts for a driver to support a PCI
Card or Memory Controller), go to step 7.
2.
If Windows did not discover the devices during boot, start Device
Manager (Control Panel – System) and look for a PCI Card entry in
the Other Devices category. There may be multiple entries if you
have multiple SmartRAID V controllers installed. There may also
be an entry for PCI Memory Controller.
If you are not prompted for a PCI Card or Memory Controller
Driver, open the Device Manager. If Device Manager displays
PCI Card with a yellow exclamation mark in the Other
Devices list, you might need a patch for your motherboard
chipset to recognize multifunction bridge controllers.
Software Installation
5-12
Contact your motherboard vendor and apply the patch. After
the patch is installed and the system is restarted, Windows 95
will repeat the device discovery process. If you are installing
the operating system, go to step 12 on page 9 and continue the
procedure. If you are adding a SmartRAID V to an existing
configuration, go to step 3 and continue the procedure.
3.
Select PCI Card and choose Update Driver from the Driver tab.
Insert the DPT Windows 95/98 diskette into the floppy drive and let
Windows search for the drivers.
4.
Windows should now identify the PCI Card as DPT I2O SCSI Host
Adapter. Select Finish. At this point, you might need to force
Windows 95 to copy the driver from A: rather than the CD-ROM
drive.
5.
Select Cancel for testing the Device ROM.
6.
If the Other Devices category shows a PCI Memory Controller,
select it and choose Update Driver from the Driver tab.
7.
Let Windows search for the drivers from the DPT Windows NT/95/
98 driver diskette.
8.
Windows should now identify the PCI Memory Controller as DPT
I2O Memory Controller. Select Finish. At this point, you may need to
force Windows 95 to copy the driver from A: rather than the CDROM drive.
9.
If there are multiple SmartRAID V controllers in the computer,
repeat steps 3 through 10 for each SmartRAID V controller. You
may be asked to reboot when this action is completed.
10. Device Manager should now show a DPT I2O SCSI Host Adapter in
the SCSI Controllers section.
Upgrading to Windows 98 from Windows 95
(Upgrade Version)
If the SmartRAID V controller is properly installed and recognized under
Windows 95 before starting the Windows 98 upgrade, no further
configuration is necessary. Perform the Windows 98 upgrade and the
functionality of the SmartRAID V controller will be retained.
If the controller is being added in addition to the upgrade, follow the
procedure in the Installing Windows 98 (Full Installation Version) section.
Software Installation
5-13
Installing Storage Manager for Windows
When the SmartRAID V controllers are properly recognized by Windows,
run SETUP.EXE from the DPT Windows 95/98 diskette and follow the
on-screen prompts to install the Storage Manager software.
The installation program detects which version of Windows you are using
and installs the appropriate Storage Manager files for your operating
system.
If you will use remote communication, be sure to select the
Communication Server install option.
If you are using Windows NT 4.0, do the following after Storage Manager is
installed to ensure optimum performance:
NOTE
1.
Start Storage Manager and double click on the controller icon to
display the Host Bus Adapter Info window.
2.
Select Configure to display the Configure Host Bus Adapter
window.
3.
Select Caching. When the HBA Caching Configuration dialog
appears, change the settings from Mandatory to Advisory.
4.
Select OK to exit the dialog and save the changes.
5.
Reboot the host system to enable the new settings.
If you reset the NVRAM on the controller, the changes to your
cache settings return to the factory default.
Software Installation
5-14
SCO UNIX
A diskette for SCO UNIX 3.2V4.2 and SCO UNIX OpenServer 5 (3.2V5.x) is
included with your SmartRAID V controller. The DPT diskette for SCO
contains the drivers for both versions of SCO. A second diskette contains a
Motif version of DPT Storage Manager for each version of SCO UNIX.
NOTE
When assigning numbers to controllers, SCO starts with zero.
Storage Manager numbers controllers starting with one.
Installing SCO UNIX 3.2V4.2 or OpenServer 5
This section describes the steps to install SCO UNIX 3.2V4.2 and up
(including ODT 3.x) and SCO UNIX 3.2V5.0.x (OpenServer 5) on a single
drive or array group. These SCO operating systems are configured to
support one SCSI hard drive or array group at ID 0, one tape drive at ID 2,
and one CD-ROM at ID 5. If these devices are to be installed at this time,
make sure that their IDs have been set accordingly.
To install SCO, follow the steps below:
1.
If installing SCO UNIX 3.2V4.2 or ODT 3.0, boot the system with the
N1 diskette. Enter the following string at the boot prompt :
defbootstr link=dpti4
If installing SCO UNIX 3.2V5.0.x, boot the system with the Boot
Disk. Enter the following string at the boot prompt:
defbootstr ahslink=dpti5
For SCO UNIX 3.2V5.04 or later the command is:
defbootstr link=dpti5
2.
When prompted for the SCO BTLD diskette (DPT volume), insert
the DPT BTLD diskette into the floppy drive.
3.
Insert the N1 or Boot disk again when prompted.
4.
If installing SCO 3.2V4.2 or ODT3.0, insert the N2 diskette when
prompted. During the Hardware Roster, the following line should
appear:
% adapter type=dpti ha=0 id=7 channel=0, PCI
If you are installing SCO UNIX 3.2V5.0.x., during the Hardware
Roster, the following line should appear:
% adapter type=dpti ha=0 id=7, PCI
Software Installation
5.
NOTE
5-15
Continue the installation according to the SCO documentation.
If you are installing from tape or CD-ROM, SCO UNIX 3.2V4.2
expects to find the tape at Controller 0, SCSI ID 2, LUN 0 and
the CD-ROM at Controller 0, SCSI ID 5, LUN 0.
SCO OpenServer 5 will locate the tape or CD-ROM at any
valid SCSI address.
Adding to an Existing SCO UNIX 3.2V4.2 or
OpenServer 5 System
When adding a DPT controller to an existing SCO UNIX configuration,
perform the following steps.
1.
Install the DPT controller.
2.
Boot SCO UNIX.
3.
After SCO UNIX has booted, place the DPT SCO driver diskette in
the floppy drive.
4.
Run the installpkg utility.
SCO UNIX will read the diskette and prompt you to enter the
package name to be installed. Enter dpti4 for SCO UNIX 3.2V4.2 or
ODT 3.0, and dpti5 for SCO OpenServer 5.
5.
Rebuild the kernel after the package is installed. Shutdown and
reboot the system to install the new kernel and add devices on the
controller.
SCO UnixWare 7
1.
Install the SmartRAID V controller as in Chapter 3, “Controller
Configuration and Installation”.
2.
Obtain the latest I2O supplement for UnixWare 7 from SCO (it
should be PTF7067E or higher).
3.
Copy or download the ptf file to the /tmp directory. Use the pkgadd
utility to install the I2O supplement by entering the command:
pkgadd –d /tmp/ptf7067e.s
4.
Follow the on-screen prompts to install the SCO UnixWare I2O
supplement.
5.
Shutdown and reboot the system.
Software Installation
5-16
NOTE
When you reboot several WARNING messages may appear and
scroll off the screen quickly. These will not occur after
“diskadd” is run and can be ignored.
6.
Use the diskadd utility to add drives or arrays that are attached to a
SmartRAID V controller.
Run the sdiconfig –l utility in a terminal window or non-GUI screen.
This will list all controllers recognized by the operating system and
any attached devices. Use this listing to determine the proper ID to
use for the diskadd utility.
The diskadd command to add devices on a secondary controller is:
diskadd cCbBtTdD
Where:
C
specifies the ID for the controller in the system.
B
is the controller bus number (from 0 – 7) to which the disks are
attached.
T
is the target controller number (ID). The value of T can be
0 – 31.
D
is the Logical Unit Number (LUN) of the disk device, from
0 – 31.
See the diskadd(1M) man page for more information.
For example, if a new drive or array is on the DPT controller, on channel 0,
with an ID of 0, the command would be:
diskadd c1b0t0d0
Software Installation
5-17
Novell NetWare
You must follow the procedures in the following sections to installing a
SmartRAID VI controller for Novell NetWare, especially as a booting
controller. There are specific steps that must be completed to ensure that
NetWare will recognize the DPT controller and any attached devices.
The DPT diskette for NetWare contains the device drivers you need to
install a DPT controller for NetWare 4.2 and 5.0. The diskette also contains
additional modules that allow you to monitor and configure a NetWare
server using Storage Manager on a remote client workstation.
If you are using NetWare 4.11, you must download the installation files
from the DPT Technical Support web site.
NOTE
The I2O drivers in the following Novell Support Packs or
driver kits will result in unexpected behavior of the
SmartRAID VI controller or disk arrays:
I2ODRV3
NetWare 5 Support Pack 2
NetWare 5 Support Pack 2A
If any of these Support Packs or updates are applied, use the
Novell install procedure to update the drivers with the files
from your DPT NetWare diskette.
After you have installed the I2O drivers you can now use the normal
NetWare procedures for modifying disk partitions, hot fixes, or volume
maintenance.
You can use SMOR to create or modify disk arrays or add Hot Spare
drives. If you install remote communication support for Storage Manager,
you can use Storage Manager from a client workstation to manage the
storage subsystem.
Software Installation
5-18
NetWare 4.11 – Boot Controller
This section describes how to install a SmartRAID VI controller as a boot
device for NetWare 4.11 servers.
NOTE
Part of this procedure requires that you have enough free
space on your C: drive to hold the contents of the NetWare
CD-ROM.
1.
Download the NW411.ZIP file from the DPT Technical Support web
site. Extract the files to a floppy disk.
2.
Install the DPT controller as the primary controller and attach the
disk drives. If other I/O controllers are present, ensure that the
SmartRAID VI boot controller is in a lower number PCI slot than
the other controllers.
3.
Install and configure a secondary controller (SCSI or IDE) for the
CD-ROM drive and any other devices such as a tape drive.
4.
Boot the system and use SMOR to create any required disk arrays.
Press Ctrl+D during boot to start SMOR.
5.
Create a DOS partition on the logical storage unit (drive or array
group) with the lowest SCSI ID.
6.
Boot the system to DOS from the partition created in step 5. Ensure
that you load the required CD-ROM drivers.
7.
Insert the NetWare 4.11 release CD-ROM into the CD-ROM drive.
Insert the diskette you created and run DPT411.BAT. This will copy
the necessary files to the C: drive.
8.
When asked to choose a disk driver select I2OPCI.NLM. Enter the
slot number where the controller is installed and press F10.
9.
Select Save parameters and load driver.
10. Choose Yes to select another driver.
11. Select BKSTROSM.HAM. Enter the slot number where the
controller is installed and press F10.
12. Select Save parameters and load driver.
13. Install any additional drivers and continue the installation.
14. Ensure that BKSTROSM.HAM is listed after I2OPCI.NLM in your
STARTUP.NCF file.
Complete the NetWare installation according to the Novell documentation.
Software Installation
5-19
NetWare 4.11 – Secondary Controller
This section describes how to integrate a DPT SmartRAID VI controller
into a new or existing Novell NetWare 4.11 system as a secondary
controller. Ensure that your NetWare version is at the minimum patch
level specified by Novell.
NOTE
The server must have patch kit IWSP4 or later installed to use
the drivers supplied by DPT. NetWare 4 Support Pack 8
contains the latest Novell I2O drivers. These drivers are newer
than the ones from DPT but have not been tested for
compatiblity with DPT controllers.
To integrate a DPT controller into a existing Novell NetWare 4.11 system,
do the following:
1.
Download the NW411.ZIP file from the DPT Technical Support web
site. Extract the files to a floppy disk.
2.
Shutdown the server and switch off the power. Install the DPT
controller and any attached storage devices as a secondary disk
subsystem. Note the slot used by the controller, you will need this
information later in the install process.
3.
Power on the server and use SMOR to configure the storage
subsystem. Press Ctrl+D during boot to start SMOR.
4.
Start NetWare. When the console prompt appears type, LOAD
INSTALL and press Enter.
5.
Select PRODUCT OPTIONS and press Enter.
6.
Select INSTALL A PRODUCT NOT LISTED and press Enter.
7.
Insert the diskette you created in drive A: and install I2O and
BLOCK STORAGE support only.
8.
Accept the defaults for the remaining install options.
9.
Return to the main install screen and select NCF OPTIONS.
10. Select EDIT STARTUP.NCF.
Ensure that BKSTROSM.HAM is listed after I2OPCI.NLM.
11. Shutdown and restart the server.
Software Installation
5-20
NetWare 4.2 – Boot Controller
This procedure is for installing a SmartRAID VI controller as the boot
device in a NetWare 4.2 server.
1.
Install the SmartRAID VI controller as the primary controller and
attach the disk drives or storage subsystem.
Ensure that you are using a CD-ROM that is not attached to the
DPT controller.
2.
Create a directory on the C: drive named NWUPDATE.
3.
Insert the DPT NetWare diskette into the floppy disk drive.
4.
Copy A:\NW42 to C:\NWUPDATE.
5.
Start the NetWare installation program from the CD-ROM. Use the
Custom Installation option.
6.
On the Server Drivers - Summary menu, highlight Disk and CDROM
Drivers and press Enter.
7.
Highlight Select an Additional Driver and press Enter.
Select I2OPCI.NLM and press Enter.
Enter the slot number.
Highlight Save Parameters and Continue. Press Enter.
8.
Answer Yes when prompted to select an additional driver.
9.
Select BKSTROSM.HAM and press Enter.
Enter the slot number.
Highlight Save Parameters and Continue. Press Enter.
10. Answer No when prompted to select an additional driver.
11. Highlight Continue Installation and press Enter.
12. Verify that BKSTROSM.HAM is listed after I2OPCI.NLM in your
STARTUP.NCF file.
Continue with the NetWare 4.2 installation according to the Novell
documentation.
Software Installation
5-21
NetWare 4.2 – Secondary Controller
This section describes how to integrate a DPT SmartRAID VI controller
into a new or existing Novell NetWare 4.2 system as a secondary controller.
Ensure that your NetWare version is at the minimum patch level specified
by Novell.
NOTE
NetWare 4 Support Pack 8 contains the latest Novell I2O
drivers. These drivers are newer than the ones from DPT but
have not been tested for compatiblity with DPT controllers.
To integrate a DPT controller into a existing Novell NetWare 4.2 system,
do the following:
1.
Download the NW411.ZIP file from the DPT Technical Support web
site. Extract the files to a floppy disk.
2.
Shutdown the server and switch off the power. Install the DPT
controller and any attached storage devices as a secondary disk
subsystem. Note the slot used by the controller, you will need this
information later in the install process.
3.
Power on the server and use SMOR to configure the storage
subsystem. Press Ctrl+D during boot to start SMOR.
4.
Start NetWare. When the console prompt appears type, LOAD
INSTALL and press Enter.
5.
Select INSTALL A PRODUCT and press Enter.
6.
Select INSTALL A PRODUCT NOT LISTED and press Enter.
7.
Insert the diskette you created from the download file in drive A:
and install I2O and BLOCK STORAGE support only.
8.
Accept the defaults for the remaining options.
9.
Return to the main install screen and select NCF OPTIONS.
10. Select EDIT STARTUP.NCF.
Ensure that I2OPCI.NLM is listed after BKSTROSM.HAM.
11. Shutdown and restart the server.
Software Installation
5-22
NetWare 5.0 – Boot Controller
Use the following procedure to install support for a SmartRAID VI
controller during installation of NetWare 5. The controller should be
installed at the beginning of the procedure so that it can be detected by the
install routine.
1.
Install the DPT controller as the primary controller and attach the
disk drives. You can use SMOR to create disk arrays or Hot Spare
drives when you boot the system.
2.
Install and configure a secondary controller (non-I2O SCSI or IDE)
for the CD-ROM drive and any other storage devices such as a tape
drive.
3.
Create a directory on the C: drive named NWUPDATE.
4.
Insert the DPT NetWare diskette into the floppy disk drive.
5.
Copy A:\NW5 to C:\NWUPDATE.
6.
Start the NetWare installation program from the CD-ROM. Use the
Custom Installation option.
7.
On the Server Drivers - Summary menu, highlight Disk and CDROM
Drivers and press Enter.
When prompted choose I2OPCI.NLM and BKSTROSM.HAM as disk
drivers.
8.
Complete the NetWare installation according to the Novell
documentation.
Verify that I2OPCI.NLM is listed after BKSTROSM.HAM in your
STARTUP.NCF file.
9.
NOTE
Install the latest support pack for NetWare 5.0.
Installing a NetWare Support Pack will overwrite the I2O
drivers supplied by DPT. If you encounter problems after
installing the Support Pack, you can restore the DPT files by
copying them from the DPT NetWare diskette.
Software Installation
5-23
NetWare 5.0 – Secondary Controller
This section describes how to integrate a DPT SmartRAID VI controller
into a new or existing Novell NetWare 5.0 system as a secondary controller.
Ensure that your NetWare version is at the minimum patch level specified
by Novell.
NOTE
NetWare 5 Support Pack 4 contains the latest Novell I2O
drivers. These drivers are newer than the ones from DPT but
have not been tested for compatiblity with DPT controllers.
To integrate a DPT controller into an existing Novell NetWare 5.0 system,
do the following:
1.
Shutdown the server and switch off the power. Install the DPT
controller and any attached storage devices as a secondary disk
subsystem. Note the slot used by the controller, you will need this
information later in the install process.
2.
Power on the server and use SMOR to configure the storage
subsystem. Press Ctrl+D during boot to start SMOR.
3.
Start NetWare and insert the DPT NetWare diskette in the floppy
disk drive.
4.
When the console prompt appears type, LOAD NWCONFIG and
press Enter.
5.
Select DRIVER.
6.
Select DISK DRIVER.
7.
Install I2OPCI.NLM and BKSTROSM.HAM. You will need tospecify
A:\NW5 as the location for the files.
8.
Ensure that I2OPCI.NLM is listed after BKSTROSM.HAM in your
STARTUP.NCF file.
Software Installation
5-24
Installing Remote Communication Support
1.
Insert the DPT NetWare diskette in the server’s floppy disk drive.
2.
Enter the following command at the prompt:
LOAD A:\DPTINST.NLM
3.
You will be prompted to choose whether to load the remote
communication support immediately or delay loading until the
server is restarted.
If you choose to wait until the server restarts, you will not be able to
monitor the DPT subsystem from a remote client until the server is
shut down and restarted. Otherwise, the remote communication
support is loaded and available as soon as you exit the install program.
Red Hat Linux
To install a DPT controller under Red Hat Linux 5.2 or 6.x, you must have
the DPT Linux diskettes. The diskette images can be downloaded from the
DPT web site (www.dpt.com).
NOTE
The DPT drivers are for Linux kernel versions 2.0.36 and 2.2.5
or higher. If your kernel has been updated by a patch, the
drivers might not function. In this case, contact DPT Technical
Support for assistance. Refer to the READ.ME file for the latest
information Linux support for DPT SmartRAID V controllers.
The REDHAT.TXT file is a HOWTO document that describes the
installation of a SmartRAID V controller in a system using Linux.
DPT controller and RAID management is available through the DPT
SMOR utility or Storage Manager for Linux. Storage Manager for Linux
requires Linux 6.x (kernel version 2.2.5 or higher). See Chapter 4, “Storage
Manager on ROM” or Chapter 6, “Storage Manager” for additional
information about using these programs.
Software Installation
5-25
Red Hat Linux 5.2 and 6.0
These instructions are for installation of the DPT device driver for Linux
on a system using Red Hat Linux 5.2 or 6.0.
1.
Install and configure your DPT controller according to the
instructions in Chapter 3, “Configuration and Installation”.
2.
If you want to install Linux on a disk array, use SMOR to configure
any disk arrays before beginning the operating system installation.
NOTE
Any partitions created on disks that are not part of an array
will be unavailable if you subsequently use the disk in a RAID
configuration.
3.
Boot the system from the DPT Red Hat Linux boot diskette and
continue with the Linux installation according to the instructions in
the REDHAT.TXT file.
Red Hat Linux 6.1
These instructions are for installation of the DPT device driver for Linux
on a system using Red Hat Linux 6.1.
1.
Install and configure your DPT controller according to the
instructions in Chapter 3, “Configuration and Installation”.
2.
If you want to install Linux on a disk array, use SMOR to configure
any disk arrays before beginning the operating system installation.
NOTE
Any partitions created on disks that are not part of an array
will be unavailable if you subsequently use the disk in a RAID
configuration.
If your hard drives do not have partition tables or do not
contain a recognizable partition), you will be prompted to
initialize these disks. The operating system can automatically
remove the data or you can manually partition the drives. For
more information, refer to the Red Hat documentation.
3.
Boot the system with the Red Hat Linux 6.1 boot diskette in the
floppy drive and CD Disk 1 in your CD-ROM drive.
4.
When the Red Hat 6.1 Welcome screen appears, type:
expert
at the boot: prompt.
Software Installation
5-26
5.
When prompted, insert the DPT driver diskette and press OK.
6.
You will be prompted for the language and keyboard type.
7.
You will be asked to specify the location of the media packages to
be installed.
If you are installing from CD-ROM, select Local CD-ROM. If you
selected CD-ROM, you will be asked what type of CD-ROM you
have.
8.
The system will ask you to specify a driver. Scroll down and select
the DPT I2O driver.
9.
The system will inform you that it has found DPT I2O SmartRAID.
10. Unless there are other third party devices to install in your system,
select Done.
Continue with the Red Hat Linux 6.1 installation according to the Red Hat
documentation.
Installing Storage Manager for Linux
The installation program detects which version you are using and installs
the appropriate Storage Manager files for your operating system. If you
will use remote communication, be sure to select the Communication
Server install option.
Storage Manager for Linux requires X Window System support. The DPT
distribution includes the LessTif components. DPT has tested Storage
Manager operation with both the K Desktop Environment (KDE) and
GNU Network Object Environment (GNOME) graphical user interface.
NOTE
Use of more than 256 colors in your GUI may result in Storage
Manager displaying incorrect colors. If this occurs, reset your
color setting to use only 256 colors.
1.
Create a temporary directory. To create a directory in Linux, type:
mkdir <directory name>
2.
Change the working directory to the one you just created:
cd <...>
Download the Storage Manager for Linux archive from the DPT
web site to this directory.
Software Installation
4.
5-27
When the download is complete, run the following command from
the prompt:
tar -xvf linsm218.tar
5.
After the file extraction is finished, run the following command
from the prompt:
./install
You will be asked if you want the DPT Communication Server to
load on boot. The default is No. If you will access this host from a
remote Storage Manager client, enable the DPT Communication
Server.
6.
Reboot the system, when the installation is complete.
7.
Open a command prompt window and change the working
directory to /usr/dpt.
8.
To launch Storage Manager, type:
./dptmgr
This will open a Storage Manager window on your desktop.
5-28
Software Installation
CHAPTER 6:
Storage Manager
DPT’s Storage Manager gives you complete
control over your storage subsystem. You
can manage your storage locally, remotely
across a network or by using a modem.
Physical View
Storage Manager brings out the best in your
DPT hardware, from checking your equipment configuration – to managing your disk
arrays – to providing online event logging
and performance statistics.
Logical View
SmartRAID V controllers also include
Storage Manager on ROM, which lets you
build disk arrays prior to installing your
operating system and Storage Manager. See
Chapter 4 for additional information about
Storage Manager on ROM.
Each physical or logical storage
device is shown as an icon. To
display information or configure the
device, click on its icon.
Storage Manager’s Physical View
presents the storage subsystem as
it is physically connected.
Storage Manager’s Logical View
presents the storage subsystem as
it appears to the operating system.
Icon Based
Event Notification
DPT controllers store events – fault
conditions as well as status
changes – in controller RAM.
Events can thus be viewed even if
the storage subsystem is not
operating.
Remote Communication
Storage Manager can be used to
configure controllers and disk
arrays on remote systems across a
TCP/IP network.
6-1
Storage Manager
6-2
Introduction
DPT Storage Manager is included with your SmartRAID V controller. This
utility performs several functions:
• Checks hardware configuration.
• Allows you to create, expand or delete disk arrays.
• Provides online functions for the DPT storage subsystem such as
event logging and notification, array status and I/O statistics.
• Provides remote access to DPT hardware and attached storage
devices across a TCP/IP network.
Storage Manager versions for supported operating systems are on the
DPT diskettes that were included in your controller kit.
System Requirements
Storage Manager can be installed on a computer with one of the following
operating systems:
• Windows 95/98
• Windows NT 4.0
• SCO UNIX 3.2V4.2
• SCO OpenServer 5
• SCO UnixWare 7.0
• Linux 6.x
NOTE
The versions of Storage Manager for SCO platforms uses the
Motif graphical user interface. Storage Manager will not
display correctly if your display is set to use more than 256
colors.
Before running Storage Manager, be sure that your mouse driver is
installed.
Storage Manager
6-3
Running Storage Manager
You can run Storage Manager:
• On the same computer that contains the DPT hardware and
peripheral devices (local operation).
• Remotely, across a TCP/IP network. This lets you view and configure
servers from remote locations.
• In demo mode, where Storage Manager simulates a storage
subsystem that contains various types of DPT controllers and
peripheral devices. This lets you experiment with various storage
configurations.
Using Storage Manager Locally
Storage Manager initially scans for DPT hardware installed on the
computer on which it is run. If one or more DPT controllers are found, the
storage subsystem hardware configuration is displayed.
Using Storage Manager Remotely
Windows NT, NetWare, SCO UNIX servers and Windows 95/98 systems can
be viewed and configured across a TCP/IP network from Windows NT,
Windows 95/98, or SCO UNIX workstations running Storage Manager.
To install the DPT communication server software, ensure that
Communications Server is selected in the DPT Storage Manager Setup
dialog. Refer to the Remote Communication section in this chapter for
detailed information about using the remote communication feature.
6-4
Storage Manager
Storage Manager on ROM
Storage Manager on ROM (SMOR) is a ROM-based version of Storage
Manager which is provided on your SmartRAID V controller. If you are
installing a new system, use SMOR to configure your RAID subsystem
prior to installing your operating system.
You can access SMOR during the system boot by pressing Ctrl+D after the
DPT I2OBIOS is loaded. Refer to “Chapter 4, Storage Manager on ROM”
for more information.
Physical Configuration View
The first window displayed by Storage Manager is the Physical
Configuration View. This window displays each DPT controller in the
system along with the peripheral buses and attached devices. Icons
representing disk drives, CD-ROMs, tapes, printers, bridge controllers,
scanners and jukeboxes are displayed. Devices are sorted by controller
number and device ID from lowest to highest. To see a list of all icons and
their meaning, select Help – Legend of Icons in Storage Manager.
Disk drives that are part of an array have the word RAID on the drive icon.
Disk drives that are assigned as Hot Spares have a red circle with a white
cross on the drive icon. Select Legend of Icons from the Storage Manager
Help menu to see a list of the various icons and their meaning.
Storage Manager
6-5
The Switch View button toggles between the Physical Configuration View
and the Logical Configuration View window. Create Array Group starts the
process of creating a RAID logical disk. Print will print a text report of the
subsystem configuration.
6-6
Storage Manager
Logical Configuration View
On the right side of the Logical Configuration View window are all physical
devices that are attached to DPT controllers. On the left side of the
window are the associated logical devices as seen by the host computer.
Non-disk devices use the same icon for both logical and physical views.
Disks drives can be seen either as individual drives or as members of
arrays. In either case, the drive or array is represented on the left side of
the window as a Logical Storage Unit (LSU). Arrays that make up a Parity
Group are displayed as RAID 1 or RAID 5 icons that appear between the
LSU icon on the left and the drives on the right.
Devices are displayed in order of device type, with all non-disk devices
displayed first, followed by all disk drives not assigned to an array, Hot
Spares, and finally, all arrays by RAID level.
Storage Manager
6-7
Logical Device Address
Every device and array is assigned a logical device address by Storage
Manager. This is the address used by the host operating system to access
the device or array. Logical device addresses appear in parentheses under
the logical device and LSU icons on the Logical Configuration View
window.
The logical device address is composed of four fields (HBA, Bus, Device,
LUN) and are assigned to devices as follows:
HBA (Host Bus Adapter)
The controller to which the device is attached. PCI bus slots are scanned
from lowest to highest looking for DPT controllers. As DPT controllers are
found they are assigned numbers incrementally, starting with 0.
Bus
The controller bus to which the device is attached. SmartRAID V
controllers can have up to three buses.
Storage Manager
6-8
Device
The unique ID for that device. For an array, this is the lowest ID among
the drives that make up the array.
LUN
Logical Unit Number for that device (normally 0).
NOTE
Array groups are automatically assigned an address that
corresponds to the lowest logical device address used by a
device in that array group. For example, if the lowest member
device address for an array is HBA:0 Bus:1 ID:12 LUN:0, the
LSU address will be HBA:0 Bus:1 ID:12 LUN:0.
When an Array Group has been created, its logical device address will not
change if the drive with the lowest ID is replaced by a Hot Spare. When a
Hot Spare replacement occurs, the failed drive automatically becomes the
new Hot Spare. If you choose not to use that drive as a Hot Spare, Storage
Manager will prompt you to select an unused logical device address for
that drive.
Status
Status is reported by SmartRAID V controllers for arrays and drives.
Some status conditions are indicated by Storage Manager through status
flags on the drive or array icons. View the information window for the drive
or array to see specific status information. Changes in status conditions
are logged and can also be broadcast to selected recipients.
The following table lists the general status conditions that can be indicated
for drives or arrays. The Status field in the device or array information
window displays additional information for any reported status condition
other than Optimal.
Storage Manager
Building
Status flag: blue - This flag appears on an array icon
when an array that was created is being built.
Status flag: white - This flag appears on the drive icon
that data will be copied to in a RAID 1 array build.
Failed
Status flag: red - This flag appears on a drive icon
when the drive has failed.
On an array icon, this flag indicates that one drive has
failed in a RAID 0 array or two or more drives have
failed in a redundant array.
Impacted
Status Flag: white - The array is in a degraded mode
but no host I/O request has occurred since the
degraded status was reported.
Missing
Status flag: black - This flag appears on a drive icon
when the drive is physically missing or will not
respond to commands on the peripheral bus.
On an array icon, this flag indicates that the array has
been created or modified within Storage Manager but
the array has not yet been built.
On a Hot Spare icon this flag indicates that the Hot
Spare has not yet been created by the controller.
Optimal
Status flag: none - The drive or array is fully
functional and is not running any operations that
impact performance.
Pending
Status flag: blue - This flag appears on an array icon
when the array has been created and the build is
queued on the controller, but is not yet started.
Rebuilding
Status flag: blue - This flag appears on an array icon
when the array is being rebuilt after a drive failure.
Status flag: white - This flag appears on a drive icon
when data is being rebuilt onto that drive.
Verifying
Status flag: white - This flag appears on drive or array
icons to indicate that redundancy is being verified on
the array. Performance may be affected.
Warning
Status flag: yellow - This flag appears on a drive icon
(with S.M.A.R.T. failure prediction) to indicate that
failure may be imminent.
This flag appears on a RAID 1 or RAID 5 array that
has a failed or missing drive.
6-9
6-10
Storage Manager
Information Windows
Double click on a controller or device icon to display an Information
window for that controller or device.
Host Bus Adapter Information Window
The Host Bus Adapter Info window displays configuration information
reported by that controller.
The Controller section displays the Model, Serial #, Firmware revision and
amount of installed Cache. ECC is checked only if DPT ECC SIMMs or
DIMMs are installed. The current temperature reported by the onboard
sensor is also displayed.
The Attached Modules section shows the expansion modules and type and
capacity of memory modules installed.
The SCSI Bus and Host Bus sections display the current configuration of
the respective bus. Buttons available include Configure, Event Log, I/O
Stats and Print. Battery displays the Battery Backup Configuration dialog if
a BB4050 module is installed.
Storage Manager
6-11
Battery Backup Configuration
NOTE
During the initial calibration cycle for a BB4050 module, the
controller disables automatic, low-battery write-through mode.
After the calibration, use the Battery Backup Configuration
dialog to set a threshold for entering write-through mode when
the battery charge drops below a predetermined level.
This option lets you view the status of the BB4050 Battery Backup Module
and set operating parameters when the battery capacity reaches a
predetermined level. The battery Status and available Backup capacity (in
hours) is displayed. The Backup value is monitored periodically by
software and changes whenever the battery pack is charging or
discharging.
The Backup Capacity Warnings parameters let you activate the following
options when the BB4050 module reaches a predetermined level of
remaining backup capacity (in hours). The capacity warning thresholds
can change depending on the capacity and number of memory modules.
Generally, more cache will result in lower battery backup capacity.
Low: Auto Write-Through
When checked this option automatically sets the cache to write-through
mode when the specified number of hours (Hrs.) of battery capacity
remain. The number of hours must be greater than or equal to the number
of hours for a Predictive Failure Warning.
Predictive Failure Warning
When this option is checked a warning message will be issued when the
battery backup capacity is almost depleted. Enter the number of hours
(Hrs.) of remaining backup capacity when this message should be issued.
Storage Manager
6-12
Click the Defaults button to set the Battery Capacity Warnings to their
default values. Actual default values are calculated by the controller based
on the size of the cache memory.
BB4050 Battery Status Messages
Battery Module status is reported as one of the following:
Full
The battery is fully charged.
Charging
The battery is charging after being
partially or fully discharged.
Initial calibration charge
The battery is in the first phase
(charging) of the initial calibration
cycle.
Initial calibration discharge
The battery is in the discharge
phase of the initial calibration
cycle.
Initial calibration recharge
The battery is in the final phase
(recharge after discharge) of the
initial calibration cycle.
Maintenance calibration discharge
The battery is being discharged as
part of the maintenance cycle.
Maintenance calibration charge
The battery is charging as part of
the maintenance cycle.
Initial Calibration (BB4050)
The Initial Calibration operation is started when a BB4050 Battery Backup
Module is first installed on a controller. This function ensures the battery
is fully charged for subsequent backup operation. The calibration cycle
requires approximately 24 hours to complete. The BB4050 module will not
be able to provide backup capability during this operation. When the
process is completed, the BB4050 module is ready for normal operation.
This function has three phases:
1.
Initial charge: The battery is charged to its full capacity.
2.
Discharge: The battery is discharged until no backup capacity
remains.
3.
Recharge: The battery is recharged to full capacity.
Storage Manager
6-13
Battery Maintenance (BB4050)
IMPORTANT
The BB4050 module will not be able to provide backup capability during this
operation.
The Maintenance operation can be used to ensure the battery remains
capable of accepting a full charge. The Maintenance cycle completely
discharges and then recharges the battery pack. This helps to eliminate
any voltage-depression effects resulting from the battery pack being
partially discharged and then recharged repeatedly during normal
operation.
When this option is selected, the date of the most recent Maintenance
cycle is displayed.
HBA Configuration
To modify hardware parameters for the SmartRAID V controller, click the
Configure button in the Host Bus Adapter Info window. The Configure Host
Bus Adapter window appears.
Storage Manager
6-14
SCSI ID
DPT controllers are set at ID 7 by default. The ID for
a SmartRAID V Fibre Channel controller can change
dynamically as the Fibre Channel loop self-configures.
Refer to Chapter 3, “Configuration and Installation”
for information about selecting an alternate SCSI ID.
SCSI TERMPWR
By default SmartRAID V controllers supply
termination power through the TERMPWR line on the
SCSI cable. Refer to the information on TERMPWR in
Chapter 7, “Theory of Operation” before you change
this parameter.
Ext. Cable Detect
This option causes the SmartRAID V controller to
check for the presence of an external SCSI cable and
automatically lower the maximum SCSI bus rate to
5MHz when an external cable is found. Some external
SCSI cables are not capable of handling higher
transfer rates without producing data errors. Enable
this feature to prevent data integrity problems if you
have external SCSI devices with long cables.
Transfer
The maximum possible SCSI transfer rate.
SmartRAID V controllers automatically negotiate with
each SCSI device at power-up, or reset, to determine
the maximum SCSI transfer rate. This parameter
should not be changed except when troubleshooting
SCSI bus errors. If data errors are eliminated by
setting the value to a lower rate, there might be
problems with the length of the bus or the bus
termination.
Termination
This parameter sets SCSI termination for the
controller. The default (Auto) should not be changed
unless both internal and external Wide SCSI cables or
8-bit (Narrow) cables are connected to the controller.
Refer to Chapter 3, “Configuration and Installation”
for information about configuring SCSI bus
termination.
Address
This field displays the controller memory address
assigned by the host BIOS. This value cannot be
changed.
Storage Manager
6-15
Buttons available are:
Test Alarm – lets you test the audible alarm on the controller.
Flash – displays the Flash Configuration dialog to update the controller
firmware and BIOS.
Caching – allows you to select the controller cache parameters.
Defaults – resets the controller configuration to factory default settings.
Cancel – cancels any changes you have made and returns to the Host Bus
Adapter Info window.
NOTE
The specific buttons you see depend on the controller model
and configuration.
Flash Configuration
This dialog is displayed when you click on the Flash button in the
Configure Host Bus Adapter window.
NOTE
It is easier to perform an update if the image files are placed
in the root directory of a diskette. Each component must be
upgraded as a separate operation.
Storage Manager
6-16
The current controller Model, Firmware Version and Type are displayed. To
specify an image file for the flash operation, you can type a path and
filename in the Filename field or click on Browse to use the standard file
selection dialog.
Firmware images are contained in a xxxxxxxx.FWI file, where the
8-character file name consists of the 4-digit controller model and a 4-digit
release number. A DPT I2O BIOS image is contained in a file named
I2OBIOS.xxx (where, xxx is the version number). SMOR updates are
contained in a file named SMORyyyy.xxx. Where, yyyy is the version
number and the file extension (xxx) is the build number. You can obtain
the latest files by contacting DPT Technical Support or download the files
from our FTP site at ftp.dpt.com.
When you select an image file, Storage Manager reads the file to
determine the type of image selected, firmware, I2O BIOS, or SMOR. The
Version, Date and Type are displayed in the corresponding section of the
dialog.
Click OK to begin the flash operation. Click Cancel to return to the
Configure Host Bus Adapter window.
Controller Caching for Windows NT
The SmartRAID V controller manages its data cache according to built-in
algorithms (see Caching Configuration in this chapter and Chapter 7,
“Theory of Operation” for additional information). If you are using
Windows NT 4.0, do the following to ensure optimum performance:
NOTE
1.
Click Caching in the Configure Host Bus Adapter window to display
the HBA Caching Configuration dialog.
2.
When the HBA Caching Configuration dialog appears, change both
of the settings to Advisory.
3.
Click OK to exit the dialog and save the changes.
4.
Reboot the Windows NT system to enable the new settings.
If you reset the NVRAM on the controller, the changes to your
cache settings return to the factory default of Mandatory. In
that case, repeat this procedure to ensure optimum
performance under Windows NT.
Storage Manager
6-17
Device Information Window
This window displays the following information:
Description
The manufacturer and model.
Revision
The drive firmware version.
Address
The logical address of the device.
Capacity
Storage capacity of the device in megabytes.
For removable media disk devices, Capacity is reported for the currently
inserted media. Disk devices also display the number of Sectors on the
media along with the Physical and Logical sector size.
Other information such as the Transfer rate or whether the device supports
Removable media are also reported depending on the system
configuration.
The current device Status is displayed. General status conditions (other
than Optimal) are also indicated by colored flags on the device icon. SCSI
Capabilities shows which supported features are enabled [x] for the device.
Members of an array display the name and RAID level of the array to
which they belong and the Stripe Size for the array.
6-18
Storage Manager
Various buttons are available depending upon the type of device. Disk
drive devices will have an Event Log and I/O Stats button. Disk drives that
are not members of arrays have Make HotSpare, Configure and Format
buttons. Print is always available.
Disk drives that are members of arrays display a Fail Drive button.
If the drive is a Hot Spare, the Remove Hotspare button replaces the Make
Hotspare button.
Caching Configuration
Clicking the Configure button in the SCSI Device Information or the Array
Group Information window displays the Device Configuration dialog. This
dialog lets you change the Write caching parameters.
Click Defaults to set the cache operation to the default setting.
Click Cancel to exit this dialog without saving changes. Click OK to exit
with changes.
Write back mode defers writes to disk until after command completion and
generally provides better performance. Write through mode writes all data
to disk for each write command before command complete status is
returned to the host. The data can also be cached for subsequent read
commands.
The Predictive Cache checkbox enables and disables the predictive
caching feature on SmartRAID V controllers. By default the box is not
checked and the feature is disabled.
Storage Manager
6-19
Predictive caching is a feature of SmartRAID V controllers that attempts
to reduce average disk access time by determining when the host is
requesting data that it read previously and reading in additional sequential
data before it is actually requested.
This feature can provide enhanced performance when you have a host
with 20 or more users who are frequently accessing large, individual
sequential files. Unless your configuration is similar to this, you will not
see any benefit from using Predictive Cache and using it will adversely
affect overall controller performance. For additional information about
predictive caching, see Predictive Caching Algorithms in Chapter 7.
Array Groups
The tool bar at the top of the Logical Configuration View window contains
Create Array Group, Expand Array Group (Windows NT only) and Delete
Array Group buttons. The Expand Array Group button allows you to add
drives to an existing array and dynamically resize the logical drive.
SmartRAID V controllers can implement any combination of RAID level 0,
1 or 5 disk arrays. RAID 0 arrays can be any combination of individual
drives. RAID 1 arrays are composed of two drives. RAID 5 arrays must
contain at least three drives. One or more arrays of the same RAID level
can be combined into a Parity Group. (This is also referred to as RAID 0+1
for multiple RAID 1 arrays or RAID 0+5 for multiple RAID 5 arrays.) The
drives in a Parity Group appear as a single Logical Storage Unit (LSU) to
the host computer.
NOTE
On SmartRAID V controllers, all drives in an array must be
attached to the same controller. For controllers with drives on
multiple SCSI bus channels, arrays are built using the drives
in the order in which they were selected regardless of the bus to
which the drive is attached.
This behavior can be used to create fault-tolerant SCSI bus
configurations using pairs of drives on alternate channels in a
RAID 1 or RAID 0+1 array or with drives distributed across all
three channels for a RAID 5 or RAID 0+5 array.
6-20
Storage Manager
From the Logical Configuration View, you can double-click on an Array
Group icon to display the Array Group Information window. Information
windows for arrays that are part of a Parity Group can be viewed by
clicking on the icons for those arrays.
Storage Manager
6-21
The Array Group Information window displays the following information:
NOTE
Name
The descriptive name assigned to the array. An icon in the
upper right corner of the window indicates the RAID level.
Address
This is the logical device address used by the host
operating system to access the logical drive. The address
is the same one as the lowest device address disk drive in
the array.
Capacity
The total usable storage capacity of the array in
megabytes (MB).
Status
The current status of the array as reported by the
controller.
Hotspares
Displays a list of any Hot Spare drives that are available
to protect the array in the event of a drive failure.
Components
Displays the logical address, model and stripe size for
each member of the array. If this a Parity Group
information window, the list displays the address or name
and stripe size for each disk array that is a member of the
parity group.
The availability of the various buttons depends on the current
configuration and state of the array.
Buttons include Configure, Print, Event Log, I/O Stats, Verify/Fix and Name.
A Rebuild button will appear for redundant arrays that have a failed drive.
For arrays that are building or rebuilding, a Stop Bld button is displayed.
For arrays that have a build pending, a Build button is displayed.
Arrays running Verification will have Stop Vfy button.
Storage Manager
6-22
Creating an Array Group
To create an Array Group, do the following:
1.
Click the Create Array Group button.
2.
The Select Array Type dialog appears.
Specify whether or not you want Drive fault tolerance (RAID 1 or 5)
or No fault tolerance (RAID 0)
Select whether the array is to be Optimized for Capacity (RAID 5) or
Optimized for Performance (RAID 1).
As you make your selections, the Chosen Array Parameters change
to indicate which RAID level and stripe size best suit your selection.
You can customize the RAID level and stripe size defaults by
selecting the Override button.
3.
Click Continue to select the drives you want to use. The Logical
Configuration View window will appear with the caption Choosing
drives for Array (RAID n), where n is the RAID level chosen.
Select the drives you want to use in the Array Group as follows:
• To add drives:
Mark each drive to be added by clicking on it. A green check mark
indicates that a drive is selected.
Click Include Drive to add the marked drives to the new Array
Group. You may need to scroll the window to view the Array
Group.
• To remove drives:
Click on the drive icons you want to remove and then click Remove
Drive.
Storage Manager
6-23
During the drive selection process some drives might be displayed
in a blue color. This indicates that these drives cannot be included
in the array unless you change the configuration. You must either
select more drives for the array or remove one or more drives
from the array. See the introduction to the Array Groups section
for rules regarding the number of drives that can be included in
arrays.
4.
When you finish selecting the drives to be included in the new Array
Group, click Done.
The icon for the Array Group will appear with a black flag until you
start the build process by saving your changes.
5.
When you are finished creating arrays, exit Storage Manager.
You will be prompted to save the configuration changes. If you save
the configuration, the build operation will start automatically. If you
have created multiple arrays, they are built one at a time in the
order they were created.
You can also start the build without exiting Storage Manager by
selecting File–Set System Configuration.
NOTE:
For large redundant arrays, this process can take several
hours to complete. You can exit Storage Manager and perform
other activities on the system while the build continues. The
arrays being built can be accessed during the build as nonredundant arrays. The arrays will be redundant when the
build is complete.
6.
RAID 1 arrays are created by copying the data on one drive of the
mirrored pair to the other. If you have specified a RAID 1 array, you
are prompted to select the direction of the copy.
Storage Manager
6-24
If you want to monitor the progress of the build operation, display the
Array Group Information window for the new array group. The build
progress is displayed as a percentage of completion in the Status field. You
can also view the Information window for an array that is a member of a
Parity Group to monitor the progress for that component of the Parity
Group.
Naming an Array Group
Click the Name button in the Array Group Information window to assign a
unique name to an Array Group or Parity Group. This name will be
displayed under the array icon and other locations that the display the
array identifier. The name can be 1 – 13 characters in length.
NOTE
You must restart the host computer before the new array name
will take effect.
Dynamic Array Expansion
NOTE
Dynamic Array Expansion is available only for arrays created
with SmartRAID V controllers running in a Windows NT host
system.
Dynamic Array Expansion allows you to increase your storage capacity by
adding one or more drives to your RAID 0 and RAID 5 arrays while your
system remains online. As additional drives are added, the controller
redistributes the data on the array, placing the new space at the end of the
LSU and increasing the size of the logical drive as seen by Windows NT.
Before this feature was available, to increase the size of an array you had
to backup the data on the array, shutdown the host system, delete the old
array, and then build a larger array that included the new drives. After the
larger array was built, you would reinstall the operating system (if
necessary) and restore the data from the backup. This entire process
could keep your system off-line for at least one day and possibly longer.
Dynamic Array Expansion lets you add the new drives to the array while
the system is active and users are logged in and accessing data. After the
array is expanded, use Windows NT Disk Administrator to add the
additional space to the volume set of the array, then shutdown the system
and restart. When the system restarts, Windows NT recognizes the
additional space as part of the existing logical drive.
Storage Manager
NOTE
6-25
There is a significant impact on system performance during
array expansion. The relative priorities given to system
performance and the array expansion process can be adjusted
by use of the Background Task Priority setting in Storage
Manager.
Dynamic Array Expansion maintains the performance gains of RAID 0
and RAID 5, spreading accesses randomly across the drives in the array.
The DPT controller presents the same number of LSUs to the operating
system after the expansion. There is no need for manual load balancing,
which would be required if the new space were presented as an additional
LSU.
Requirements and Restrictions
Expand refers to the process of adding additional space to an existing
array group using the Dynamic Array Expansion feature. Extend refers to
the process of adding the new space created by the array expansion to the
existing Windows NT volume (LSU).
Dynamic Array Expansion has the following system requirements:
• Windows NT Server or Windows NT Workstation (Version 4.0 and later
with the most recent Service Pack).
• Dynamic Array Expansion requires the I2O OSM supplied by DPT.
The Microsoft-supplied OSM does not support Dynamic Array
Expansion.
• Dynamic Array Expansion and volume extension is supported only for
RAID 0 or RAID 5 array groups using the NTFS file system.
You cannot extend a volume that has a FAT partition. The expanded
space must be added as a separate LSU.
You cannot extend the Windows NT boot partition.
• The array must have Optimal status before beginning the expansion.
If the array status is not Optimal, correct the problem and complete a
rebuild of the existing array before continuing with expansion.
Storage Manager
6-26
NOTE
If a drive fails during expansion, the expansion will complete
successfully. However, the new, larger array will be in a
degraded state until the problem is fixed.
If there is a Hot Spare associated with the array, the degraded
array is rebuilt using the Hot Spare. If there is no Hot Spare,
replace the defective drive, and start a rebuild of the array.
• Each new drive added to an array must be at least equal to the
capacity of the smallest capacity drive already in the array.
In any RAID configuration, the drive with the least capacity in the
array determines the usable capacity of all the drives in the array.
Therefore, there is no advantage in adding a drive with a capacity
larger than the smallest capacity drive already in the array.
Expanding an Array Group
IMPORTANT
DPT recommends that you backup your data before performing an operation
that affects the configuration of a disk array. Do not power off the host system
while the expansion operation is running.
To expand an existing Array Group, do the following:
1.
NOTE
Connect the additional drives to the peripheral bus and power-on
the drives. Refer to the Chapter 3, “Configuration and Installation”
if necessary.
If the drives are not in hot-pluggable carriers, power down the
system before adding drives to the peripheral bus.
2.
Start Storage Manager and click Switch View to change to the
Logical Configuration View.
3.
Select the Array Group to which you want to add drives by clicking
the corresponding RAID 0 or RAID 5 icon.
4.
Click Expand Array Group.
5.
Mark the drives to be added to the array by clicking on them. A
green check mark indicates that a drive is selected.
6.
Click Include Drive. This will cause the marked drives to join the
existing Array Group. The drives to be added are now marked New.
Storage Manager
6-27
7.
When you are finished choosing drives, click Done. The icon for the
Array Group appears with a black flag until the expansion process
is started.
8.
Select File–Set System Configuration to start the array expansion.
The status flag on the Array Group turns blue and the flags on the
components turn white during the expansion process.
You can perform other activity on the system while the expansion
continues, because the array is fully functional during the
expansion process.
NOTE
For large arrays, the expansion can take several hours to
complete. Host I/O activity can prolong the expansion process.
If you want to monitor the progress of the expansion operation, you can
use Storage Manager to view the Array Group Information window. Status
is Expanding during the expansion process, Optimal when it has completed.
Reconfiguring Windows NT After Array Expansion
When the array expansion is complete, do the following to allow Windows
NT to recognize the additional space:
1.
NOTE
Shut down and restart Windows NT.
When restarting, Windows NT will run CHKDSK to verify the
new space. This might take an extended period of time for
large volumes.
2.
Start Disk Administrator. The new space appears as Free Space at
the end of the existing logical drive.
3.
Select both the original logical drive and the free space by
highlighting both segments.
4.
Select Partition–Extend Volume Set...
Click Yes when prompted to the save your changes and to restart
Windows NT.
Use the RDISK.EXE utility to update your emergency repair diskette with
the new disk configuration information.
Storage Manager
6-28
Deleting an Array Group
To delete an Array Group, do the following:
1.
From the Logical Configuration View window, select the LSU or
Array Group icon of the array you want to delete. Then click Delete
Array Group.
2.
Click OK when the confirmation message appears to complete the
delete operation.
Click Cancel to exit without deleting the Array Group.
An array is not physically deleted until you exit Storage Manager and
choose to save changes or select File–Set System Configuration.
NOTE
In a Microsoft Cluster Server configuration, a logical array
that is deleted can continue to appear as available in a remote
session from the other server in the cluster or from a remote
client. The cluster servers must be restarted to ensure that the
deleted array does not continue to be reported as available.
Saving the Subsystem Configuration
The Storage Manager File menu has four options:
Read System Configuration
Causes Storage Manager to read the current hardware configuration.
Any changes that have been made and not saved are lost. This
operation also occurs automatically when Storage Manager is started.
Set System Configuration
Causes Storage Manager to save any changes that you made to the
storage subsystem configuration. If any arrays have been created or
modified, this action causes the controller to start build operations for
the new arrays.
Load Configuration File
Lets you load a previously saved configuration into Storage Manager
and apply it to the current hardware.
Save Configuration File
Lets you save the current configuration, or any changes to that
configuration, to a file for later use. This feature allows storage
subsystems to be configured for other machines with the same size
and type drives, then later loaded from the configuration file.
NOTE
If you reset the NVRAM on the controller, any changes to your
controller parameters return to the factory defaults.
Storage Manager
6-29
Events
Events are generated for detected fault conditions as well as subsystem
status changes. The events are grouped into four categories as follows:
1 – Soft Error
An operation on a disk drive that caused an error but was successful
after a retry.
2 – Recoverable Hard Error
An error on a disk drive, controller or peripheral bus, where the data
was recovered using ECC or from redundant array information.
3 – Nonrecoverable Hard Error
An error on a disk drive, controller or peripheral bus where the data
could not be recovered using ECC or from redundant array
information.
4 – Status Change
The status of an array or drive changed. Examples of this would be a
drive or array failure or an array build or rebuild operation that was
initiated or completed.
Storage Manager
6-30
Event Logging
When events occur, they are automatically logged in the cache on the
SmartRAID V controller where they occurred. In addition, you can specify
that Storage Manager maintain an event log on disk (see Event
Notification). The contents of the event log can be displayed by clicking
Event Log in any controller, drive or array information window. Only the
events pertaining to that controller, drive or array are displayed.
When the Event Log button is selected, the Event Log window appears.
The Event Log window lets you specify that the display be limited only to
events of a specific level or higher (the default is level 4).
NOTE
Some important events may not be displayed by the default
level. You should select one of the higher event levels to ensure
that you are aware of significant errors that might have
occurred.
Click on the arrow buttons to adjust the Display Threshold to the desired
level.
Click View Log to see the event messages whose levels match the selected
levels.
Storage Manager
6-31
Event Notification
If your operating system supports a Broadcaster, Storage Manager lets
you specify that event messages be sent to users, groups, devices, through
e-mail, to the system error log and the DPT Log File. Select Options–
Event Broadcast Control to display the Event Broadcasting dialog. The
options in the Event Broadcasting dialog may vary depending upon your
operating system.
Click Stop Broadcasting to stop Storage Manager from sending event
messages to the locations or addresses specified.
Click Unload Broadcaster to remove the Broadcaster service from the
current set of active services.
6-32
Storage Manager
You can select or modify the following parameters:
Time Interval
Specified time interval at which point the Broadcaster will read the event
logs of all DPT controllers with cache. Any new event messages are
broadcast to each specified destination (based on the Display Threshold
setting).
Broadcast to Computer(s)
Enable this option to send event messages to each system in the
Computer Names list.
Local Desktop Message
Enable this option to have event messages displayed on the local system
desktop.
E-Mail to Users
Enable this option to mail event messages to each e-mail address in the
Email Users list.
Log to Event Viewer
Enable this option to also log and display event messages in the Windows
NT Event Viewer.
Computer Names / Email Users
Use these fields to specify a list of computer systems or e-mail addresses
that are to receive broadcast event messages.
Display Threshold
Click on the arrow buttons to adjust the indicator to the threshold you
want to use for reporting events. Messages will be broadcast for all events
whose levels match those selected.
Storage Manager
6-33
Pager Event Messaging
DPT Storage Manager for Windows NT can also send event messages to
alphanumeric paging devices. The messages are sent as e-mail to pagers
that support e-mail text message delivery.
NOTE:
This feature should work with any alphanumeric paging
system that supports message delivery via e-mail. Contact
your service provider for specific information.
The following procedure applies to the system that is running the Storage
Manager Event Logger to monitor DPT controllers and attached disk
arrays.
To configure alphanumeric pager support, do the following:
1.
Ensure your pager is working and activated through a service
provider.
2.
Establish a permanent connection to an Internet service provider
(ISP) or create Dial-up Networking connection to your ISP.
3.
Configure your e-mail client software to access your Internet mail
server.
4.
Determine the e-mail addresses for the people you want to receive
event messages.
5.
In Storage Manager:
a.
Set the Display Threshold to the level where you want to start
broadcast messages.
b.
Enable E-Mail to Users in the Event Broadcasting dialog
(Options–Event Broadcast Control).
c.
Enter the e-mail addresses for the pagers in the E-Mail Users
list.
d. Click on OK to exit the dialog and save the changes.
When the Event Broadcaster is active, the Event Logger will send text
messages for selected events to the pager using the e-mail address you
specified.
Storage Manager
6-34
Broadcasters
Broadcasters are provided on the DPT diskettes for Windows NT, SCO
UNIX and SCO UnixWare. The Broadcaster collects events logged by DPT
controllers in the host computer on which that broadcaster is running.
The Broadcaster records these events to disk files for each controller for
use by Storage Manager. Additionally, events whose levels are greater
than or equal to the current Broadcast Threshold are sent to destinations
as specified in the Storage Manager Event Broadcasting dialog (see Event
Notification).
Broadcaster for NetWare
A Broadcaster is provided for NetWare. The Broadcaster collects
events from DPT controllers and records them to files in the
SYS:\SYSTEM\DPT directory for use by Storage Manager. Additionally,
events are broadcast to the User/Group list, the system console and the
system error log file as specified in the Storage Manager Event Broadcast
Control window.
Installing the Broadcaster
The DPT diskette for NetWare contains the Broadcaster for NetWare.
When this diskette is installed, the Broadcaster is placed, by default, in the
SYS:\SYSTEM\DPT directory on the server. During Broadcaster
installation, you will be prompted to specify if the Broadcaster is to be
loaded automatically when the server is booted. Doing so will allow events
to be gathered and recorded without user intervention.
Stopping/Restarting the Broadcaster
The Broadcaster is a NetWare Loadable Module (NLM) that can be loaded
or unloaded from the server’s command prompt:
unload DPTELOG
load SYS:\SYSTEM\DPT\DPTELOG
unloads the Broadcaster
loads the Broadcaster
Viewing Events
Events can be viewed on a network workstation through the Event Log
Display window in Storage Manager (see the Event Logging section).
Events can also be viewed by examining the System Error Log file
SYS$LOG.ERR on the server if the Broadcaster has been configured to
send events to that file.
Storage Manager
6-35
Broadcaster for SCO UNIX Systems
DPT provides a Broadcaster for SCO UNIX 3.2V4.2, SCO OpenServer 5
and SCO UnixWare 7. The Broadcaster collects events from DPT
controllers and saves them to files in the /usr/dpt directory for use by
Storage Manager. Events can also be sent to an ASCII file, specified
devices, the User/Group list or e-mailed to users as specified in the Event
Broadcasting dialog.
Installing the Broadcaster
The DPT diskette for SCO contains the Broadcaster for SCO. When you
install the software on this diskette, the Broadcaster is copied to /usr/dpt.
During Broadcaster installation, you are prompted to specify if the
Broadcaster will be loaded automatically when the system is started in
multi-user mode.
If you specify automatic loading, events are gathered and recorded
without user intervention. This also copies a script file, S33dpt, to the /etc/
rc2.d directory. This script automatically loads the Broadcaster as a
background process when the system goes into multi-user mode.
NOTE
Do not use the S33dpt script file for UnixWare 7 Broadcaster
operation.
Stopping/Restarting the Broadcaster
A script file (dptlog) is provided that lets you stop and restart the
Broadcaster from the UNIX prompt. Use the following commands to stop
or start the Broadcaster:
dptlog stop
dptlog start
stops the Broadcaster
restarts the Broadcaster
Viewing Events
Events can be viewed from the system console or a terminal by using the
Event Logs window in Storage Manager (see Event Logging).
Storage Manager
6-36
Broadcaster for Windows NT
The Windows NT Broadcaster collects events from DPT controllers and
records them to files in the \DPTMGR directory for use by Storage
Manager. Additionally, events are sent to the Windows NT Event Viewer.
Installing the Broadcaster
When the DPT installation program runs, the Broadcaster is copied to the
C:\DPTMGR subdirectory by default. The Broadcaster is installed as a
Windows NT service and runs automatically whenever the system is
started. This allows events to be gathered and recorded without user
intervention.
Stopping/Restarting the Broadcaster
Because the DPT Broadcaster is a Windows NT Service it must be
managed from one of the Windows NT service management applications:
• The Broadcaster can be controlled using the Services Control Panel.
Services is accessed by selecting the Services icon in the Control
Panel folder.
Highlight the DPTSERV entry in the Services list. Then click on Start
or Stop to start or stop the Broadcaster service.
• The Broadcaster can also be started or stopped from a command line
prompt by using the NET command as follows:
NET START DPTSERV
NET STOP DPTSERV
starts the Broadcaster
stops the Broadcaster
Viewing Events
Events can be viewed either through the Event Log Display window in
Storage Manager (see Event Logging) or the Windows NT Event Viewer.
The Event Viewer is part of the Windows NT operating system.
To run the Event Viewer, select Event Viewer from the Administrative Tools
group on the Start menu. The Event Viewer lets you view events that have
been placed in the System, Security, and Application logs. To view DPT
events, select Log–Application.
Storage Manager
6-37
This window displays a list of the events submitted to the Application Log
by the DPT Broadcaster, as well as other applications. Events are singleline entries that contain the following information:
Icon
An icon that indicates the Windows NT severity level for the
event. DPT events can be assigned severity levels of Error,
Warning, Information, Operation, or Unknown.
Date
The date the event was logged by the controller.
Time
The time the event was logged by the controller.
Source
The software component that triggered the event. This can be
an application, a component of Windows NT or a device driver.
DPT events can have DPTELOG, DPTENG32 or DPTSCOM in
their Source field.
Category
DPT events can display None, Operation, Warning or a
hexadecimal event code.
Event ID
A number assigned by the Broadcaster to identify the event for
Windows NT.
User
DPT events always display N/A.
Computer
The name of the computer where the event occurred.
Additional information about an event can be obtained by selecting View–
Detail. The Description and Data fields display additional information about
the event. The Description field contains a detailed text description of the
event. The Data field contains the original DPT Event Log data, generated
by the controller. This data might be requested by DPT Technical Support
when troubleshooting problems.
Event information can be saved by the Event Viewer to a file. If the file is
saved in text format then only the event description is saved. Archiving in
Event Log (.EVT) file format saves all event information which can then be
sent to another person for troubleshooting purposes.
Storage Manager
6-38
Formatting Drives
SmartRAID V controllers can perform a low-level format on attached
fixed disk drives. This function is available from the SCSI Device
Information window in Storage Manager. SmartRAID V controllers format
drives in standard 512-byte format.
A low-level format is not normally required before using a disk drive.
However, if a drive has been previously formatted with a different sector
size, it must be reformatted with 512 byte sectors before it can be
recognized by the SmartRAID V controller.
!
IMPORTANT
!
Do not power-down the drive or reboot the computer until after the format
operation is complete. Doing so can cause drives to be left in an unusable
state and require them to be returned to the manufacturer for repair.
To perform a low-level format on a disk drive:
1.
Click Format in the drive’s SCSI Device Information window. The
Format Options dialog appears.
2.
Click Format in the Format Options dialog to start the operation.
3.
You can now exit Storage Manager. The format operation will
continue even though Storage Manager is not running.
To determine if the format has finished, run Storage Manager and look at
the drive’s icon. A blue flag indicates that the format is still in progress. If
the drive icon has no flag, the format is complete.
Storage Manager
6-39
Drive Failures
Drive failures are indicated by flags on the icons associated with the Array
Group, array and individual drives. Failure conditions are indicated as
follows:
A drive belonging to an Array Group fails
The drive icon displays a red flag in both the Physical and Logical
Configuration Views.
The failed drive belongs to a RAID 0 Array Group
The icon for the failed drive displays black flag. Loss of any drive in a
RAID 0 array means the array has failed and data on that array has
been lost.
The failed drive belongs to a RAID 1, 0+1, 5, or 0+5 Array Group
The array icon displays a yellow flag, indicating that the array is
currently running in degraded mode. If two or more drives belonging
to the same array show a red flag, the yellow flag on the array changes
to red, indicating that the array has failed and that data has been lost.
Audible Alarm
DPT controllers with caching capability (Millennium models or Decade/
Century models using an RA4050 module) have an audible alarm. The
failure of a drive which is a member of an array attached to the controller
causes the audible alarm to sound. The alarm stops automatically (after
the initial system scan) when you start Storage Manager or SMOR.
Alarms that occur while Storage Manager is running can be stopped by
selecting Options–Turn Off Audible Alarms.
Storage Manager
6-40
Rebuilding a Degraded Array
NOTE
You can select Rebuild even if the failed drive has not been
replaced and try using the drive again. If the rebuild attempt is
not successful, replace the drive before starting another
rebuild.
When a drive in a RAID 1 or RAID 5 array fails, and that drive is not
protected by a Hot Spare, do the following to restore the array to Optimal
status:
1.
Replace the failed drive according to the procedure in your
hardware documentation.
2.
After the failed drive has been replaced, choose the Logical
Configuration View in Storage Manager.
3.
Double click the array group icon to open the Array Group
Information window.
4.
Click Rebuild in the Array Group Information window to start the
rebuild process.
The drive will display a white flag to indicate that a rebuild operation is in
process. The array and LSU icons will display yellow flags. The percentage
completion of the rebuild operation is displayed in the Array Group
Information window. When the rebuild is complete, the flags will disappear
and the array status should be Optimal.
NOTE
If you are using a DPT RAIDstation storage cabinet with a
SmartRAID V controller, failed drives can be replaced and
rebuilt without using Storage Manager. RAIDstation storage
cabinets and drives have hot swap (SCA-2) connectors so that
failed drives can be removed or replaced without first quieting
the bus. The RAIDstation subsystem will detect that a failed
drive has been physically replaced and the controller will
automatically start a rebuild operation.
Storage Manager
6-41
Hot Spares
To assign a drive as a Hot Spare, click Make Hotspare in the drive’s SCSI
Device Information window. Click Remove Hotspare to reassign an
existing Hot Spare drive as a normal drive.
Any drive that is not assigned to an array can be assigned as a Hot Spare.
Hot Spares are reserved to automatically replace failed drives in RAID 1
or 0+1 and RAID 5 or 0+5 arrays and cannot be accessed by the operating
system for data storage. Hot Spares can only protect drives of equal or less
capacity that are attached to the same controller as the Hot Spare.
When a drive failure occurs in an array protected by a Hot Spare, the
controller automatically starts rebuilding data onto the Hot Spare. During
this process, Storage Manager swaps the positions of the failed drive and
the Hot Spare in the Logical Configuration View. The failed drive will
appear with a red failed flag in the former position of the Hot Spare, and
the Hot Spare will appear as a member of the Array Group with a white
flag indicating that a rebuild operation is in process. The array and LSU
icons appear with yellow (degraded) flags.
When the rebuild is complete, the Hot Spare icon and flags will disappear
and the drive will be displayed as a normal member of the array. The red
flag will remain on the failed drive until that drive is replaced or returned
to Optimal status.
Do the following to replace the failed drive:
1.
Follow the steps in your hardware documentation to remove and
replace the failed drive.
2.
Click Make Optimal in the new drive’s information window.
The new drive will become the a Hot Spare, replacing the previous
Hot Spare that is now a member of the rebuilt Array Group.
NOTE
You can select Make Optimal without replacing the drive if you
want to try using the drive again. If the operation fails, you
should replace the drive.
Storage Manager
6-42
Verify
NOTE
Data inconsistencies should not occur under normal
conditions. However, a power failure that interrupts an array
write operation can cause inconsistencies. Using the Verify
function to make the data consistent does not ensure that
the new consistent data is the correct data.
Running a Verify operation for RAID 1 and RAID 5 arrays ensures that the
redundant information contained in the array is consistent. This operation
is performed by the SmartRAID V controller concurrent with normal
system operation and requires no user or host computer intervention.
To start data verification on an array, select Verify in the Array Group
Information window.
If any inconsistencies in the data redundancy are found, they are made
consistent.
For RAID 1 arrays, the mirrored drive pairs are compared sector by
sector to ensure that both drives contain identical data.
For RAID 5 arrays, parity is recalculated and checked against the stored
parity information.
For RAID 0, only a disk media ECC check is performed.
Running Verify on an array that is in a degraded state can result in
significantly reduced I/O performance. Although there is no impact on the
host CPU, system I/O performance can be affected due to increased
demand on controller resources. If this is a concern, use the Background
Task Priority feature to assign a lower level priority to background
processes.
Storage Manager
6-43
Background Task Priority
You can run rebuild and verification operations without taking an array
offline. These functions are performed as background tasks on the
SmartRAID V controller and are transparent to the host operating system.
However, the controller interleaves I/O from the operating system with I/O
from the background task and this can affect system performance. The
relative priority given to I/O from the operating system and the
background task can be controlled by selecting Options–Background
Task Priority.
The Task Priority section of the dialog assigns the background task
priority for all DPT controllers in the host computer. Set the Priority by
using the arrow buttons to move the indicator . The indicator moves
between Background and Foreground in ten increments.
If the indicator is closer to Background, background tasks are processed
only when there is no disk I/O from the host for a period of at least 250
milliseconds. As the indicator is moved toward Foreground, more of the
controller resources are allocated to background tasks. All of the
resources are allocated to background tasks during periods when there is
no disk I/O from the host system.
Storage Manager
6-44
I/O Statistics
SmartRAID V controllers keep a cumulative record of I/O operations in
cache RAM for use in analyzing the efficiency of the storage subsystem.
You can view this data by selecting I/O Stats in the Information window for
any, controller, drive or array. By analyzing these statistics, the array
configuration, cache and stripe size can be optimized for your particular
system configuration.
Controller I/O Statistics
Click the I/O Stats button in the Host Bus Adapter Information window to
see cache statistics for that controller. These statistics include:
Cache Statistics
Total Pages
The total number of pages contained in the controller
cache.
Used Pages
The number of pages that currently contain disk data.
Dirty Pages
The number of pages that contain dirty data.
Read-Ahead Pages
The number of cache pages that contain data that has
been loaded from disk as a result of read-ahead
functions.
ECC Fault Pages
The number of pages that have been mapped for
non-use after a RAM fault was discovered by the
controller ECC feature
Storage Manager
6-45
Command Statistics
Total Comm ands
The total number of commands received from
the host computer. This includes read and
write commands and other commands that
may not involve device I/O.
M isaligned Transfer
The number of commands that required data
to be transferred starting at a RAM address
location that was not an even byte value.
SCSI Bus Resets
The total number of SCSI bus resets that have
been issued by the controller. A large number
of resets can indicate a problem with the SCSI
bus or an attached device.
Device I/O Statistics
Write-Backs + Write-Throughs = the total number of sectors written to disk
by the controller.
Storage Manager
6-46
I/O Commands
The number of read and write commands issued
by the computer to the controller. If you use RAID
and caching, the number of commands issued to
the drives can differ significantly from this value.
Total Sectors
The total number of sectors read and written
from the host computer.
The total number of sectors (read or written from
the host computer) that were found in the
controller cache and did not require a disk access.
Cache Hits
Read-Ahead Hits
Write-Backs
Write-Throughs
The number of the cache hits for data read
requests satisfied by data held in the cache from
previous disk read-ahead operations.
The number of sectors written to disk that were
held in the controller cache and written some
time after the host write command reported as
completed.
The number of sectors written directly to disk
before the write command ended.
In RAID 1 arrays, (Write-Backs + Write-Throughs) can be up to twice the
value of Total Sectors because each sector written from the host results in
a write to each mirrored disk. In RAID 5 arrays, each write from the host
can generate up to two disk reads and two disk writes. If the controller has
cache memory, this can result in the actual number of sectors read from
or written to disk being reported as less than this value.
Environments with a large number of sequential reads should generate a
high number of Read-Ahead Hits relative to Total Sectors. These hits
reduce the number of seek operations and increase performance. The
read-ahead hits count can be increased by adding more cache memory to
the controller. A high percentage of 4KB or smaller I/O operations also
indicates that increasing the controller cache would be beneficial.
Storage Manager
6-47
If cache hits are low, adding more cache RAM can also increase the hit
count. Systems with a large number of disk writes also derive significant
performance benefits from the controller cache.
For RAID 0 and RAID 5 arrays, the default stripe size is set for optimal
performance in most environments. However, some specific environments
can benefit from smaller or larger stripes. For example, when a RAID 5
write operation accesses all drives there will be a large number of stripe
crossings. By using a larger stripe size, the RAID 5 parity data can be
generated more efficiently. This results in better write performance.
However, if the write crosses one or more stripes but does not involve all of
the drives, the performance will be less efficient with the larger stripe size.
You should choose the stripe size relative to both the I/O segment size and
the number of drives in the array, so that most I/O operations either:
• Do not cross stripes and involve only a single drive
OR
• Cross many stripes and involve all the drives in the array.
6-48
Storage Manager
Remote Communication
The Storage Manager Remote Communication feature lets you use
Storage Manager running on your local workstation to manage remote
server systems that contain DPT controllers. The Available Connections
window shows the types of connections you can use and any predefined
connections you have created.
6-49
Storage Manager
Communicating Across a Network
Storage Manager can run as a client/server application across a network
using a TCP/IP connection. The Storage Manager client runs on a
supported workstation operating system and connects to one of the
supported networked servers running the DPT communication engine.
The following illustration shows the supported workstation and server
connections for a TCP/IP network.
Client Workstations
Windows
95/98
Windows
NT
SCO
TCP/IP
Windows
95/98
Windows
NT
NetWare
SCO
Servers
Installation and Configuration
Remote Communication uses both workstation-based and server-based
software components. The workstation (client) remote communication
software is automatically installed during Storage Manager setup. You are
given the option whether to install the Communication Server software
when you run the Storage Manager setup program on an operating
system that supports a DPT communication server (Windows 95/98,
Windows NT, NetWare or SCO platforms).
Storage Manager
6-50
You must configure Remote Communication before it can be used. For
both workstations and servers, this is done by editing the DPT
configuration file (DPTMGR.INI). There may be additional server
configuration steps, depending on your operating system. You must have
physical connections between the servers and client workstations.
NOTE
Under Windows 95/98, the server must be started manually by
selecting the DPT Communication Server icon. This version of
the communication server runs in a DOS window.
Editing the Configuration File
Both the workstation and server communication engines use the DPT
configuration file DPTMGR.INI. This file is contained in the directory in
which you installed Storage Manager.
The [OPTIONS] section of the file controls the behavior of the
communication server. To control the display of messages from the server,
change the Verbose parameter:
[OPTIONS]
Verbose=x
where x is:
0 No messages
1 Basic messages (such as errors, connects and disconnects)
2 More messages (option 1 messages + socket numbers and TCP/IP
addresses)
3 All messages (option 2 messages + message tracing)
The default is 1.
The [MODULES] section of the file tells the communication server the
protocol you want to use for communication.
[MODULES]
TCP
(Specify as many as needed)
(Comm server will use TCP protocol)
Each protocol can have its own section, in which you can set additional
parameters specific to the protocol. The following list shows all supported
options. The value listed is the default; optional values are in parentheses.
[TCP]
(Available options for TCP protocol)
SOCKET=2091
(TCP socket on which to listen
Can be any valid socket number)
Storage Manager
6-51
Setting up the Server
If you are using an operating system that supports the DPT
communication server and selected the Communication Server option
during Storage Manager setup, the server was automatically installed at
that time. Additional steps may be required to complete the installation,
depending on your operating system. Some operating systems (such as
Windows NT) require setting up a user name and/or password before the
server can be accessed. The default password is password.
Windows NT Server
Under Windows NT, the server is installed as a service named DPT
Communication Service and starts automatically when Windows NT
starts.
Windows NT server access requires both a user name and password. The
user name must be defined through the Windows NT User Manager.
Additionally, the user name must have Administrator level privileges.
Setup the user account for use by performing the following steps:
1.
Log in with Administrator privileges.
2.
From the Start menu select Programs–Administrative Tools–User
Manager.
3.
Select Policies-User Rights and check the Show Advanced User
Rights box when the User Rights Policy dialog appears.
4.
Grant the Act as part of the operating system and Log on as batch
job rights to the Administrators group.
5.
Reboot Windows NT before you attempt to use Remote
Communication.
Windows 95/98
Under Windows 95/98, the communication server is installed as a DOS
command line application. You can start the communication server by
clicking on DPT Communication Server in the DPT folder on the Start
menu.
The Windows 95/98 server is accessed by a password only; a user name is
not required. The initial default password is password. To change the
password, run Change Communication Password from the DPT folder in
the Start menu.
Storage Manager
6-52
SCO UNIX
Under SCO UNIX, a communication server entry is placed into the rc.d
files, which causes it to be started when the system is booted into multiuser mode.
During installation, the file dptcom.chk is created in the directory into
which Storage Manager is installed. The ownership of this file is set to
root/dtadmain, and permissions are -rw-rw---(6608).
Access to SCO servers requires both a user name and password. After
determining that the user name and password are valid on the system on
which it is running, the communication server will attempt to open
dptcom.chk using the user name.
Connecting Servers and Workstations
The Storage Manager client must have a valid network connection to
connect to a remote server. Use the following guidelines to ensure that the
workstation and server can communicate.
Servers and workstations communicating by TCP/IP must be
connected by a local area network or the Internet. Some operating
systems may require additional TCP/IP protocol configuration before
use. Refer to your operating system documentation for more
information.
The client workstation specifies the TCP/IP address of the server,
along with a user name and password as required by the server.
Storage Manager
6-53
Connecting to Remote Systems
Select Communication–Make Connection to use Storage Manager to
manage DPT hardware in a remote server system using the remote
communication feature. This menu item displays the Available
Connections window.
NOTE:
You can also use the IP=address command line parameter
when you start Storage Manager. This parameter causes
Storage Manager to connect to the remote system at the
specified IP address instead of the default Local connection.
From this window, you can make a connection either by selecting an
available protocol under Manual Connections and entering the name,
address and password of the server; or by selecting an entry under
Custom Connections. Custom connections are those for which you
previously stored address information in the Phonebook (see Using the
Phonebook for more information).
6-54
Storage Manager
Making a Manual Connection
Under the Manual Connections section of the Available Connections
window, there is an icon the TCP/IP protocol that you configured in the
DPTMGR.INI file, and an icon for the system you are using (Local).
To make a connection to a remote system, click the icon that represents
the protocol for that system. (To select the system from which you are
running Storage Manager, select the Local icon). For icons except Local,
the Connection Information dialog is displayed. Enter the address, user
name (if required) and password for the server selected. Refer to
Connecting Servers and Workstations for more information.
When you click OK, Storage Manager attempts to connect to the remote
system. If the connection is successful, Storage Manager scans the DPT
hardware on the remote system. When the hardware scan is complete, the
Physical Configuration View window for the remote system appears.
Storage Manager
6-55
Using the Phonebook
You can save the server name, address, user name and protocol of
systems you frequently access in the Phonebook. After you have entered
data for a system, you can place an icon under Custom Connections to
make future connections to that system without having to reenter the
connection information each time.
To make an entry in the Phonebook, fill in the appropriate fields.
Select Display icon for this entry if you want an icon for this connection
displayed in the Available Connections window. You can customize the icon
to be used by clicking Choose Icon.
Click Add to add the new entry. Click Delete to remove unwanted entries.
When you are finished, click Save to save your changes.
To make a connection from within the Phonebook, highlight the desired
system and click Connect.
Storage Manager
6-56
Using a Custom Connection
Under the Custom Connections section of the Available Connections
window, there is one icon for each system that you marked for display in
the Phonebook.
To make a connection to a remote system, double-click the icon. The
Connection Information dialog appears. Enter the password of the server
selected. Refer to Connecting Servers and Workstations for details.
When you click OK, Storage Manager will attempt to connect to the remote
system. If the connection is successful, Storage Manager scans the DPT
hardware on the remote system. After the scan is complete the Physical
Configuration View window for the remote system appears.
NOTE:
You can also use the IP=address command line parameter
when you start Storage Manager. This parameter causes
Storage Manager to connect to the remote system at the
specified IP address instead of the default local connection.
CHAPTER 7:
Theory of Operation
This chapter provides background and
reference information about:
Chapter Topics:
• Intelligent RAID Controllers
• Operation of SmartRAID V controllers
• Pipelined Parallel Processing
• Disk array configuration
• Caching Algorithms
• SCSI device interfaces and cabling
• RAID Levels
• Controller Architecture
• I2O Messaging
• PCI Data Transfer
• On-board Microprocessor
• SCSI Interface
• Cabling and Termination
• Fibre Channel Arbitrated Loop
• I2O BIOS
7-1
7-2
Theory of Operation
Intelligent RAID Controllers
File servers achieve maximum throughput when the performance of their
component subsystems is optimized for the CPU bandwidth. If any
component is not capable of supplying data at an optimum rate, that
component will restrict the data path performance and slow down the
entire system. Conversely, a component with significantly greater
performance than the rest of the system can have much of its
performance potential unused. Any attempt to speed up a server by
increasing the performance of only some system components usually
results in a negligible overall increase.
The speed of CPUs used in currently available file servers has increased
greatly over those available a few years ago. However, the performance of
data storage subsystems has not kept up. As a result, many servers have
data storage subsystems that are not capable of supplying data to the CPU
fast enough. Consequently the CPU can be idle for significant periods
waiting for data and these servers cannot perform at their full potential.
In a traditional server, the host CPU handles the processing of I/O
interrupts from peripherals, disk storage and network devices. This design
was developed for the single-user and non-multitasking operating systems
which were in use when microcomputers first became widely available.
However, this design becomes a performance restriction when
incorporated into graphics workstations or multi-user servers that have
pipelined-architecture processors. On a busy server, the large amount of I/
O interrupts sent to the host CPU reduces the number CPU clock cycles
available to process application software instructions. Additionally, each
interrupt causes the CPU internal pipelines and caches to flush their data,
which results in additional CPU performance degradation.
Theory of Operation
7-3
Intelligent I/O
To maximize the performance of a busy server requires a data storage
subsystem that incorporates intelligent I/O. Such a subsystem would have
a microprocessor embedded in the I/O and storage subsystem that is
dedicated to handling interrupt requests and I/O processing functions in
place of the host CPU. Intelligent I/O frees the CPU to process user
application software and requests, which significantly increases data
throughput and overall system response. DPT SmartRAID V controllers
implement intelligent I/O using an onboard RISC microprocessor.
Benefits
Controllers that depend on the host CPU for processing low-level I/O
functions can require a significant portion of the total processing power of
the system: from 25% to nearly 100%, depending on the task. Nonintelligent RAID controllers (that is, controllers without embedded I/O
processors) also use the host CPU for RAID functions such as, striping
and parity calculations, RAID-generated disk interrupts, and read and
write data transfers. Intelligent controllers incorporating hardware-based
RAID relieve the CPU from handling of these additional processes.
Intelligent host adapters and RAID controllers can provide the following
benefits:
Optimum server performance – An I/O subsystem with performance
matching the current powerful host CPUs allows servers to reach
their maximum application processing potential.
I/O subsystem scalability – Server scalability through symmetric
multiprocessing (SMP) can be matched with the installation of
multiple intelligent controllers or by having multiple data channels per
intelligent controller.
A greater number of users per server – With servers free to process
more applications with greater bandwidth, more users can be
assigned to each server.
Increased end-user productivity – With more CPU cycles devoted to
servicing applications and user requests, system performance is
improved yielding faster response time.
Reduced total cost of ownership – With improved performance,
servers can support more users and fewer servers may be required to
meet the overall needs of the organization. In addition, fewer servers
and their support structure may be needed in the future.
Theory of Operation
7-4
I2O and Intelligent RAID Controllers
Several information technology industry leaders have formed a Special
Interest Group (SIG) to promote intelligent I/O standards for the entire
industry. These standards, referred to as the Intelligent I/O (I2O)
specification, address the performance issues caused by I/O interrupts
sent to the host CPU and the need for unique drivers for every
combination of I/O device and operating system.
I2O-compliant products include both intelligent host adapters and RAID
controllers (those incorporating a microprocessor) and non-intelligent
host adapters and RAID controllers (those without a microprocessor).
Maximizing a server’s potential requires more than just I2O. It requires an
intelligent adapter or controller with a dedicated microprocessor.
Intelligent I/O (both intelligent adapters and RAID controllers) has been
available from DPT for two decades. The I2O-compliant SmartRAID V
controllers are the latest intelligent adapter technology, with fast RISCbased embedded I/O processors and our new Pipelined Parallel
Processing (P 3) technology.
The following illustration depicts the two-part device driver architecture
used by I2O compliant products. This design allows for much faster device
driver development for specific operating system platforms. For additional
information, see the SmartRAID V Architecture section in this chapter.
I 2O Split Driver Architecture
HOST
OSM
OS Specific
OS Revision
Independent
I/O Processor
Communication Layer
HDM
OS Independent
I/O Software
Hardware
7-5
Theory of Operation
Pipelined Parallel Processing (P3)
Traditional RAID controller designs execute host commands sequentially.
Each command initiates a sequence of hardware activities that are
processed one after the other until the command is complete. If any of
these activities are delayed, hardware components must wait. These
delays compromise performance. DPT’s P3 technology addresses this
problem in five important areas:
• Real-Time Control System – P 3 technology can process a virtually
unlimited number of host commands in parallel. The SmartRAID V
Real-Time Control System (RTCS) breaks each command down into
sub-tasks that are then routed to one of the following hardware
processors on the controller:
– DMA processor
– XOR processor
– SCSI or Fibre Channel interface.
XOR Engine
DMA Engine
SCSI / Fibre Channel
Interface
Controller CPU
Host Commands
• Hardware Request Schedulers – Each hardware processor has an
associated hardware request scheduler that coordinates its queued
requests and manages the queue so that all subtasks are processed
efficiently. The disk hardware request scheduler manages one request
queue for each disk drive attached to the controller. To achieve
maximum throughput from each drive, sorted disk queues are
maintained by grouping requests when possible. This results in an
increased number of read and write operations per drive.
Theory of Operation
7-6
• Hardware RAID Processor – DPT’s P 3 architecture is based on a new
DPT-exclusive ASIC. This specially designed integrated circuit provides
memory-to-memory and disk-to-memory data transfers that are fully
concurrent with controller CPU execution. Depending on whether the
PCI bus is 32-bit or 64-bit, data transfers occur at 132 or 264 MB/sec data
rates, respectively. Independent DMA and XOR processors execute
Scripts-based programs, enabling DMA and XOR tasks to be chained
and executed without the intervention of the controller CPU. The
hardware RAID processor also generates 128-bit parallel errorcorrecting code (ECC) for the controller cache memory. This powerful
ECC is capable of correcting an error of up to 32 bits for each 512-byte
block of data.
• Intelligent Caching – Because most popular operating systems are
heavily cached, a SmartRAID V controller does not cache data read by
the host. Instead, the controller cache is used for predictive and writeback caching and provides groups write data for efficient RAID 5
operation. For those environments where read caching on the controller
will improve performance (particularly those without software disk
caching), read caching can be enabled using Storage Manager.
SmartRAID V controllers also incorporate predictive caching
algorithms. These algorithms are described in the Software and
Hardware Caching section.
• Virtual Cache – In addition to normal caching functions, SmartRAID V
controllers also maintain a large virtual cache used for pattern
recognition and prediction. The total amount of cache space supported
by a SmartRAID V controller is fixed. Some of this total cache space is
used when physical memory modules are installed on the controller to
be used for standard cache. The space not used for standard cache is
available for virtual caching. SmartRAID V Century controllers support
a total cache space of 128MB with up to 64MB of physical cache
(depending on the number of memory modules installed). SmartRAID V
Millennium controllers support a total cache space of 512MB with up to
256MB of physical cache. As an example, if 128MB of cache memory
were installed on a SmartRAID V Millennium controller then
(512 – 128)MB = 384MB of virtual cache would be available.
Disk Caching
SmartRAID V Millennium controllers can be configured with up to 256MB
of cache RAM. SmartRAID V Century controllers (equipped with an
RA4050 RAID Accelerator) can be configured with up to 64MB of cache
RAM. Adding more cache to a SmartRAID V controller can result in better
RAID performance – especially for RAID 5.
7-7
Theory of Operation
Software and Hardware Caching
The SmartRAID V cache is designed to enhance performance in
environments that are already software cached, such as UNIX, Novell and
Windows NT. Although software caches are beneficial, hardware caching
offers a significant advantage in larger, heavily loaded systems when
combined with the software cache for the operating system.
The purpose of the operating system cache is to minimize disk reads. If
data requested by an application is found in the operating system cache,
a disk read is eliminated. A software cache can also improve application
response time by immediately accepting data to be written to disk from
application programs and postponing the actual writes until the disk is
idle.
LIGHTLY LOADED SYSTEM
Without Hardware Cache
Application programs write to OS cache:
Bus
activity:
Idle
time
Idle
time
OS cache flushes to disk (~11ms per write)
Although write caching through the operating system can benefit a lightly
loaded system, it can result in performance degradation in large multiuser environments. The greater I/O demands in such systems reduce the
periods of system idle time required to flush the operating system cache to
disk. As disk activity becomes more intense, the periods of system idle
time become shorter and less frequent. The operating system cache fills
up with dirty sectors waiting to be written to disk, which take up space that
could be used for more recently requested data. When cache RAM is
added to a SmartRAID V controller, the operating system cache buffers
must still be flushed. However, the controller receives and caches the
flushed data in a fraction of the time required to actually write to a disk.
The controller can also write the data back to disk in an elevator-sorted
order without interrupting other system activity.
Theory of Operation
7-8
HEAVILY LOADED SYSTEM
HEAVILY LOADED SYSTEM
With Hardware Cache
Without Hardware Cache
Application programs write to OS cache:
Bus
activity:
Idle
time
Idle
time
Application
programs
write to
OS cache
System Bus
activity
Idle
time
Idle
time
OS cache flushes
to SmartRAID
(1ms per write)
SmartRAID
Bus
OS cache flushes to disk (~11ms per write)
SmartRAID
elevator-sorted writes to disk
(1-5ms per write)
As an example, consider an operating system with 4MB of dirty data in its
cache buffers. With a 4KB record size and an average disk I/O time of
11ms per record, this flush operation would take 11 seconds. However, with
controller cache installed the entire flush operation from the software
cache would take approximately 1 second instead of 11 seconds. The
SmartRAID V controller can then flush its cache to disk in elevator-sorted
order using its separate I/O bus, concurrently with other system activity.
Elevator Sorting
SmartRAID V cache flush operations occur much faster than from the
operating system cache, because all data in the controller cache is kept in
elevator sorted order. Data is written to disk in order of increasing block
address, minimizing both physical disk seeks and rotational latency. This
feature, called Elevator Sorted Write-back, eliminates much of the head
thrashing commonly associated with disk-intensive operating systems like
Novell NetWare, Windows NT and UNIX. By elevator sorting the sectors in
cache, the average write access time to the drive may be reduced from the
normal 11ms (seek + rotation) time, to between 1 and 5ms.
A good analogy is an elevator in a multistory building. Consider how much
slower the elevator would be if it went to the requested floors in the same
sequence that the buttons were pressed, instead of in floor number
sequence! An additional benefit of elevator sorting is that multiple short
disk write operations can often be coalesced within the controller cache
into one large write operation, significantly reducing peripheral bus
overhead.
7-9
Theory of Operation
DISK WRITES
DISK WRITES
Without Elevator Sorting:
11ms average
disk write
With Elevator Sorting:
1-5ms average
disk write
4
65
4
65
93
11
93
17
31
11
17
31
27
27
Caching
SmartRAID V controllers can operate in write-back or write-through
mode and also use a predictive caching algorithm to optimize disk I/O for
the host system. This section describes the methods and rules used by
SmartRAID V controllers to cache data for host I/O operations.
Write-back and Write-through Algorithm
Write-back is a method of postponing data to be written to a peripheral
device, such as a disk drive, by temporarily saving the data in a cache. The
data can then be written at a later time when the device would otherwise
be idle. In Write-back mode, writes to disk are deferred until after
command completion. This type of caching generally provides better I/O
performance.
In addition, the data can be processed by techniques such as elevator
sorting, to increase write operation performance.
In Write-through mode, all data is written immediately to disk for each
write command before command complete status is returned to the host.
The data may also be held in cache for subsequent read requests.
NOTE
When you first install a PM3755U2B controller with a BB4050
module, automatic low-battery write-through mode is disabled.
Use the Battery Backup Configuration dialog in Storage
Manager to set a threshold for when write-through mode
should be used if the battery backup capability drops below a
predetermined level.
Theory of Operation
7-10
Flush Strategy
DPT SmartRAID V controllers continually flush all dirty cache pages to
disk, with bursts occurring every second and increasing to 4 times per
second as the number of dirty pages increases. The data is written to disk
in elevator sorted order. The cache flushing routine establishes a flush
rate that writes all the dirty pages to disk within three minutes. If there is
no host activity, the cache flush will occur in less than three minutes.
Predictive Caching Algorithms
SmartRAID V predictive caching dramatically increases disk read
performance in computer systems with highly fragmented file structures.
The SmartRAID V pattern recognition algorithms are capable of
predicting disk read requests from the host computer by recognizing
patterns within seemingly random disk accesses. The SmartRAID V
pattern recognition algorithms not only look for hidden sequential strings,
but also analyze historical read/write patterns to determine when the host
is accessing files that were previously paged out of host cache.
Virtual Cache
SmartRAID V controllers use virtual cache to track the patterns of data
access for the most recently accessed disk data without actually caching
that data. The collected information can then be used for more intelligent
predictive algorithms. Additionally, the use of virtual cache enables the
controller to track how many hits would have resulted from installing
more controller or host cache. Virtual cache does not actually use RAM.
Read Ahead
Because disk data is often grouped as contiguous sectors on the disk, it is
sometimes beneficial to continue to read sequential sectors into cache
following a sector that has been requested by the computer. The
SmartRAID V controller uses algorithms that analyze the pattern of disk
access, looking for cases where sequential I/O may be occurring. If it is
determined that a read command from the computer is part of a pattern of
sequential reads, the controller reads additional sequential sectors into
cache so that future read commands will result in cache hits.
7-11
Theory of Operation
Pre-Fetch
The SmartRAID V pre-fetch algorithm analyses disk access patterns to
detect whether the host is reading data that it previously read. By using
records stored in virtual cache, the caching algorithm determines the
boundaries of the contiguous block of previously read data surrounding
the current host read request, then reads that data block into cache.
Predictive Caching in Multi-user Systems
Disk read-ahead is effective in a multi-user environment when a large
number of read-ahead sectors from many different areas of the disk are
cached. This is because multi-user and networking operating systems
contain highly fragmented files and time-slice I/O requests from many
different users or tasks. Data that has been cached by a read-ahead
operation may not be accessed until after other queued commands are
processed and the same user or task has another chance to access the
disk.
In addition, patterns of sequential I/O are harder to detect in large multiuser systems because they are typically interleaved with I/O from many
users. This is why the read-ahead FIFO buffer on a drive (which can store
only a limited number of data segments) provides acceptable performance
in a single-user system, but must be augmented by the larger cache on the
SmartRAID V controller in a system with multiple users.
AUTOMATIC READ-AHEAD
IS PERFORMED BY DRIVE
Read data
Command
terminates
here
Read-Ahead data
Read-Ahead data is stored
in FIFO buffer on SCSI drive.
READ-AHEAD SEGMENTS
ARE CACHED IN BACKGROUND
HARDWARE
DISK
CACHE
40MB/s Peripheral Bus
Stores
thousands of
Read-Ahead
segments
DRIVE
DRIVE
DRIVE
DRIVE
Read
Ahead
FIFOs
Theory of Operation
7-12
Caching Host Reads
Because there is no advantage in using controller cache to duplicate the
operating system cache, SmartRAID V controllers with cache only cache
predictive and dirty data. Read cache misses are transferred directly from
disk to host, bypassing the controller cache.
Optimizing Operating System Cache Size
When configuring a server, you should always consider the effect of the
operating system cache on the disk subsystem. Although many disk read
operations can be eliminated by the operating system cache, all writes
must get through to the disk. In effect, the operating system cache is
acting as a read filter for the disk subsystem.
As the operating system cache size is increased, more read hits are
serviced from the operating system cache resulting in fewer reads issued
to the disk subsystem. However, the number of disk writes stays relatively
constant. At some point, adding more cache to the operating system will
not result in significantly reduced disk I/O. At this point it is better to add
more cache to the controller to improve write throughput.
NOTE
Most operating systems that have caching capability can
automatically manage the size of their disk cache.
Optimizing Controller Cache Size
The SmartRAID V cache is used for predictive reads, elevator sorted
write-backs and as temporary storage during RAID 5 parity calculations.
Increasing the size of the cache makes these features more efficient. More
cache RAM allows the controller to hold and elevator-sort more records,
resulting in closer spacing between consecutive write-back segments. This
results in lower average access time for disk write operations and higher
disk throughput.
The SmartRAID V cache can be expanded incrementally as needed to
maintain optimum system performance. More cache is typically required
for systems with many active users than for single-user systems. Systems
with large data files also benefit from a larger cache. However, a system
that has a large number of users who all access the same data, will require
less cache than the previous examples.
NOTE
DPT recommends a minimum of 16MB of cache for
SmartRAID V controllers.
Theory of Operation
7-13
RAID
The basic idea of RAID (Redundant Array of Independent Disks) is to
combine multiple inexpensive disk drives into an array of disk drives to
obtain performance, capacity and reliability that exceeds that of a single
large drive. The array of drives appears to the host computer as a single
logical drive.
The Mean Time Between Failure (MTBF) of the array is equal to the
MTBF of an individual drive, divided by the number of drives in the array.
Because of this, the MTBF of a non-redundant array (RAID 0) is too low
for mission-critical systems. However, disk arrays can be made faulttolerant by redundantly storing information in various ways.
Five types of array architectures, RAID 1 through RAID 5, were originally
defined, each provides disk fault-tolerance with different compromises in
features and performance. In addition to these five redundant array
architectures, it has become popular to refer to a non-redundant array of
disk drives as a RAID 0 array.
Disk Striping
Fundamental to RAID technology is striping. This is a method of
combining multiple drives into one logical storage unit. Striping partitions
the storage space of each drive into stripes, which can be as small as one
sector (512 bytes) or as large as several megabytes. These stripes are then
interleaved in a rotating sequence, so that the combined space is
composed alternately of stripes from each drive. The specific type of
operating environment determines whether large or small stripes should
be used.
Most operating systems today support concurrent disk I/O operations
across multiple drives. However, in order to maximize throughput for the
disk subsystem, the I/O load must be balanced across all the drives so that
each drive can be kept busy as much as possible. In a multiple drive
system without striping, the disk I/O load is never perfectly balanced.
Some drives will contain data files that are frequently accessed and some
drives will rarely be accessed.
Theory of Operation
7-14
STRIPING DISK DRIVES
Disk 1
1
5
9
13
Disk 2
2
6
10
14
Disk 3
3
7
11
15
Disk 4
4
8
12
16
Data stripes from each drive
are interleaved to create one logical drive.
By striping the drives in the array with stripes large enough so that each
record falls entirely within one stripe, most records can be evenly
distributed across all drives. This keeps all drives in the array busy during
heavy load situations. This situation allows all drives to work concurrently
on different I/O operations, and thus maximize the number of
simultaneous I/O operations that can be performed by the array.
Definition of RAID Levels
RAID 0 is typically defined as a group of striped disk drives without parity
or data redundancy. RAID 0 arrays can be configured with large stripes for
multi-user environments or small stripes for single-user systems that
access long sequential records. RAID 0 arrays deliver the best data
storage efficiency and performance of any array type. The disadvantage is
that if one drive in a RAID 0 array fails, the entire array fails.
RAID 0
Non-Redundant Striped Array
Writes can occur simultaneously on every drive.
DATA
DATA
DATA
DATA
DATA
DATA
Reads can occur simultaneously on every drive.
7-15
Theory of Operation
RAID 1, also known as disk mirroring, is simply a pair of disk drives that
store duplicate data but appear to the computer as a single drive. Although
striping is not used within a single mirrored drive pair, multiple RAID 1
arrays can be striped together to create a single large array consisting of
pairs of mirrored drives. All writes must go to both drives of a mirrored
pair so that the information on the drives is kept identical. However, each
individual drive can perform simultaneous, independent read operations.
Mirroring thus doubles the read performance of a single non-mirrored
drive and while the write performance is unchanged. RAID 1 delivers the
best performance of any redundant array type. In addition, there is less
performance degradation during drive failure than in RAID 5 arrays.
RAID 1
Mirrored Arrays
Duplicate data is written to pairs of drives.
DATA
DATA
DATA
DATA
DATA
DATA
Reads can occur simultaneously on every drive.
RAID 2 arrays sector-stripe data across groups of drives, with some
drives assigned to store ECC information. Because all disk drives today
embed ECC information within each sector, RAID 2 offers no significant
advantages over other RAID architectures and is not supported by
SmartRAID V.
RAID 2
Parallel Array with ECC
Each write operation spans all drives.
DATA
DATA
DATA
DATA
ECC
ECC
Each read operation spans all drives.
ECC
7-16
Theory of Operation
RAID 3, as with RAID 2, sector-stripes data across groups of drives, but
one drive in the group is dedicated to storing parity information. RAID 3
relies on the embedded ECC in each sector for error detection. In the case
of drive failure, data recovery is accomplished by calculating the exclusive
OR (XOR) of the information recorded on the remaining drives. Records
typically span all drives, which optimizes the disk transfer rate. Because
each I/O request accesses every drive in the array, RAID 3 arrays can
satisfy only one I/O request at a time. RAID 3 delivers the best
performance for single-user, single-tasking environments with long
records. Synchronized-spindle drives are required for RAID 3 arrays in
order to avoid performance degradation with short records. Because
RAID 5 arrays with small stripes can yield similar performance to RAID 3
arrays, RAID 3 is not supported by SmartRAID V.
RAID 3
Parallel Array with Parity
Read and write operations span all drives.
DATA DATA
DATA
DATA DATA
PARITY
Parallel access decreases data transfer time
for long sequential records.
RAID 4 is identical to RAID 3 except that large stripes are used, so that
records can be read from any individual drive in the array (except the
parity drive). This allows read operations to be overlapped. However, since
all write operations must update the parity drive, they cannot be
overlapped. This architecture offers no significant advantages over other
RAID levels and is not supported by SmartRAID V.
RAID 4
Parallel Array with Parity
Every write must update dedicated parity drive.
DATA
DATA
DATA
DATA
PARITY
Reads can occur simultaneously
on every data drive.
7-17
Theory of Operation
RAID 5, sometimes called a Rotating Parity Array, avoids the write
bottleneck caused by the single dedicated parity drive of RAID 4. Under
RAID 5 parity information is distributed across all the drives. Since there
is no dedicated parity drive, all drives contain data and read operations
can be overlapped on every drive in the array. Write operations will
typically access one data drive and one parity drive. However, because
different records store their parity on different drives, write operations
can usually be overlapped.
RAID 5
Striped Array with Rotating Parity
Writes require parity to be updated.
DATA/
PARITY
DATA/
PARITY
DATA/
PARITY
DATA/
PARITY
DATA/
PARITY
DATA/
PARITY
Reads can occur simultaneously on every drive.
Theory of Operation
7-18
In summary:
• RAID 0 is the fastest and most efficient array type but offers no faulttolerance. RAID 0 requires a minimum of two drives.
• RAID 1 is the best choice for performance-critical, fault-tolerant
environments. RAID 1 is the only choice for fault-tolerance if no more
than two drives are used.
• RAID 2 is seldom used today since ECC is embedded in all hard drives.
RAID 2 is not supported by SmartRAID V.
• RAID 3 can be used to speed up data transfer and provide faulttolerance in single-user environments that access long sequential
records. However, RAID 3 does not allow overlapping of multiple I/O
operations and requires synchronized-spindle drives to avoid
performance degradation with short records. Because RAID 5 with a
small stripe size offers similar performance, RAID 3 is not supported
by SmartRAID V.
• RAID 4 offers no advantages over RAID 5 and does not support
multiple simultaneous write operations. RAID 4 is not supported by
SmartRAID V.
• RAID 5 combines efficient, fault-tolerant data storage with good
performance characteristics. However, write performance and
performance during drive failure is slower than with RAID 1. Rebuild
operations also require more time than with RAID 1 because parity
information is also reconstructed. At least three drives are required
for RAID 5 arrays.
Dual-Level RAID
In addition to the standard RAID levels, SmartRAID V controllers can
combine multiple hardware RAID arrays into a single array group or
parity group. In a dual-level RAID configuration, the controller firmware
stripes two or more hardware arrays into a single array.
NOTE
The arrays being combined must both use the same RAID level.
Dual-level RAID achieves a balance between the increased data
availability inherent in RAID 1 and RAID 5 and the increased read
performance inherent in disk striping (RAID 0). These arrays are
sometimes referred to as RAID 0+1 or RAID 10 and RAID 0+5 or
RAID 50.
Theory of Operation
7-19
Creating Data Redundancy
RAID 5 offers improved storage efficiency over RAID 1 because only the
parity information is stored, rather than a complete redundant copy of
all data. The result is that three or more drives can be combined into a
RAID 5 array, with the storage capacity of only one drive dedicated to
store the parity information. Therefore, RAID 5 arrays provide greater
storage efficiency than RAID 1 arrays. However, this efficiency must be
balanced against a corresponding loss in performance.
The parity data for each stripe of a RAID 5 array is the XOR of all the data
in that stripe, across all the drives in the array. When the data in a stripe is
changed, the parity information is also updated. There are two ways to
accomplish this:
The first method is based on accessing all of the data in the modified
stripe and regenerating parity from that data. For a write that changes
all the data in a stripe, parity can be generated without having to read
from the disk, because the data for the entire stripe will be in the
cache. This is known as full-stripe write. If only some of the data in a
stripe is to change, the missing data (the data the host does not write)
must be read from the disks to create the new parity. This is known as
partial-stripe write. The efficiency of this method for a particular
write operation depends on the number of drives in the RAID 5 array
and what portion of the complete stripe is written.
The second method of updating parity is to determine which data bits
were changed by the write operation and then change only the
corresponding parity bits. This is done by first reading the old data
which is to be overwritten. This data is then XORed with the new data
that is to be written. The result is a bit mask which has a 1 in the
position of every bit which has changed. This bit mask is then XORed
with the old parity information from the array. This results in the
corresponding bits being changed in the parity information. The new
updated parity is then written back to the array. This results in two
reads, two writes and two XOR operations. This is known as readmodify-write.
Theory of Operation
7-20
RAID 5
Updating Parity During Writes
XOR
WRITE
READ
DATA
XOR
READ
WRITE
PARITY
The cost of storing parity, rather than redundant data as in RAID 1, is the
extra time required for the write operations to regenerate the parity
information. This additional time results in slower write performance for
RAID 5 arrays over RAID 1. Because SmartRAID V controllers generate
XOR in hardware, the negative effect of parity generation is primarily from
the additional disk I/O required to read the missing information and write
the new parity. SmartRAID V controllers can generate parity using either
the full- or partial-stripe write algorithm or the read-modify-write
algorithm. The parity updated method chosen for any given write
operation is determined by calculating the number of I/O operations
needed for each type and choosing the one with the smallest result. To
increase the number of full stripe writes, the cache is used to combine
small write operations into larger blocks of data.
Handling I/O Errors
SmartRAID V controllers maintain two lists for each RAID 5 array: a Bad
Parity List, and a Bad Data List. These lists contain the physical block
number of any parity or data block that could not be successfully written
during normal write, rebuild or dynamic array expansion operations.
These lists alert the controller that the data or parity in these blocks is not
valid. If the controller subsequently needs data from a listed block and
cannot recreate the data from existing redundant data, it returns an error
condition to the host.
Blocks are removed from the Bad Parity List or the Bad Data List if the
controller successfully writes to them on a subsequent attempt.
Theory of Operation
7-21
Degraded Mode
When a drive fails in a RAID 0 array, the entire array fails. In a RAID 1
array, a failed drive reduces read performance by 50%, as data can only be
read from the remaining drive. Write performance is increased slightly
because only one drive is accessed. A RAID array operating with a failed
drive is said to be in degraded mode.
RAID 5 arrays synthesize the requested data by reading and XORing the
corresponding data stripes from the remaining drives in the array. For
RAID 5, the magnitude of the performance impact in degraded mode
depends on the number of drives in the array. An array with a large
number of drives will experience more performance degradation than an
array with small number of drives.
Rebuilding a Failed Hard Drive
A failed drive can be replaced in a RAID 1 or RAID 5 array by physically
removing the drive and replacing it or by a designated Hot Spare.
SmartRAID V controllers will rebuild the data for the failed drive onto the
new drive or Hot Spare. This rebuild operation occurs online while normal
host reads and writes are being processed by the array.
RAID 1 arrays are rebuilt relatively quickly, because the data is simply
copied from the duplicate (mirrored) drive to the replacement drive. For
RAID 5 arrays, the data for the replacement drive must be synthesized by
reading and XORing the corresponding stripes from the remaining drives
in the array. RAID 5 arrays that contain a large number of drives will
require more time for a rebuild than a small array.
Theory of Operation
7-22
SmartRAID V Architecture
SmartRAID V controllers can process commands from the operating
system using two different protocols:
• ASPI (Advanced SCSI Programming Interface), which is compatible
with tape and CD-ROM application software.
• I2O Messaging, which uses the full performance capability and
function of the PCI bus.
ASPI Protocol
SmartRAID V controllers can communicate with any ASPI protocol device
under Windows NT and Windows 95/98.
ASPI support under MS-DOS is limited to CD-ROM drives only. MS-DOS
support is provided by the I2ODDL.SYS and DPTCDROM.SYS driver files.
I2O Messaging Protocol
Intelligent I/O architecture (I2O) is a device driver architecture that
provides a standardized interface between I/O devices and operating
systems. I 2O supports PCI bus mastering, overlapped I/O processing,
command queuing, scatter/gather memory access and auto-request
sense.
The I2O architecture uses a split-driver model, where an I/O driver
consists of two parts:
An Operating system Services Module (OSM) that translates
operating system specific I/O calls into a neutral message-based
format.
A Hardware Device Module (HDM) that accepts the OSM messages
and processes them on a specific controller and devices. (The
SmartRAID V controller functions as an HDM.)
This messaging system enables the modules to communicate without
specific knowledge of the underlying operating system or I/O bus.
Theory of Operation
7-23
Command Processing
The I 2O messaging protocol is based on the existence of a messaging
FIFO (first-in, first-out) queue and a reply FIFO queue. Each queue
consists of a list of message frames. The messaging queue contains
requests for I/O processing that are read by the HDM. The reply queue is
where completion statuses are posted by the controller and read by the
OSM. DPT SmartRAID V controllers use I2O messaging to process a
command from the host computer as follows:
1.
The OSM acquires a frame in controller RAM from the messaging
queue on the controller, fills it out and returns it to the messaging
queue to initiate the I/O command.
2.
The controller HDM firmware processes the frame.
3.
The controller completes the I/O command; transferring data if
necessary.
4.
The controller acquires a frame from the reply queue, fills it out and
returns it to the queue and issues an interrupt.
5.
On interrupt, the OSM retrieves the completed reply frame from
the reply queue and uses the information in the frame to complete
the I/O operation. It then returns the frame to the reply queue.
Command Overlapping
Command overlapping is a technique that lets a controller receive and
respond to multiple host requests before any particular request is
completed. SmartRAID V controllers can receive additional commands
before the previous command has finished executing. By allowing multiple
outstanding commands, SmartRAID V controllers can optimize
performance by overlapping commands for multiple devices, thereby
keeping all devices busy as much as possible. The Decade and Century
family of controllers can queue up to 2000 commands. The Millennium
family of controllers can queue up to 8000 commands.
Theory of Operation
7-24
Command Queuing
I2O messaging allows the controller to queue multiple commands from the
host and execute them out of order for greater efficiency. When a
command is complete, the controller transfers status information to host
memory. This status message contains information indicating which
command was just completed and enables commands to be executed in a
different order than they were issued by the host. This allows commands
to be queued and rearranged so that they are executed in a more efficient
order for the particular peripheral configuration.
SmartRAID V controllers also support Tagged Command Queuing on the
peripheral bus. This feature allows the controller to transfer multiple
commands to a single device if that device can accept queued commands.
Auto Request Sense
When a SCSI error occurs, the controller automatically issues a Request
Sense command and returns the data to the host in the message reply
frame. Data returned by this command is stored in the controller’s Event
Log. You can use Storage Manager to view the Event Log.
Byte/Word Alignment
Data to be read or written by the SmartRAID V controllers need not be
aligned to word or long-word boundaries in system memory. If the data
starts on an odd memory address, the necessary bus control signals will
be generated by the SmartRAID V to access 8-bit, 16-bit, 24-bit and 32-bit
data as required. Data blocks of any byte length are correctly handled by
SmartRAID V.
Scatter/Gather
SmartRAID V controllers support Scatter/Gather. This is a method of
providing multiple memory addresses for data transfer in one host
command. This ability greatly increases performance in environments
such as UNIX, Novell NetWare and Windows NT. Scatter/Gather is
necessary because when using virtual memory addressing schemes,
system memory that may appear contiguous to the user can actually be
fragmented into many widely scattered physical address locations.
Because of this, when accessing a large amount of contiguous data from a
peripheral device, it is often necessary to split the transfer into many
different locations in system memory.
Theory of Operation
7-25
PCI Data Transfer and Bus Mastering
The Peripheral Component Interconnect (PCI) bus provides a means to
interconnect peripheral chips and expansion cards to the computer
backplane. PCI is optimized for peripheral device controllers that transfer
data in blocks. PCI allows for data transfer rates of 132MB/sec (32-bit PCI)
and 264MB/sec (64-bit PCI). If a Plug-and-Play BIOS is available,
automatic expansion card configuration is also available.
All SmartRAID V controllers use a PCI bus for internal communication as
well as their interface to the host computer. SmartRAID V Century
controllers are configured by the host BIOS as multifunction PCI devices.
SmartRAID V Millennium controllers are seen by the host computer as
standard PCI-to-PCI bridge devices. Millennium controllers are available
in both 32-bit and 64-bit PCI versions.
The PCI bus uses a 33MHz clock rate as opposed to the 8.33MHz clock
used by ISA and EISA buses. To maintain data integrity at the higher clock
rate, the PCI bus is limited to four card slots and six devices integral to the
motherboard. However, motherboard designs can include multiple PCI
buses that are interconnected through bridge devices.
PCI supports two modes of data transfer: Bus mastering and PIO. Third
party DMA is not supported by the PCI architecture.
Each PCI device contains configuration registers that allow the system to
be configured for maximized throughput. Parameters such as Latency
Timer, Cache Line Size, Minimum Bus Grant, Maximum Latency, as well
as configurable address and IRQs, can be read or written by the PCI
system BIOS. SmartRAID V controllers support the full range of required
and optional features described by revision 2.1 of the PCI specification.
SmartRAID V controllers are also fully compliant with revision 2.1 of the
PCI specification for bus drivers and receivers. These transceivers are
unique in their ability to vary the drive current in order to minimize signal
reflections.
All mandatory and optional memory read/write commands are supported
by SmartRAID V controllers. Mandatory commands include Memory
Read and Memory Write, which are used by all PCI bus-mastering devices.
Optional commands include Memory Read Line, Memory Read Multiple
and Memory Write and Invalidate. These commands maximize
performance in PCI systems with intelligent bridges that perform CPU
cache line optimization.
Theory of Operation
7-26
On-Board Microprocessor
SmartRAID V controllers use embedded processors to maximize their
performance when installed in fully loaded servers. The onboard
microprocessor performs high-level processing of commands from the
host and controls the sequencing of command execution on the controller.
A fast microprocessor enables the controller to process more I/O
commands per second which is especially important when caching and
RAID are used.
SmartRAID V Millennium controllers use a high-performance RISC
processor. SmartRAID V Decade and Century controllers use an Intel
i960® embedded I/O processor.
SCSI Interface
The Small Computer Systems Interface (SCSI) is a standard parallel
interface designed to communicate with intelligent peripheral devices.
SmartRAID V controllers conform to the ANSI SCSI specification defined
in ANSI document number X3.131-1986 and subsequent standards defined
in ANSI X3.131-1994 (SCSI-2) and various SCSI-3 specification documents.
SCSI Device IDs
The original SCSI specification allowed up to eight SCSI devices, including
the controller, to be connected to a single 8-bit SCSI bus. Wide (16-bit)
SCSI busses support up to fifteen devices and one controller per bus. All
SCSI devices, including the controller, must be assigned a unique numeric
identifier. These SCSI IDs can be any number from 0 – 7 or, for Wide SCSI
devices, 0 – 15. The SmartRAID V controller ID is set by default to ID 7.
In 8-bit SCSI systems, the device with ID 7 has the highest priority and ID
0 has the lowest priority when arbitrating for use of the SCSI bus. In a
Wide SCSI system, IDs have the following priority (highest to lowest):
7, 6, 5, 4, 3, 2, 1, 0, 15, 14, 13, 12, 11, 10, 9, 8
The IDs 8 – 15 always have a lower priority during arbitration than 8-bit
devices, this allows 8-bit devices which do not recognize IDs 8 – 15 to
coexist on a SCSI bus with Wide SCSI devices which may be assigned
those IDs. However, when controlling 8-bit devices, a Wide SCSI controller
cannot be assigned an ID greater than 7 because 8-bit devices would not
be able to recognize the controller.
Theory of Operation
7-27
Logical Unit Numbers (LUN)
A SCSI device may contain up to eight sub-devices or Logical Units,
assigned Logical Unit Numbers (LUN) 0 – 7. SmartRAID V controllers
support multiple LUNs. However, most devices other than bridge
controllers consist of only one LUN and are accessed as LUN 0 by
SmartRAID V controllers.
Bridge Controllers
Bridge controllers are single devices that can, in turn, control multiple
peripheral devices. These devices may be disk drives configured as a
remote RAID subsystem. Each of the peripheral devices attached to the
bridge controller is assigned a Logical Unit Number. A SmartRAID V
controller supports 250 physical devices. These devices can be individual
SCSI IDs for actual physical devices or LUNs within an ID (such as on a
bridge controller).
Commands and Messages
All SCSI commands are supported by SmartRAID V. SmartRAID V
controllers pass through any SCSI command from the host computer to
the SCSI peripheral device.
SmartRAID V also supports one byte, two byte, and extended messages,
such as Initiator Detected Error, Parity Error, Simple Queue, Wide
Negotiation and Synchronous Negotiation. The SCSI message protocol is
normally handled automatically by SmartRAID V controllers without
requiring intervention by the host computer. This allows much faster SCSI
protocol handling and reduces the load on the host CPU, freeing it for
other tasks.
Disconnect/Reconnect
SmartRAID V provides full support for Disconnect and Reconnect
commands by SCSI peripheral devices. All Disconnect/Reconnect
commands are handled automatically by SmartRAID V controllers,
without host intervention.
Disconnection allows a SCSI device or controller to temporarily release
control of the SCSI bus, during periods of non-activity while a SCSI
command is in process, so that another SCSI device can share the bus.
Theory of Operation
7-28
Transfer Padding and Residue Reporting
In unusual cases, a SCSI peripheral device might read or write more data
than was requested by the host. If this occurs during a write operation,
SmartRAID V controllers automatically transfer pad data by adding
additional bytes to the end of the data. If this occurs during a read
operation, the controller automatically strips off additional bytes of data
read from the device. This prevents the data from overwriting valid data in
host memory.
It is also possible for a SCSI peripheral device to read or write less data
than was requested by the host. If this occurs, SmartRAID V controllers
will report the actual number of bytes transferred as less than the
requested number of bytes.
Single-Ended and Differential SCSI
The SCSI electrical specification has two transceiver specifications:
1.
Single-Ended TTL transceivers which allow:
Asynchronous data transfer at a maximum cable length of 6 meters
(19.6 feet). No maximum transfer rate is specified but typical rates
range from 1.5MHz to 3MHz.
Synchronous data transfer up to 5MHz at a maximum cable length
of 6 meters.
10MHz (Fast) synchronous data transfer at a maximum cable
length of 3 meters (9.8 feet).
20MHz (Ultra) synchronous data transfer at a maximum cable
length of 3 meters with up to 4 devices and 1.5 meters (4.9 feet) with
up to eight devices.
2.
Differential SCSI: Differential electrical signal protocol transmits
information through a current loop rather than by changes in
voltage, thereby reducing susceptibility to electrical interference.
Differential SCSI has two variants:
High Voltage Differential (HVD) SCSI uses RS-485 transceivers that
allow up to 20MHz data transfer with a maximum cable length of 25
meters (82 feet).
Ultra2 SCSI Low Voltage Differential (LVD) transmission. This
transmission method has the benefits of differential SCSI without
the need for external, high-voltage bus transceivers. Ultra2 SCSI
allows synchronous transfers up to 40MHz with a maximum cable
length of 12 meters. SmartRAID V products support multimode
Ultra2 SCSI, which includes LVD operation.
Theory of Operation
7-29
Multi-Mode SCSI
To ensure backward compatibility with existing SCSI devices, all
SmartRAID V Ultra2 controllers support both LVD and single-ended
devices. This is referred to as multimode SCSI. Both types of devices can
be attached to the same SCSI bus with the following restrictions:
• If all devices on the bus are Ultra2, the bus operates in LVD mode and
runs at the maximum Ultra2 speed (40 MHz).
• If the devices are a mix of 40 MHz Ultra2 and 20 MHz Ultra singleended, the bus operates in single-ended mode and runs at 20MHz.
!
CAUTION
!
Do not use both LVD and High Voltage Differential devices on the
same SCSI bus.
Do not attach High Voltage Differential devices to an Ultra2 bus.
Wide SCSI
The SCSI-2 specification defines 8-bit, 16-bit and 32-bit bus widths. All
SmartRAID V controllers use a 16-bit bus and support both 8-bit and 16-bit
SCSI devices.
SCSI Transfer Rate
SmartRAID V controllers are capable of handling a combination of SCSI
devices with different transfer speeds, and both synchronous and
asynchronous protocols on the same SCSI bus. Using the SCSI message
protocol, SmartRAID V controllers automatically negotiate with each
device during the power-on sequence to determine the fastest transfer
speed and protocol for that device.
SmartRAID V controllers can also detect the presence of an external SCSI
cable and automatically lower the maximum SCSI bus transfer rate when
an external cable is present. (Both the external and internal cable transfer
rates are lowered.) Because some external SCSI cabling methods are not
capable of handling the maximum possible transfer rates without causing
data errors, this feature protects you from data integrity problems.
Theory of Operation
7-30
The original SCSI specification synchronous transfer rates of up to 5MHz.
The SCSI-2 specification increased the maximum synchronous rate to
10MHz. Devices that support the faster 10MHz rate are called Fast SCSI
devices. The SCSI-3 specification defines transfer rates up to 20MHz (Fast20/Ultra SCSI).
The SCSI-3 SPI-2 specification defines synchronous data transmission
speeds of up to 40MHz. The 40MHz transfer speed provides data transfer
rates up to 40MB/sec on an 8-bit bus and 80MB/sec on a 16-bit SCSI bus.
Devices that support these rates are called Ultra2 SCSI devices. Ultra2
SCSI is also known as Fast-40 SCSI. A device must use Low Voltage
Differential signaling to achieve the maximum Ultra2 speed.
Fibre Channel Arbitrated Loop (FC-AL) controllers transfer data at a
maximum speed of 1GHz per loop (100 MB/sec).
Cabling Single-Ended SCSI
The SCSI cable is an electrical transmission line that has a characteristic
impedance. The value of the impedance depends on the type and
configuration of the cable. Differences in this impedance can cause signal
reflections. These impedance variations can be the result of extra
capacitance internal to SCSI devices, connectors, incorrect termination,
mixing of different cable types, cable stubs, and so on.
A properly configured Single-Ended SCSI bus can reliably transfer data at
up to 20MHz. However, the following guidelines should be used to ensure
success:
Use as short a cable length as possible. For 10MHz data transfers
the total SCSI bus length should not exceed 3 meters (9.8 feet). For
20MHz data transfers, total SCSI bus length should not exceed 3
meters with up to 4 devices on the bus, or 1.5 meters (4.9 feet) with
between 5 and 8 devices on the bus.
NOTE
When an external SCSI storage cabinet is used, the cabinet’s
internal SCSI backplane will add additional length to the bus.
Avoid stub clustering. Avoid spacing SCSI devices on the cable closer
than 0.3 meters (11.8 in). When devices are clustered closely together
on the SCSI cable, their capacitance adds together to create an
impedance discontinuity and thus reflections. SCSI devices should be
spaced as evenly as possible.
Cable stub length should not exceed 10 cm (3.9 in). Some SCSI
devices may create stubs internal to the device which exceed this
value. This can result in excessive capacitive loading and signal
reflections. This factor is controlled by the SCSI device manufacturer.
The SCSI cabling itself should not include stubs.
Theory of Operation
7-31
Beware of capacitance changes. As devices are added to a SCSI bus,
capacitance is introduced to each signal from the connectors,
receivers, and circuit board traces. The SCSI-3 working specification
limits this capacitance to 25pF because added capacitance lowers the
impedance of the cable near the device and adds delay. Look for input
filters that may be attached to the SCSI front-end of the device’s
printed circuit board. These filters add capacitance.
Avoid unnecessary connector converters. They cause impedance
discontinuity and signal reflections.
Route the cable with care. Avoid rolling the cable up on itself,
running the cable alongside of metal for long lengths or routing the
cable past noise generators (such as power supplies). Placing the
cable near ground planes created by grounded metal components also
reduces its impedance. The SCSI-3 draft specification suggests that to
minimize discontinuity due to local impedance variation, a flat cable
should be spaced at least 1.27 mm (0.050 in.) from other cables, any
other conductor, or the cable itself when the cable is folded.
Follow the SCSI specification for cable impedance. Cable
impedance for 10MHz SCSI systems should be limited to 84+/-12
Ohms (Ω). Cable impedance for 20MHz SCSI systems should be
limited to 90+/-12Ω for REQ and ACK signals and 90+/-10Ω for all
other signals.
Avoid mixing cable types. Select either flat or round, shielded or
unshielded. Typically mixing cables mixes impedance. Cable
impedance mismatch is a common problem that results in signal
reflections. Internal cables are normally unshielded flat ribbon cables,
while external cables should be shielded. Where they offer easier
routing, size advantages, and better air flow, round cables can also be
used internally. This may be beneficial if it allows for better impedance
matching to the external cable.
When round cable is used, select a cable that uses an optimal
placement of signals within the cable. Ribbon cable has good
crosstalk rejection due to the GND-Signal-GND layout. With standard
25-pair round cable, pairs are arranged in three layers. The closer a
pair is to the cable center, the higher the impedance. Using centrally
located high impedance pairs for speed-critical signals such as REQ
and ACK is a good choice. Locating the data pairs in the outermost
layer of the cable minimizes crosstalk between REQ, ACK and the
data lines. The middle layer might contain status lines such as C/D,
I/O, MSG, ATN, and so on. Make sure that the lowest impedance wire
in the cable is used for TERMPWR to minimize transmission line
effects on this voltage supply line. Some SCSI cables have a lowimpedance conductor specifically for this purpose.
7-32
Theory of Operation
Cabling Ultra2 (LVD) SCSI
When the total length of a SCSI cable must exceed the maximum length
for single-ended buses, DPT recommends the use of an Ultra2 Low
Voltage Differential SCSI interface.
The following guidelines will help ensure reliability when configuring
Ultra2 SCSI systems:
• Use twisted-pair cable. Twisted-pair cable (either twisted-flat or
discrete wire twisted-pairs) provides greater signal integrity over
longer distances because noise coupled into a twisted-pair generally
appears equally on both wires. Because the differential receivers
respond to differences between the conductors of the twisted-pair, the
coupled common-mode noise is rejected.
Flat non-twisted ribbon cable can cause two problems. First, noise
introduced into parallel conductors tends not to be common mode.
Second, while the single-ended conductor arrangement interleaves
ground wires between signal wires, there are not enough conductors
to interleave grounds between each differential signal pair. These
factors lead to increased crosstalk between adjacent conductors on a
flat non-twisted ribbon cable.
• The maximum cable length for an LVD bus is 12 meters (39.3 feet).
• Because Ultra2 cables are sensitive to cable capacitance, the following
table of minimum spacing between devices should be followed:
Cable Capacitance
Minimum Spacing
40 pF/m
36 cm
65 pF/m
22 cm
90 pF/m
16 cm
115 pF/m
13 cm
140 pF/m
10 cm
• Stub lengths should not exceed 0.1 meters (3.9 in.). The difference in
stub length should be less than 1.27 cm (0.05 in.) for the REQ, ACK,
DATA and PARITY signals.
• Terminators should be installed only at each physical end of the bus,
either on the last device attached or the unused connector at the end
of the cable. Ultra2 buses must be terminated with LVD terminators.
DPT cables are supplied with an LVD terminator installed at the end
of the cable.
Theory of Operation
7-33
SCSI Termination
SCSI termination for SmartRAID V controllers is controlled through
Storage Manager or SMOR. Refer to Chapter 3, “Configuration and
Installation” for information about setting SCSI termination for
SmartRAID V controllers.
SmartRAID V controllers contain onboard active SCSI terminators for
maximum reliability with Fast and Ultra SCSI devices. Active termination
for a single-ended bus uses one 110Ω resistor per signal, pulled up to a
locally supplied voltage that is regulated at 2.85V. Active termination for
Ultra2 SCSI uses a differential impedance of 100Ω with a 1.25V bias.
Features of active termination include:
• Better immunity to fluctuations in TERMPWR.
• Closer match to the characteristic impedance of the cable minimizes
reflections.
• Increased signal-to-noise ratio.
Passive termination employs one 220Ω pull-up and one 330Ω pull-down
resistor per signal. Active terminators should be used whenever possible.
However, active and passive termination can be mixed on a SCSI bus and
will result in better reliability than using only passive termination.
IMPORTANT
Do not use passive termination for Fast-20 or Ultra SCSI in a single-ended
bus configuration.
Theory of Operation
7-34
TERMPWR
SCSI terminators need power to operate. SmartRAID V controllers always
supply power for their own onboard SCSI terminators and, by default,
supply termination power for other devices using the TERMPWR line on
the SCSI cable. Most SCSI devices allow the way in which TERMPWR is
supplied to their onboard SCSI terminators to be selected using jumpers
on the device. Devices that supply SCSI bus termination should be
configured to supply their own isolated TERMPWR. This prevents loss of
receiver noise margin due to a DC voltage drop across the cable.
TERMPWR should remain enabled if there is a possibility of power-loss on
an external SCSI device that supplies its own TERMPWR.
If all devices supplying termination for the SCSI cable are also capable of
supplying their own termination power, then TERMPWR supplied to the
SCSI cable from SmartRAID V controllers can be disabled using SMOR or
Storage Manager.
SmartRAID V controllers supply termination power onto the SCSI cable
through a circuit breaker which protects the controller from a short-toground on the TERMPWR line. Some SCSI devices use a fuse for
protection; this can require the replacement of the part if a short occurs or
if the SCSI cable is inserted incorrectly.
The ANSI XT3T9.3 SCSI-3 Parallel Interface working group recommends
that all TERMPWR lines be decoupled at each terminator to minimize
TERMPWR glitch coupling. The minimum recommended values are a
2.2µF solid tantalum capacitor along with .01µF ceramic capacitor in
parallel to suppress high frequency, low voltage noise. These capacitors
supply the high frequency, low impedance path to ground necessary to
provide a clean supply voltage.
Theory of Operation
7-35
Fibre Channel Arbitrated Loop (FC-AL)
Fibre Channel is an industry standard set of protocols for information
transfer using a serial bus. Fibre Channel provides for reliable high-speed
transfer of large amounts of information over long cable lengths.
SmartRAID V Fibre Channel controllers support transfer rates of 1GHz
among up to 126 devices over a maximum of 30 meters (98 feet) of copper
or 10 km (6.2 miles) of optical cable.
DPT Fibre Channel controllers use Arbitrated Loop (FC-AL) topology,
where all devices and the controller are connected in a serial (loop)
fashion. FC-AL provides a cost-effective method of implementing a Fibre
Channel storage subsystem, while maintaining high bandwidth and
reliability characteristics.
Redundancy can be added to a Fibre Channel storage subsystem by using
a dual loop configuration. By forming two arbitrated loops on a dual
channel controller and connecting dual-port devices to both loops, fault
tolerance is provided for cable failure. In normal operation, the controller
will use both loops to communicate with the devices, thus providing an
effective controller bandwidth of 200MB/sec.
DPT SmartRAID V Fibre Channel controllers can have a one or two
HSSDC (copper) connectors. If you want to use a cable length greater than
30 meters, a Media Interface Adapter (MIA) can be used to convert
electrical signals to optical signals for use with fiber optic cable. Optical
cable has complete immunity to electrical noise and the ability to use
significantly longer cable lengths without signal loss.
Theory of Operation
7-36
DPT I2O BIOS
SmartRAID V controllers are shipped with the DPT I 2O BIOS loaded on
an EEPROM. This BIOS intercepts and processes Int13 requests with an
embedded DOS driver.
The host system Plug-and-Play BIOS can automatically disable or change
the address of the I 2O BIOS ROM. In systems with multiple DPT
controllers, the first DPT I2O controller found during boot loads its BIOS
and installs all of the DPT I2O hardware on the system. DPT I2O
controllers found subsequently automatically detect the presence of the
first controller and disable their BIOS code. If another manufacturer’s
BIOS occupies address C8000H, then an alternate BIOS ROM address will
be selected for either the DPT controller or the other card. The disk
controller that has the lowest BIOS ROM address (typically, the lowest
PCI slot) will become the booting controller. Therefore, in a system with
multiple controllers, ensure that the DPT ROM occupies the lowest
address if you want the DPT controller to be the booting controller.
If the DPT I2O BIOS is correctly configured, the following message
appears during boot-up, followed by a list of attached devices:
DPT
I2O SCSI BIOS V001.xx (yyyy/mm/dd)
Distributed Processing Technology
©Copyright 1996-99
All Rights Reserved
Hit <CTRL+D> for DPT setup
APPENDIX A:
Assembly Drawings
This chapter contains outline drawings of
DPT’s SmartRAID V products with jumpers
and connectors labeled for easy reference.
Millennium Controllers
PM3755F
PM3755U2B
PM3754U2
See the SmartRAID V Controller LEDs
section in Chapter 2 for specific information
about the LED indicators on your controller
or module.
Decade/Century Controllers
PM1554U2
PM2554U2
PM2654U2
Bus Expansion Modules
SX4054U2-1
SX4054U2-2
SX4055U2-1
SX4055U2-2
SX4055F
RAID Accelerator
RA4050
Battery Module
BB4050
A-1
A-2
Assembly Drawings
PM3755F
ECC EN
pin-1
P9
P6
P4
ECC ERR
pin-1
87654321
IRQ
Audible Alarm
J9
SIMM 1
SIMM 2
SIMM 3
J8 J7
SIMM 4
P4
Pins 1-2 NVRAM Clear
Pins 3-4 Misc (Reserved, do not use)
P6
Disk Activity LED Connector
P9
Pins 1-2 Load
Pins 3-4 Run
J7
32-bit Bus Expansion Module Connector
J8
64-bit Bus Expansion Module Connector
J9
External Fibre Channel HSSDC Connector (Bus 0)
1…8, IRQ
Adapter Activity LEDs
ECC EN
ECC Enabled LED (Green)
ECC ERR
ECC Error LED (Red)
Assembly Drawings
A-3
PM3755U2B
J10
pin-1
ECC EN
87654321
P6
IRQ
ECC ERR
pin-1
P4
DIMM 3 & 4
J12
pin-1
J7
J11
J8
P9
pin-1
Audible Alarm
RUN LOAD
P4
Pins 1-2 NVRAM Clear
Pins 3-4 Misc (Reserved, do not use)
P6
Disk Activity LED Connector
P9
Pins 1-2 Load
Pins 3-4 Run
J7
32-bit Bus Expansion Module Connector
J8
64-bit Bus Expansion Module Connector
J10
Internal Wide Ultra2 SCSI Connector (Bus 0)
J11
External Wide Ultra2 SCSI Connector (Bus 0)
1…8, IRQ
Adapter Activity LEDs
ECC EN
ECC Enabled LED (Green)
ECC ERR
ECC Error LED (Red)
DIMM 1 & 2
A-4
Assembly Drawings
PM3754U2
P4
J10
pin-1
ECC EN
P6
87654321
IRQ
ECC ERR
pin-1
Audible Alarm
pin-1
J11
pin-1
SIMM 1
P9
J4
SIMM 2
SIMM 3
SIMM 4
P4
Pins 1-2 NVRAM Clear
Pins 3-4 Misc (Reserved, do not use)
P6
Disk Activity LED Connector
P9
Pins 1-2 Load
Pins 3-4 Run
J4
Bus Expansion Module Connector
J10
Internal Wide Ultra2 SCSI Connector (Bus 0)
J11
External Wide Ultra2 SCSI Connector (Bus 0)
1…8+IRQ
Adapter Activity LEDs
ECC EN
ECC Enabled LED (Green)
ECC ERR
ECC Error LED (Red)
Assembly Drawings
A-5
PM1554U2, PM2554U2 and PM2654U2
pin-1
87654321
IRQ
P6
P9
pin-1
pin-1
J11
J7
P4
Pins 1-2 NVRAM Clear
Pins 3-4 Misc (Reserved, do not use)
P6
Disk Activity LED Connector
P9
Pins 1-2 Retry (Reserved, do not use)
Pins 3-4 Reset (Reserved, do not use)
J7
Bus Expansion Module Connector
J10
Internal Ultra2 Wide SCSI Connector (Bus 0)
J11
External Ultra2 Wide SCSI Connector (Bus 0)
1…8, IRQ
Adapter Activity LEDs
P4
RA4050 Module Connector
J10
A-6
Assembly Drawings
SX4054U2-1 Bus Expansion Module
pin-1
J2
pin-1
J3
P5
J2
Internal Ultra2 Wide SCSI Connector (Bus 1)
J3
External Ultra2 Wide SCSI Connector (Bus 1)
P5
Mating Connector for SmartRAID V Controller
Assembly Drawings
A-7
SX4054U2-2 Bus Expansion Module
pin-1
J4
pin-1
J2
pin-1
J3
P5
pin-1
J1
J1
External Ultra2 Wide SCSI Connector (Bus 2)
J2
Internal Ultra2 Wide SCSI Connector (Bus 1)
J3
External Ultra2 Wide SCSI Connector (Bus 1)
J4
Internal Ultra2 Wide SCSI Connector (Bus 2)
P5
Mating Connector for SmartRAID V Controller
A-8
Assembly Drawings
SX4055U2-1 Bus Expansion Module
pin-1
J2
pin-1
J3
P5
P3
J2
Internal Wide Ultra2 SCSI Connector (Bus 1)
J3
External Wide Ultra2 SCSI Connector (Bus 1)
P3
64-bit SmartRAID V Adapter Connector
P5
Mating Connector for SmartRAID V Controller
Assembly Drawings
A-9
SX4055U2-2 Bus Expansion Module
pin-1
J4
pin-1
J2
pin-1
J3
P5
P3
pin-1
J1
J1
External Wide Ultra2 SCSI Connector (Bus 2)
J2
Internal Wide Ultra2 SCSI Connector (Bus 1)
J3
External Wide Ultra2 SCSI Connector (Bus 1)
J4
Internal Wide Ultra2 SCSI Connector (Bus 2)
P3
64-bit SmartRAID V Adapter Connector
P5
Mating Connector for SmartRAID V Controller
A-10
Assembly Drawings
SX4055F FC-AL Bus Expansion Module
P5
P3
P2
P2
Fibre Channel (FC-AL) Connector (Bus 1)
P3
64-bit SmartRAID V Adapter Connector
P5
Mating Connector for SmartRAID V Controller
Assembly Drawings
A-11
ECC ERR
ECC EN
RA4050 RAID Accelerator
Docking Guide
J1
SIMM 1
Docking Connector
SIMM 2
SIMM 3
SIMM 4
Docking Guide
J1
Docking Connector
ECC EN
ECC Enabled LED (Green)
ECC ERR
ECC Error LED (Red)
Audible Alarm
A-12
Assembly Drawings
BB4050 Battery Module
TRICKLE
Stability Bracket mounting location
J1
CHARGE
Mounting Hole
P12
Mounting Holes
Board shown without Stability Bracket
J1
Battery Cable Connector
P12
Connector for SmartRAID V PM3755U2B
TRICKLE
LED indicator for trickle charge activity
CHARGE
LED indicator for charging/recharging cycle
!
CAUTION
!
Danger of explosion if battery is incorrectly replaced. Replace only
with the same or equivalent type recommended by the
manufacturer. Dispose of used batteries according to the
manufacturer’s instructions.
APPENDIX B:
Troubleshooting
This chapter provides answers to many
commonly asked questions. If a situation
occurs that is not covered in this chapter, or
if the recommendations here do not correct
the problem, contact DPT Technical Support. We are always ready to assist you.
DPT Technical Support
Phone:
407-830-5522
(Press 6 and follow the
prompts.)
Fax:
407-830-4793
Internet:
http://www.dpt.com
Product information and the latest
versions of DPT drivers and
utilities can be obtained at no
charge from the DPT FTP site
(ftp.dpt.com) or from the Technical
Support section of our World Wide
Web site 24 hours a day.
DPT also offers priority Technical
Support as a fee-based option. If
you choose one of these options,
your call is given priority over all
other support calls:
900-555-4378
at the rate of $1.35 per minute
407-830-5522
Press 6 and select the Priority
Technical Support option. The rate
is $30 for the first hour and $1.00
per minute after the first hour.
B-1
B-2
Troubleshooting
Problem:
When the DPT I2O BIOS displays the peripheral devices at system boot, a
device does not appear.
Solution:
The following conditions can cause this to occur:
The device ID might be set to the same ID as the DPT controller (ID 7).
Ensure that all devices have a unique ID. Fibre Channel devices
automatically configure a unique ID for each device on the bus.
The device might not be powered on.
The device is not connected to the SCSI cable or the connection is loose.
Problem:
In addition to the SmartRAID V controller, the system contains another
manufacturer’s SCSI controller and hangs during boot.
Solution:
The other controller does not correctly implement EBDA usage rules. Use
SMOR to try a different setting for the EBDA Relocation parameter or
rearrange the controller slot assignments.
If your system BIOS supports configuring the boot order, you can also try
changing those settings.
Problem:
The system contains another manufacturer’s SCSI controller in addition
to the SmartRAID V controller. During boot, messages from each
controller’s BIOS appear, but one controller cannot communicate with its
attached drives.
Solution:
The other controller does not properly implement EBDA usage rules. Use
SMOR to try a different setting for the EBDA Relocation parameter or
rearrange the controller slot assignments.
If your system BIOS supports configuring the boot order, you can also try
changing those settings.
Problem:
Windows NT displays a blue screen error message that references the
system video controller.
Solution:
The video controller does not properly implement extended BIOS data
area (EBDA) usage rules. Use SMOR to enable the EBDA Relocation
option.
Problem:
The controller does not respond and the IRQ LED (and possibly other
LEDs) remains lit. See Appendix A for the location of the LEDs on your
controller.
Solution:
The IRQ LED indicates that the controller IRQ assignment is pending.
This usually indicates an IRQ conflict with another card. Ensure that each
card is set to a unique IRQ.
Troubleshooting
Problem:
B-3
The controller does not respond and one of the following LED patterns
occurs at power-up:
LEDs 6 and 7 alternating with LEDs 5 and 8
LEDs 5 and 6 alternating with LEDs 7 and 8
LEDs 5 and 7 alternating with LEDs 6 and 8
Solution:
These patterns indicate that the DPT controller is not being configured by
the motherboard BIOS. SmartRAID V Century controllers require a
motherboard BIOS that supports multifunction devices, where one of the
devices is a PCI-PCI bridge. All SmartRAID V controllers require a BIOS
that supports large memory-mapped address ranges.
Refer to the SmartRAID V READ.ME file on the DPT diskette for
information about motherboard compatibility and a list of motherboards
that DPT has tested with SmartRAID V products.
Problem:
The controller does not respond and one of the following patterns of LEDs
flash once per second at power-up:
PATTERN
Solution:
MEANING
7, 6, 5, 2, 1
None
7, 6, 5, 3, 1
High
7, 6, 5, 3, 2
Mismatch
7, 6, 5, 3, 2, 1
Invalid
These patterns indicate that there is a problem with the memory modules
on the controller.
None:
Either no memory modules were detected on the controller,
or there is no module in socket 1. Socket 1 must always have
a 16MB or 64MB 60ns EDO memory module installed.
High:
Too much memory has been detected on a Decade or
Century controller. Remove memory so that the total is less
than or equal to 64MB.
Mismatch:
Memory modules of mixed sizes have been detected or a
SIMM slot was skipped. All installed modules must be the
same size and must be filled sequentially from socket 1 to
socket 4.
Invalid:
A memory module of a size other than 16MB or 64MB has
been detected. Use only 16MB or 64MB memory modules.
NOTE:
Do not install non-EDO SIMMs or DIMMs. This will cause data
corruption.
Troubleshooting
B-4
Problem:
The controller does not respond and various LEDs in the 1–4 range flash
once per second.
Solution:
This pattern indicates an internal microprocessor trap occurred in the
controller. Remove all attached devices, cables and option modules and
retry. If the trap error disappears, reconnect the cables and devices, one
device at a time, until the faulty device, cable or module is isolated. If the
error persists, contact DPT Technical Support.
Problem:
Pressing Ctrl+D to access SMOR does not work or the information
displayed is garbled.
Solution:
If this happens, use the following procedure to restore the parameters in
the NVRAM to their default settings:
1.
Turn off power to the system.
2.
Place a shorting jumper across pins 1 and 2 of P4 on the controller.
3.
Power on the system and wait until the LEDs 3, 5, 7 and 8 on the
controller begin flashing.
4.
Turn off power to the system and remove the jumper.
The controller can now be reconfigured using SMOR or Storage Manager.
Problem:
During the installation of SCO UNIX with a DPT PCI controller the
following message appears:
Warning: SCSI controller cannot install interrupts
vecno=xx, type = 2, IPL=5
Vector xx is private.
Solution:
SCO UNIX reserves certain interrupts for its internal use. This error
indicates that the DPT controller has been assigned one of these reserved
interrupts. Assign a different interrupt to the controller and start the
installation again.
Problem:
You want the system to boot from a drive that is not attached to a DPT
controller. However, during boot, the DPT I2O BIOS message appears first
which indicates that a drive attached to a DPT controller will be the boot
drive.
Solution:
Use SMOR to modify the Bootable Devices parameter by selecting
Disabled. This will prevent the DPT controller from being used as the
booting controller for system.
Troubleshooting
B-5
Problem:
The SmartRAID V controller I2O BIOS reports the drive as a disk instead
of a drive.
Solution:
This typically happens when a drive that is attached to a SmartRAID V
controller has been formatted with a sector size other than 512 bytes . Use
Storage Manager or SMOR to reformat the drive with 512 byte sectors.
This can also occur if the drive is the 9th or higher logical drive attached to
the controller.
NOTE
Using SMOR to set Bootable Devices to Disabled as in the
previous problem will result in the same symptoms. If you
require access to disk drives connected to the SmartRAID V
controller during the boot process, change Bootable Devices to
Normal.
Problem:
Although the SCSI devices can be accessed by the SmartRAID V
controller, the fault LEDs on the devices in a RAIDstation storage cabinet
do not flash during boot-up and the SmartRAID V controller does not
detect drive swaps or cabinet failures.
Solution:
These symptoms indicate that the RAIDstation storage cabinet status
signals are not being properly received by the SmartRAID V controller.
For DEC Fault Bus subsystems: this can be caused by another SCSI
device or non-DPT cabinet connected to the external SCSI cable along
with the RAIDstation cabinet. Other devices will typically ground these
signals.
For SAF-TE or SES: this can result from a failed enclosure monitoring
module in the subsystem cabinet.
Problem:
After updating the SmartRAID V controller firmware or BIOS and
rebooting, the adapter does not respond.
Solution:
The update may have been unsuccessful. The controller is now in a state
in which it hangs the system during boot. If this happens, the parameters
in the NVRAM can be restored to their default settings using the following
procedure:
1.
Turn off power to the system.
2.
Place a shorting jumper across pins 1 and 2 of P4 on the controller.
3.
Power on the system and wait until the LEDs 3, 5, 7 and 8 on the
controller begin flashing.
4.
Turn off power to the system and remove the jumper.
You can now reconfigure the controller using SMOR or Storage Manager.
Troubleshooting
B-6
Problem:
After updating the SmartRAID V controller firmware or BIOS and
rebooting, LEDs 1 and 5 or 2 and 5 flash once per second.
Solution:
These patterns indicate that the adapter startup code detected a firmware
checksum error or a flash error. Attempt the firmware update procedure
again by using the procedure in the following Problem description to
recover from this condition.
Problem:
A flash ROM upgrade is unsuccessful, causing the controller to hang.
Solution:
The new firmware can be temporarily disabled and the upgrade attempted
again by following the steps below:
1.
Power-off the system.
2.
PM1554, PM2554 and PM2654 controllers – remove any RA4050
or SX405x expansion modules from the controller.
3.
PM1554, PM2554 and PM2654 controllers – place shorting
jumpers across pins 1 and 2 and pins 3 and 4 of P9 on the controller.
PM375x controllers – move the shorting jumper from pins 3 and 4
of P9 to pins 1 and 2 (from RUN to LOAD) as shown in Appendix A,
“Assembly Drawings”.
NOTE
You do not need to remove an SX405x expansion module from
PM375x controllers to access the jumper pins at P9.
4.
NOTE
Insert the DPT SMOR Boot diskette and power-up the system. This
will start SMOR.
The SMOR Boot diskette image is available from the DPT
Technical Support ftp site. The download file contains the
diskette image and instructions for use.
5.
Use SMOR to update the firmware. You must restore all three
components of the flash ROM, firmware, I2O BIOS, and SMOR.
6.
Power-off the system and return the jumpers to their original
positions.
7.
Reattach the expansion module to the controller card and insert
the card in a host system PCI slot.
8.
Remove the SMOR Boot diskette from your floppy disk drive and
power-up the system.
Troubleshooting
B-7
Problem:
After a flash ROM upgrade of the I 2O BIOS only, pressing Ctrl+D at the
system prompt displays the message Card not configurable.
Solution:
Perform a flash ROM upgrade for SMOR to correct this condition.
Problem:
The floppy disk drive cannot be accessed after installing a DPT controller.
Solution:
Use SMOR to enable the EBDA Relocation option.
Problem:
The controller’s audible alarm is sounding during normal operation.
Solution:
This indicates a drive has failed. Start Storage Manager or restart the host
system and run SMOR to identify the failed drive. The alarm will stop
when Storage Manager or SMOR finish the system scan. Replace the
failed drive and start a rebuild operation for the array.
For additional information about procedures for failed drives, refer to
Chapter 4, “Storage Manager on ROM” or Chapter 6, “Storage Manager”.
B-8
Troubleshooting
APPENDIX C:
Specifications
This chapter lists the electrical and environmental specifications for the SmartRAID V
product line.
Specifications:
• DC Power Requirements
• Environmental Specifications
• Memory Requirements
• Battery
C-1
C-2
Specifications
DC Power Requirements
Voltage:
5 V ± 5%
Ripple and Noise:
50mv peak-to-peak max
Component
Current (typical)
PM3754U2B
1.82A
PM3755U2B
1.90A
PM3755F
1.82A
PM2654U2
1.63A (with SX4054U2)
1.60A (with SX4055F)
1.35A (with RA4050; no cache)
PM2554U2
1.15A
PM1554U2
1.15A
RA4050
0.20A
SX4054U2
0.48A
SX4055U2
0.98A
SX4055F
0.45A
BB4050
0.27A (not charging)
1.16A (charging)
SM4050-16
0.16A
SM4050-64
0.16A
Voltage:
3.3 V ± 5%
Ripple and Noise:
50mv peak-to-peak max
Component
Current
DM4050-16
DM4060-16
0.16A typical
DM4050-64
DM4060-64
0.16A typical
C-3
Specifications
Environmental Specifications
Ambient Temp (operating):
10° C to 50° C
10° C to 40° C (PM3755U2B with BB4050)
Relative Humidity (operating): 10% to 90% non-condensing
Altitude (operating):
3,000 meters (10,000 feet)
NOTE Maximum ambient temperature is 40°C when using a BB4050 module.
Memory
SmartRAID V controllers accept up to four memory modules. The total
memory installed must not exceed the maximum cache capacity of the
controller.
Capacities
Product
PM1554U2 w/RA4050
PM2554U2 w/RA4050
PM2654U2 w/RA4050
PM3754U2
PM3755F
PM3755U2B
Max Cache
64MB
64MB
64MB
256MB
256MB
256MB
Module Type
SIMM
SIMM
SIMM
SIMM
SIMM
DIMM
SIMMs
Non-DPT SIMMs to be installed in SmartRAID V controllers must
conform to the following specifications:
Speed:
Size:
Width:
Configuration:
Voltage:
60ns EDO
16MB or 64MB
32-bit or 36-bit
Single-sided, low-profile
5.0V
C-4
Specifications
DIMMs
DIMMs to be installed in SmartRAID V controllers must be DPT DM4050
or DM4060 memory modules only. Use of non-DPT DIMMs can result in
data corruption.
Battery
The following applies to the BB4050 module only:
Type:
Capacity:
Backup time:
Charge time:
Operating Temp:
Nickel Metal Hydride
3800mAH (minimum)
72 hours (with one 16MB memory module installed)
3 hours
10° C to 40° C
Exceeding the temperature limits can shorten the battery life and reduce
the battery capacity. Storage Manager displays the ambient temperature
for the controller which is reported by the onboard temperature sensor.
APPENDIX D:
SNMP
Simple Network Management Protocol
(SNMP) is an industry-wide standard
protocol designed to allow you to remotely manage your computer network.
By using the DPT SNMP subagent and
an SNMP-based management console,
you can get the DPT hardware configuration and information for your servers
from a remote workstation.
Introduction
What’s Included
System Requirements
Hardware Support
Management Consoles
Operating Systems and Platforms
Installation
DPT SNMP Support
Microsoft Windows SNMP
Operation
Software Overview
DPT SNMP Subagent
DPT SNMP Trap Broadcaster
D-1
D-2
Simple Network Management Protocol
Introduction
Simple Network Management Protocol (SNMP) lets you obtain
basic DPT hardware configuration and status information
from an SNMP-based management console. You can also
receive SNMP Traps, which are messages about changes in
the hardware status. These messages alert you to important
events that affect the DPT hardware and attached devices.
Simple Network Management Protocol
Simple Network Management Protocol (SNMP) is a group of
network management specifications, which includes the
protocol itself, the definition of the database and associated
concepts. SNMP is widely used and supported on a variety of
servers, workstations, bridges, routers and hubs. SNMP
support also includes other computer system resources, such
as SCSI, Fibre Channel and RAID controllers.
SNMP implementations vary from simple device information
display and collection of statistics, to complex configuration
management systems. SNMP is implemented by creating a
Management Information Base (MIB) that describes the
objects to be managed. SNMP establishes standards to access
and manage various types of data regarding attached
resources within a computer network.
Management Information Base
As with any network management system, the core component
of SNMP is the database containing the information about the
objects to be managed. For SNMP this is referred to as the
Management Information Base or MIB. A MIB is written using
the ASN.1 (Abstract Syntax Notation One) format as described
in ISO 8825-2. This format allows the exchange of structured
data, especially between application programs over networks,
by describing data structures in a form that is independent of
machine architecture and application software.
Simple Network Management Protocol
D-3
Every system resource to be managed is represented as an
object and the MIB is a collection of these objects. In a network
environment each system (workstation or server) maintains a
copy of the MIB containing the current status of the objects
which it defines. The MIB information is kept up-to-date by a
software agent. At the operating system level, there is a Master
Agent that controls the system MIB. Each vendor for a
managed resource (such as DPT) also supplies a subagent that
contributes its own MIB to the system MIB. This sub-agent
also responds to requests from the Master Agent for
information as needed.
NOTE
For additional security all of the DPT MIB data is
read-only. The DPT SNMP feature is intended only
for gathering inventory information and for
processing status and alert information.
Management Console
The SNMP management console is usually a client
workstation running SNMP-based management software. The
console software can be from any vendor who provides SNMP
management console software.
DPT MIB Information
The specific hardware and configuration information in the
DPT MIB includes groups for the following:
• DPT SCSI System Modules
• DPT SCSI Controllers
• DPT SCSI Busses
• DPT SCSI Devices
• DPT SCSI Arrays
• DPT SCSI Statistics
• DPT SCSI Events
D-4
Simple Network Management Protocol
Refer to the MIB itself for the exact contents of each of these
groups. The following lists are the groups’ contents:
DPT SCSI System Modules group identifies the versions,
creation date and various capabilities of the software modules.
The following specific modules are included:
DPT SNMP Subagent
DPT SNMP Engine
DPT SCSI Driver
DPT SCSI Logger
MIB revision information
DPT SCSI Controller group is a list of all DPT controllers in
the system. This group contains the following information for
each Controller:
Controller Number
Controller Vendor
Firmware Version
Address
IRQ
IRQ Type
Host Bus Type
Max Transfer Rate
Controller Modules (RAID, Caching, SIMMs/DIMMs)
DPT SCSI Bus group is a list of the SCSI Buses with a set of
parameters that describe and control a SCSI Bus. These
parameters include:
SCSI bus number
SCSI bus width
SCSI bus type
SCSI bus transfer rate
DPT controller SCSI ID (on this bus)
Simple Network Management Protocol
D-5
DPT SCSI Device group is a list of SCSI Devices managed by
the DPT SCSI subsystem and represents the physical
configuration. For each device in the system there is:
Device SCSI Address information:
Controller
Bus
ID
LUN
Device SCSI Inquiry data
Device Capacity and Block Size
Device RAID level and status (Optimal, Failed, etc.)
DPT SCSI Array group contains all the RAID-specific
information within the DPT SCSI subsystem. For each Array
there are:
Configuration parameters
SCSI address information
Background task information
DPT SCSI Statistics group contains statistical information
regarding the DPT SCSI controllers, SCSI devices and arrays.
For each SCSI controller, there are statistics on:
Cache Pages
Commands
Transfers
For Devices and Arrays there are statistics on:
Cache hits/misses
Stripe boundary crossings
Physical I/O commands (read/write)
DPT SCSI Events group is used to send traps on Controller
event log entries. These event log entries are converted to
SNMP traps and are identical to those used by the DPT
Logger/Broadcaster. The specific traps are listed in the MIB.
D-6
Simple Network Management Protocol
What’s Included
The DPT SNMP software components are included on the
diskettes in the DPT controller package. Specific components
vary by operating system.
The Windows 95/98 and Windows NT components are on the
DPT Windows diskette. The components are as follows:
• DPT Event Logger
• SNMP Event Logger Extensions
• DPT SNMP sub-agent
• DPT MIB
System Requirements
The DPT SNMP feature has the following system
requirements:
• The TCP/IP network protocol must be enabled on your
system. Refer to your operating system documentation for
information about installing the TCP/IP protocol.
• The SNMP Service for your operating system must be
installed.
Hardware Support
The DPT SNMP agent is designed to work with all DPT
SmartRAID V controllers. The specific controller information
available will vary depending on whether or not that device has
RAID or cache capability.
Management Consoles
The DPT SNMP agent conforms to the SNMP Version 1
specification. DPT has developed and tested this feature to
work with various operating systems that provide built-in
SNMP capability.
Simple Network Management Protocol
D-7
Operating System Support
The following table lists the environments in which DPT
hardware is manageable from an SNMP management console:
Operating System
Comments
Novell NetWare
Versions 4.11, 4.2 and 5.0
Windows 95/98
Windows NT 4.0
Workstation and Server versions
Installation of DPT SNMP Software
The following topics describe how to install DPT SNMP subagent software components and how to install SNMP support
for Microsoft Windows 95/98 and Windows NT 4.0.
NOTE
Ensure that your operating system has SNMP
support enabled before installing the DPT subagent
software.
Installing DPT SNMP Support
The DPT SNMP agent can be installed during the DPT
Storage Manager installation process. The setup utility
displays a Select Components dialog that includes a check box
for the DPT SNMP software. To install this feature, check the
box labeled SNMP System Agent.
By default this check box is not checked. Continue with the
installation according to the installation procedure for your
operating system in Chapter 5, “Software Installation”.
If your operating system is Windows NT, refer to the Microsoft
Windows NT documentation for information about installing
the SNMP Service.
Simple Network Management Protocol
D-8
Installing SNMP for Microsoft Windows 95
If your operating system is Windows 95, the SNMP Service
must be installed by following these steps:
1.
Insert the Windows 95 Installation CD-ROM into your
CD-ROM drive. If the Autostart window appears, close
the window.
2.
Launch the Network icon in Control Panel. Click Add
and select Service as the type of network component to
install.
3.
Click Add, then click the Have Disk button. Browse to
the ADMIN\NETTOOLS\SNMP directory on the CDROM. The file SNMP.INF will be selected. Click OK.
4.
Click OK in the Install From Disk window. The
Microsoft SNMP agent will be selected. Click OK to
complete the installation.
Installing SNMP for Microsoft Windows 98
If your operating system is Windows 98, the SNMP Service
must be installed by following these steps:
1.
Insert the Windows 98 Installation CD-ROM into your
CD-ROM drive. If the Autostart window appears, close
the window.
2.
Launch the Network icon in Control Panel. Click Add
and select Service as the type of network component to
install.
3.
Click Add, then click the Have Disk button. Browse to
the \TOOLS\RESKIT\NETADMIN directory on the CDROM. The file SNMP.INF will be selected. Click OK.
4.
Click OK in the Install From Disk window. The
Microsoft SNMP agent will be selected.
Click OK to complete the installation.
Simple Network Management Protocol
D-9
Operation
Before you can view information about your DPT hardware,
you must install the DPT Management Information Base
(MIB) into the Management Console database. During the
installation of the SNMP feature, the DPT MIB is installed by
default as DPTMGR\DPTSCSI.MIB. Refer to your SNMP
Management Console documentation for more information
about adding this MIB to your existing database.
The DPT SNMP sub-agent can be accessed from any SNMP
Management Console.
SNMP Software Overview
This section describes the architecture of the DPT SNMP
functions in a platform independent manner.
The DPT SNMP software architecture is an addition to the
current DPT software components. The components used in
the SNMP implementation are the DPT Engine, Event
Logger/Broadcaster and operating system device driver. The
DPT Engine is used to gather all information on the system
configuration and to perform the defined management
functions. An SNMP Trap broadcast module has been added to
the Event Logger to handle messages that are intended
specifically for the SNMP console.
D-10
Simple Network Management Protocol
TC P/IP
SNM P M anagem ent
Console
SNM P M aster Agent
DPT SCS I
Sub-Agent
SN MP Trap
Broadcast M odule
DPT Engine
DPT Event Logger
DPT Driver
D P T C ontroller
DPT SNMP Sub-Agent
The DPT SNMP sub-agent attaches to the operating systemspecific Master Agent to handle SNMP requests for objects
defined in the DPT MIB. It also broadcasts the DPT-specific
Traps to the designated management consoles.
The design and implementation of the DPT SNMP sub-agent
complies with the operating system-specific implementations
of the SNMP specification. This assures compatibility and
functionality within each operating system environment. On
the client side, all SNMP management consoles are supported.
Simple Network Management Protocol
D-11
DPT SNMP Trap Broadcaster Module
The SNMP Trap implementation uses the Broadcaster feature
of the DPT Event Logger. The Event Logger has individual
broadcast modules that have complete control over how
events are received from the Event Logger. The modules
define what events are to be sent and how they want to receive
the event information. Each broadcast module has control
over which events are to be broadcast, how they are broadcast
and where they are sent.
An SNMP Trap Broadcast module controls all DPT SNMP
Traps. The SNMP Trap Broadcast module registers with the
DPT Event Logger to receive all events generated for the DPT
subsystem. This module then forwards all events to the DPT
SNMP sub-agent for processing and delivery to the SNMP
Master Agent.
D-12
Simple Network Management Protocol
Glossary
This glossary provides brief definitions of
selected terms and acronyms used throughout this manual.
G-1
G-2
Glossary
Adapter
SCSI Host Bus Adapter or HBA
ANSI
American National Standards Institute.
Arbitrated Loop
See FC-AL.
Bridge Controller
A device which appears as a single device on the
bus, but which bridges to multiple devices.
Bridge controllers are typically used to
overcome limitations on the number of devices
that can be on a single bus or to control remote
subsystems.
Array Group
A group of disk drives that appear to the
computer as a single logical drive. RAID 0,
RAID 1 and RAID 5 array groups can be
composed of any combination of individual disk
drives.
Build
The operation of initializing a redundant array
by creating consistent redundant information.
In the case of a RAID 1 array, data from one
drive is copied to a second drive. In the case of
RAID 5 arrays, parity information is generated
by XORing the disk data.
ASIC
Application Specific Integrated Circuit. An
integrated circuit chip designed for a specific
task rather than a general purpose design such
as a microprocessor.
Burst
A term used to describe data words which are
transmitted as a single group across a bus
without interruption by another device.
ASPI
Advanced SCSI Programming Interface. A
protocol used by many SCSI drivers to
communicate with SCSI controllers under
MS-DOS, OS/2 and NetWare.
Asynchronous
Data transfer protocol that is not synchronized
to a set timing interval. Asynchronous SCSI data
transmitting devices must wait after each byte
for acknowledgment from the receiving device.
Either device can take as long as it needs to
send or acknowledge data. Asynchronous SCSI
has no defined maximum transfer rate but is
typically limited to between 1.5 and 3MHz.
BIOS
Basic Input Output System. A ROM-based
collection of device drivers and system start-up
routines which is provided as an integral part of
a computer. The BIOS provides enough
intelligence to enable the computer to
understand some simple keyboard commands
and load the operating system from disk on
power-up.
Bus Mastering
A method of data transfer that allows data to be
moved between a peripheral controller and
system memory without interaction with the
host CPU or a DMA controller. This technique
allows the peripheral controller to take control
of the system bus and move data at the
maximum speed supported by the bus.
Cache
A temporary storage area (usually RAM) for
data that would normally be accessed from a
slower storage device. A cache management
algorithm monitors the data access patterns
and selects which data from the slower device is
to be kept in the cache for quick access. Caches
are normally transparent to or hidden from the
accessing device.
CD-ROM
Compact Disk Read-Only Memory. A read-only
storage device that can retrieve data from a
removable optical storage disk similar to an
audio compact disk.
Command Queuing
A feature that allows multiple I/O commands to
be executed by a peripheral controller in a more
efficient order.
Glossary
Controller
A device that controls the transfer of data
between a computer and a peripheral device.
For example, disk drives, video displays,
keyboards, and printers all require controllers.
The DPT SmartRAID V is a controller for disk
drives and disk arrays. See also, HBA.
DAE
Dynamic Array Expansion. A feature that
enables you to increase your storage capacity
under Windows NT by adding one or more
drives to your RAID 0 and RAID 5 arrays while
your system remains online.
Degraded Mode
The mode of operation of a redundant array in
which it can continue to be accessed after a
member drive has failed. For read accesses,
data is synthesized from the other drives in the
array. For write accesses, data is stored on the
remaining drives in such a way that it can be
restored when the failed drive is replaced.
Device
SCSI devices include peripherals such as disk
and tape drives, optical devices, scanners, and
printers. SCSI controllers are also considered
SCSI devices.
Differential
An electrical signal protocol which transmits
information through a current loop rather than
by changes in voltage, thereby reducing the
susceptibility to electrical interference. High
Voltage Differential SCSI uses RS485
transceivers to transfer data at distances up to
25 meters (82 feet). Low Voltage Differential
SCSI (Ultra2) achieves a 25 meter cable length
without the need for external transceivers.
Digital High Availability Fault Bus
A proprietary Compaq Corporation (formerly
Digital Equipment Corp.) bus standard that
provides signals to host adapters for remote
system management and device fault
notification.
G-3
DIMM
Dual In-line Memory Module. A standard
packaging for RAM on a small circuit board with
a defined edge connector. See also SIMM.
Dirty
Refers to a cache page in which data has been
written or modified but which has not yet been
copied to the storage device. When the data has
been copied to disk, the page is said to be clean.
DMA
Direct Memory Access. A method of data
transfer that allows data to be moved between a
peripheral controller and system memory
without interaction with the host CPU. The data
may be moved by the peripheral controller itself
or by a separate DMA controller.
Driver
A system level software module that receives I/
O requests from higher levels within the
operating system and converts those requests
to the protocol required by a specific hardware
device such as a peripheral adapter.
Dual Loop
A method of providing Fibre Channel controller
redundancy. By forming two arbitrated loops on
a dual channel controller and connecting the
devices to both loops, fault tolerance is provided
for cable failure. In normal operation, the
controller will use both loops to communicate
with the devices, which maximizes controller
bandwidth and performance.
EBDA
Extended BIOS Data Area. An area of system
RAM that may be used by controller devices to
store and execute their BIOS code. Standard
rules define the way in which devices must
share this space.
ECC
Error Correcting Code. A method of generating
redundant information (checksums), which can
be used to detect and correct bit errors in
stored or transmitted data.
G-4
Glossary
EDO DRAM
Extended Data Output DRAM. A type of RAM
that improves memory read performance on
systems that are designed to use EDO memory.
Flash ROM
A nonvolatile memory device on the controller
that can be reprogrammed using special
software without removing it from the board.
Elevator Sorting
A method of sorting records or cache pages by
physical location on disk so that the information
can be written to disk with less seek and
rotational latency.
Flush
The action of writing all dirty data in the cache
to disk.
Events
Messages generated by DPT controllers for
detected fault conditions or subsystem status
changes.
Failed
The mode of operation of a drive or array in
which the drive or array, because of a
malfunction, can no longer be accessed.
Fast SCSI
The SCSI-2 standard for synchronous transfers
at up to 10MHz. This provides a transfer speed
of 10MB/sec for an 8-bit bus and 20MB/sec for a
16-bit bus. Devices that utilize these faster
timings are called Fast SCSI devices. The
original SCSI standard defined a 5MHz transfer
rate.
Fast-20 SCSI
See Ultra SCSI.
Fast-40 SCSI
See Ultra2 SCSI.
FC-AL
Fibre Channel Arbitrated Loop – A Fibre
Channel topology where up 126 devices are
connected in a serial (loop) fashion. The devices
negotiate individually for use of the bus. See
also, Fibre Channel.
Fibre Channel
A standard set of protocols for information
transfer. Fibre Channel provides for high speed
transfer of information over long cable lengths.
Fibre Channel controllers can transfer data at
up to 100MB/sec. among up to 126 devices over
30 meters of copper or 10 km of optical cable.
Hardware Array
A group of disk drives that are all members of
the same RAID level 1 or 5 array managed by a
DPT controller and which appears to the
computer as one storage device.
HBA
Host Bus Adapter – Refers to a DPT or other
controller card that provides host computer
access to the peripheral bus.
HDM
Hardware Device Module. The part of the I2O
split device driver that resides within the
firmware of the device.
Hit
A data access in which the requested data is
found in the cache.
Hit Ratio
The ratio of cache hits to total disk accesses. A
hit ratio of 100% means that all disk accesses
were serviced from the cache.
Hot Plug
The operation of adding or removing a device
from a bus while transactions involving other
devices are occurring over the bus.
Hot Spare
A disk drive that is assigned to automatically
replace a failed disk drive.
Hot Swap
The operation of removing a failed disk drive
that is a member of a redundant array and
replacing it with a good drive while transactions
involving other devices are occurring over the
bus. Synonymous with Hot Plug.
Glossary
HSSDC
High Speed Serial Data Connector. The industry
standard modular connector used on Fibre
Channel controllers.
I 2O
Intelligent I/O Architecture. A proprietary
specification that provides a standardized
software interface between peripheral devices
with built-in intelligence and a host operating
system.
ID
A numeric value used by SCSI devices to
address one another. (IDs can be from 0 to 7 or
0 to 15 for Wide SCSI.) Fibre Channel devices
can have ID numbers from 0 to 126.
Information View
The view in SMOR that provides a display of
device specific information when that device is
highlighted in the Tree View. The information
on this view may be further grouped into Tab
Pages.
Initiator
A SCSI device, such as a SmartRAID V
controller, that sends commands to SCSI Target
devices such as disk drives.
IRQ
Interrupt ReQuest. A hardware interrupt on a
computer. Systems that use the IBM/Intel
architecture have 16 IRQ lines used to signal
the CPU when a peripheral event has started or
terminated. Except for PCI devices, two devices
cannot use the same line. The PCI bus allows
IRQs to be shared between devices.
Latency
The time required by a device to access stored
data, excluding the data transfer time. Reducing
disk latency results in more I/O operations per
second being performed on a disk drive.
LBA
Logical Block Address. A method of addressing
storage blocks on a disk drive in a linear
fashion, rather than by Cylinder, Track and
Sector. This technique also overcomes storage
addressing limitations on some systems.
G-5
LED
Light Emitting Diode. An electronic device that
gives off light when power is applied.
LRU
Least Recently Used. A cache management
algorithm employed by SmartRAID V to
determine the next cache page to delete and
reuse when all pages have been filled with data.
By deleting the page in cache that has gone the
longest without an access, the algorithm
ensures that the most frequently accessed data
is resident in the cache.
LSU
Logical Storage Unit. A device on which the
computer can store and retrieve information.
This can be an individual disk drive or an array
group.
LUN
Logical Unit Number. Each SCSI device can
contain up to eight sub-devices or logical units.
The logical units are assigned from 0 to 7.
Typically, SCSI devices such as a disk or tape
drive contain only one subunit (LUN 0).
LVD SCSI
Low Voltage Differential SCSI. A SCSI-3
transmission protocol that provides long cable
lengths without the need for external high
voltage bus transceivers. The maximum
synchronous transfer rate is 40MHz. Also known
as Ultra2 SCSI.
MIA
Media Interface Adapter. A device that converts
electrical signals to optical fiber signals for
Fibre Channel connections.
Mirroring
A popular term for RAID 1. It refers to the
method of creating disk-fault tolerance by
storing duplicate information on pairs of drives.
Miss
A data access in which the requested data is not
found in cache.
G-6
Glossary
MTBF
Mean Time Between Failure. The average time
between expected failures of a device in a large
sample group of devices.
Multilevel RAID
A method of combining multiple RAID 1 or
RAID 5 arrays into a single array, providing
increased storage and performance in multiuser environments.
Non-Redundant Array
An Array Group with no fault tolerance
(RAID 0). If only one drive in a non-redundant
array fails, the entire array will fail.
NVRAM
Nonvolatile Random Access Memory. Hardware
memory that stores data even when system
power is turned off. SmartRAID controllers
store their setup information in NVRAM.
Operating System
Software that manages the resources of a
computer and provides the operating
environment for application programs.
Optimal
The mode of operation of a disk array in which
no drive failures have occurred.
OSM
Operating System Module. A system-level
software module that provides communication
between the host operating system and the
intelligent hardware controller. See also HDM.
Page
The smallest region of cache which can be
allocated to store data. Each page stores one
512-byte sector from disk.
Parity
A method of generating redundant information
which can be used to detect errors in stored or
transmitted data. Parity is used in standard
memory modules and over the peripheral bus to
detect data errors. Parity is used in RAID 5 disk
arrays to reconstruct flawed or missing data
sectors.
PCI
Peripheral Component Interconnect. An
intelligent computer bus specification that
provides automatic configuration of peripheral
cards. The bus supports 32-bit data paths at a
132 MB/sec data transfer rate or 64-bit data
paths at a 264 MB/sec data transfer rate.
PIO
Programmed Input/Output. A method of data
transfer where data is moved between a
peripheral controller and system memory by the
CPU.
Plug-and-Play
The ability to install peripheral cards or devices
without requiring a user to configure interrupts
or addressing for the device.
Plug-and-Play BIOS
A part of the system BIOS in PCI computers
that has the ability to automatically configure
PCI peripheral cards so that there are no
conflicts with other installed devices.
Predictive Caching
A method of analyzing disk I/O requests to
improve system performance by predicting
which data is likely to be needed in the near
future. The data is then read into the physical
cache.
RAID
Redundant Array of Independent Disks. A
method of combining disk drives into one logical
storage unit that provides disk-fault tolerance
and can operate at higher throughput levels
than a single disk drive.
Rebuild
The operation of restoring data belonging to a
failed member of a redundant disk array by
reconstructing the data from the other disks
and writing that data to a replacement disk
drive.
Redundant Array
A fault-tolerant disk array (for example, RAID 1
or RAID 5).
Glossary
SAF-TE
SCSI Accessed Fault-Tolerant Enclosure. A
specification (co-developed by nStor
Corporation, Inc. and Intel Corp.) which
provides a standardized method for monitoring
and reporting on the condition of disk drives,
power supplies and cooling systems used in
high availability LAN servers and storage
subsystems. The specification is independent of
hardware I/O cabling, operating systems, server
platforms, and RAID implementation because
the enclosure itself is treated as a device on the
SCSI bus.
Scatter/Gather
A feature that allows data to be transferred to or
from multiple noncontiguous areas of host
computer memory with a single I/O command.
SCSI
Small Computer Systems Interface. An ANSI
standard parallel interface for communication
with intelligent peripheral devices. The original
SCSI definition is defined in ANSI document
number X3.131-1986. The standard is now in its
third major revision.
SES
SCSI Enclosure Services. A SCSI-3 command
set that provides a method to manage and sense
the state of the power supplies, cooling devices,
displays, indicators, individual drives, and other
non-SCSI components installed in an enclosure.
See also SAF-TE.
SIMM
Single In-line Memory Module. A standard
packaging for RAM on a small circuit board with
a defined edge connector. See also DIMM.
Single-ended
An electrical signal protocol which transmits
information through changes in voltage. Singleended SCSI uses standard TTL signal-andground pairs to transmit information over the
SCSI bus. Single-ended SCSI cables can be up
to 3 meters in length.
G-7
S.M.A.R.T.
Self-Monitoring, Analysis and Reporting
Technology. A standard for disk drives that
monitor their own condition and report
potential problems. It is part of the ATA-3
specification for IDE drives and is also used on
some SCSI drives.
SPX
Sequenced Packet Exchange. The transport
layer protocol used by Novell NetWare
networks.
Stripe
A contiguous region of disk space. Stripes can
be as small as one sector or can be composed of
many contiguous sectors.
Striping
A method of distributing data evenly across all
drives in an array by concatenating interleaved
stripes from each drive.
Synchronous
A data transmission protocol which is
synchronized to a defined time interval.
Synchronous SCSI can transmit data faster than
asynchronous SCSI because the transmitting
device does not wait for acknowledgment of
each byte from the receiving device.
Tagged Command Queuing
A feature of SCSI-2 and SCSI-3 protocols that
allows SCSI commands to be executed out of
order.
Target
A SCSI device, such as a disk drive, that
receives and executes commands from a SCSI
Initiator device such as a DPT SmartRAID V
controller.
TCP/IP
Transmission Control Protocol/Internet
Protocol. A widely used network communication
protocol that is the standard for transmitting
data through the Internet.
G-8
Glossary
Termination
A method of matching the transmission
impedance of an electrical bus so that signal
reflections are minimized or eliminated.
TERMPWR
A signal line on the SCSI bus that supplies
electric power for SCSI bus terminators.
Throughput
A term used to describe the amount of data that
can be processed or transmitted by a system in
a given amount of time.
Tree View
The view in SMOR that provides a graphic
display of all controllers and devices on the
system in a hierarchical format. When a device
is highlighted in the Tree View, a corresponding
Information View is displayed that provides
specific information about that device.
Ultra SCSI
The SCSI-3 specification that defines
synchronous data transmission rates of up to
20MHz. This transfer rate provides data
transfer of up to 20MB/sec on an 8-bit bus and
40MB/sec on a 16-bit bus. Ultra SCSI is also
called Fast-20 SCSI.
Ultra2 SCSI
The SPI-2 specification that defines
synchronous data transmission rates of up to
40MHz. This transfer rate provides data
transfers of up to 40MB/sec on an 8-bit bus and
80MB/sec on a 16-bit SCSI bus. Ultra2 SCSI is
also called Fast-40 SCSI.
VHDCI
Very High Density Cable Interconnect. The
small industry standard 68-pin external
connector used on DPT SmartRAID V SCSI
controllers.
Wide SCSI
A SCSI protocol and signal definition that
provides a data path more than 8-bits wide.
Wide SCSI buses can support both 8- and 16-bit
data transfers. All SmartRAID V controllers use
a 16-bit SCSI interface.
WORM
Write Once Read Multiple – An optical storage
device similar to a CD-ROM that can write data
one time only to any location on a removable
optical disk. WORM drives are primarily used as
data archive devices because once written, the
data can never be erased.
Write-Back
A method postponing an actual data write
operation to a slower device, such as a disk
drive, by saving the data in a cache. The data
can then be written at a time when the device
would otherwise be idle.
Write-Through
Data is written to directly to the final destination
before a write operation is reported as
complete.
XOR
Exclusive OR. A logical operation performed on
two binary operands which yields a 0 for every
bit position where the operands are both 1 or
both 0, and a 1 in every bit position when the
operands are dissimilar.
Index
Index
address
controller 3-16, 4-13
I2O BIOS 7-36
logical device 6-7
alarm
audible 2-13, 6-39
disable 2-13, 6-39
testing 6-15
architecture, SmartRAID V 7-22
array group
building with SMOR 4-21
building with Storage Manager 6-23
caching configuration 6-18
copying data 6-23
creating 4-20, 6-22
deleting 4-23, 6-28
expanding 6-26
hot spare rebuild 6-41
information (SMOR) 4-17
modifying 6-24
naming 6-24
number of drives 6-19
rebuilding 6-21, 6-40
starting a build 6-28
task priority 6-43
ASPI protocol 7-22
asynchronous SCSI 7-29
AUTOEXEC.BAT
Windows 95/98 5-8, 5-10
Windows NT 5-7
battery module
features 2-10
installing 3-14
warning 2-10
battery specifications C-4
BB4050
configuration 6-11
features 2-10
installing 3-14
status 6-12
bootable devices 4-10
booting controller 3-16
bridge controller 7-27
Broadcaster 6-31, 6-34
NetWare 6-34
pager event messages 6-33
SCO Unix 6-35
SCO UnixWare 6-35
Windows NT 6-36
BSD/OS 5-1
building arrays
SMOR 4-2, 4-20
Storage Manager 6-23, 6-28
bus expansion module
compatability 3-9
features 2-8
installation 3-9
types 2-8, 3-9
bus mastering
byte/word alignment 7-24
PCI 7-25
B
C
background task priority 6-43
battery backup
auto write-through 6-11
initial calibration 6-12
maintenance 6-13
predictive failure warning 6-11
status messages 6-12
status window 6-11
cable
adapters 1-5
disk activity 3-15
external cable detect 7-29
Fibre Channel 3-3
guidelines 7-30, 7-32
included with product 1-5
LVD 7-32
SCSI cable installation 3-3
single-ended 7-30
Ultra2 7-32
A
I-1
I-2
cache
caching host reads 7-12
configuring 6-18, 7-6
DIMM, installing 3-13
DOS/Windows 3.1 setting 4-10
flushing 7-7, 7-10
operating system 7-7
operating system, size of 7-12
optimizing size 7-12
predictive 2-4, 6-19, 7-10
requirements 3-11, 3-13
SIMM, installing 3-11
virtual 7-6, 7-10
write-back vs. write-through 7-9
Century
expansion modules 2-8
features 2-7
SIMM installation 3-11
system requirements 1-5
types 2-7
cluster server
cables 3-20
enable/disable, support for 4-14
overview 3-18
requirements 3-19
SmartRAID V 3-20
command
overlapping 7-23
queuing 7-24
configuration, setting 4-19
controller
booting 3-16
ID 3-6
installling 3-15
intelligent 7-3
IRQ 3-16
NVRAM reset 3-17
SCSI ID, changing 4-14, 6-15
creating data redundancy 7-19
D
DAE. See Dynamic Array Expansion
data verification 6-42
Decade
cache 2-7
features 2-2, 2-7
SIMM installation 3-11
system requirements 1-5
Index
degraded mode 7-21
deleting arrays 6-28
device
caching configuration 6-18
capabilities 4-16
configuring IDs 3-6
formatting 4-24
SCSI capabilities 6-17
status info 4-16
differential SCSI 7-28
Digital Fault Bus G-3
DIMM
installing 3-13
restrictions 3-13
types 2-6
DM4050, installing 3-13
documentation
chapter summaries 1-2
overview 1-2
DOS/Windows 3.1 cache setting 4-10
drivers
BSD/OS 4.1 5-1
FreeBSD 5-1
Linux 5-24, 5-25
Novell NetWare 5-17
SCO Unix 5-14
Windows NT 5-2, 5-7
Dynamic Array Expansion. See See DAE
overview 6-24
performance 6-25
system requirements 6-25
using 6-26
E
EBDA. See extended BIOS data area
ECC
enabled LED 2-12
error LED 2-12
enable extended int13 4-10
environmental specifications C-3
event
levels 6-29
logging 6-30
send message to pager 6-33
SNMP D-11
viewing 6-34, 6-36
Index
event notification
broadcasters 6-34
configuring 6-31
NetWare 6-34
pager messages 6-33
SCO Unix 6-35
SCO UnixWare 6-35
SNMP D-11
Windows NT 6-36
expanding arrays 6-26
extended BIOS data area
definition of G-3
relocation, enabling 4-10
SMOR parameter 4-10
troubleshooting B-2, B-7
extended Int 13 4-10
F
Fibre Channel
cable 3-3
configuration 3-3, 3-6
connector 3-3, 7-35
controller 2-5, A-2
description of 7-35
device ID 3-6
expansion module 2-6, 2-8
firmware
file name 4-27
restoring from disk 4-28
saving to disk 4-27
upgrading 4-25, 6-15
flash
SMOR 4-25
Storage Manager 6-15
unsuccessful, recovering from B-6
flushing, cache
operating system 7-7
strategy 7-10
formatting disks
SMOR 4-24
Storage Manager 6-38
FreeBSD 5-1
full stripe write 7-19
H
hardware device module (HDM) 7-22, 7-23
Hot Spare 2-4, 6-41
assigning 4-23
icon 6-4
removing 4-23
I
I/O statistics 6-44
I2O
command processing 7-23
features 7-22
intelligent controllers 7-4
split driver model 7-22
I2O BIOS
configuration 7-36
flash upgrading 4-25
powerup message 7-36
icons
SMOR 4-4
Storage Manager 6-4
installation
battery module 3-14
bus expansion module 3-7, 3-10
configuring 3-3
controller 3-15
drivers
BSD/OS 4.1 5-1
FreeBSD 5-1
Linux 5-25
NetWare 5-17
SCO Unix 5-14
Windows 95/98 5-7
Windows NT 5-2, 5-7
memory modules 3-11, 3-13
overview 3-2
procedure 3-2
RA4050 3-8
RAID Accelerator 3-8
roadmap 1-3
SIMMs 3-11
SNMP D-7
Storage Manager
Linux 5-26
SCO UNIX 5-14
Windows 5-13
I-3
I-4
Int 13, extended 4-10
IRQ 3-16
K
keys, SMOR navigation 4-3
L
LED
battery module 2-13, A-12
cache status 2-12
Century/Decade, location of A-5
controller active 2-12
controller idle 2-11
disk activity 3-15
ECC enabled 3-11, 3-13
PM375x, location of A-3
power up sequence 2-11
RA4050, location of A-11
Linux
driver installation 5-25
Storage Manager 5-26
locality of reference 7-10
logical block addressing 4-10
logical device address 6-7
logical storage unit (LSU) 6-19
logical unit number (LUN) 7-27
LVD SCSI 7-28
Index
Millennium
connectors 2-6
DIMM installation 3-13
expansion modules 2-6
features 2-5
memory modules 2-6
SIMM installation 3-11
mirroring 7-15
module installation 3-7
MSCS. See Microsoft Cluster Server (MSCS)
multi-mode SCSI 7-29
N
naming arrays 6-24
Novell NetWare
boot controller, 4.11 5-18
boot controller, 4.2 5-20
boot controller, 5.0 5-22
driver installation 5-17
remote communication 5-24
secondary controller, 4.11 5-19
secondary controller, 4.2 5-21
secondary controller, 5.0 5-23
NT. See Windows NT
NVRAM
clearing 3-17
resetting 6-28
restoring defaults 3-17
M
O
management console, SNMP D-3
management information base D-2
memory capacities C-3
memory modules
Century/Decade 2-9
Millennium 2-6
memory specifications C-3
microprocessor 7-26
Microsoft Cluster Server
cables 3-20
enable/disable, support for 4-14
overview 3-18
SmartRAID V 3-20
system requirements 3-19
operating system
NetWare remote communication 5-24
operating system drivers
BSD/OS 4.1 5-1
FreeBSD 5-1
Linux 5-24, 5-25
NetWare 4.11 - booting 5-18
NetWare 4.11 - secondary 5-19
NetWare 4.2 - booting 5-20
NetWare 4.2 - secondary 5-21
NetWare 5 - booting 5-22
NetWare 5 - secondary 5-23
SCO
OpenServer 5 5-14
UNIX 3.2v4.2 5-14
UnixWare 7 5-15
Index
Windows 95/98 5-7
Windows NT 5-2
operating system services module (OSM)
description of 7-22
dynamic array expansion 6-25
Microsoft Windows 5-2, 5-6
P
pager event notification 6-33
parity group 7-18
partial stripe write 7-19
password, remote communication
changing 6-51
default 6-51
SCO servers 6-52
Windows 95/98 6-51
Windows NT 6-51
PCI bus 2-5, 7-25
performance
caching 7-6
intelligent I/O 7-3
pipelined parallel processing 7-5
physical configuration view 6-4
pipelined parallel processing 7-5
powerup
booting controller 3-16
I2O BIOS message 7-36
predictive caching 2-4, 6-19, 7-10
prefetch algorithm 7-11
R
RA4050
features 2-9
installation 3-8
SIMM installation 3-11
SIMMs 2-9
RAID
array groups 4-17
bad data list 7-20
bad parity list 7-20
degraded mode 7-21
dual-level 7-18
I/O errors 7-20
intelligent controller 7-3
parity 7-19
RAID 0 7-14
RAID 1 7-15
RAID 2 7-15
RAID 3 7-16
RAID 4 7-16
RAID 5 7-17
RAID Accelerator
features 2-9
installation 3-8
SIMM installation 3-11
read-ahead algorithm 7-10
read-modify-write 7-19
rebuilding arrays
button 6-21
hot spare 6-41
task priority 6-43
using SMOR 4-24
using Storage Manager 6-40
remote communication
command line parameter 6-53
configuration file, editing 6-50
configuring 6-49
default password 6-51
NetWare support 5-24
network attached 6-49
phonebook 6-55
servers 6-51
TCP/IP 6-50, 6-52
reset NVRAM 3-17
S
SAF-TE G-7
scan delay 4-10
SCSI
ANSI specification 7-26
asynchronous 7-29
cable installation 3-3
commands and messages 7-27
Common Command Set (CCS) 7-26
configurable parameters
controller termination 4-14
SCSI ID 6-15
termination 3-4
TERMPWR 4-13, 4-14
controller ID 3-6, 6-14
device capabilities 6-17
device IDs 3-6
I-5
I-6
devices, number of 3-6
differential 7-28
LUNs 7-27
LVD 7-28
LVD cabling 7-32
mixing LVD and non-LVD devices 7-29
multi-mode 7-29
residue reporting 7-28
single-ended cabling 7-30
singled-ended 7-28
synchronous 7-29
termination 3-4, 7-33
TERMPWR 4-13, 4-14, 7-34
transfer padding 7-28
transfer rate 7-30
Ultra2 cabling 7-32
wide 3-3, 3-6, 7-29
SES (SCSI Enclosure Services) G-7
SIMM
installing 3-11
restrictions 3-11
single-ended SCSI 7-28
SM4050, installing 3-11
SmartRAID V
architecture 7-22
configuration 6-13
controller ID 3-6
features 2-4
installation 3-15
modules, installing 3-7
product overview 2-2
products 2-3
SMOR
array information 4-18
controller configuration 4-13
controller information 4-11
device capabilities 4-16
device information 4-15
device status 4-16
icons 4-4
information view 4-6, 4-11
keyboard reference 4-3
menu reference 4-4
navigation 4-5, 4-6
overview 4-2
Index
parameters
bootable devices 4-10
DOS/Windows 3.1 cache 4-10
EBDA relocation 4-10
enable extended int13 4-10
scan delay 4-10
termination 3-4, 4-14
TERMPWR 4-14
running 4-8
screen layout 4-5
tree view 4-6
upgrading 4-25
SNMP
components D-6, D-10
description of D-2
installing D-7
management console D-3
MIB D-2
MIB, DPT specific D-3
operating system support D-7
operation D-9
software overview D-9
sub-agent D-10
system requirements D-6
trap broadcaster D-11
status
array 6-8
device 4-16
drive 6-8, 6-17
failed drive 6-39, 6-41
Storage Manager
array group
creating 6-22
deleting 6-28
modifying 6-24
naming 6-24
background task priority 6-43
Broadcaster 6-31, 6-34
NetWare 6-34
SCO Unix 6-35
SCO UnixWare 6-35
Windows NT 6-36
configuration files 6-28
controller information 6-10
device information 6-17
event logging 6-30
Index
event notification 6-31, 6-34
NetWare 6-34
SCO Unix 6-35
SCO UnixWare 6-35
Windows NT 6-36
I/O statistics 6-44
Linux installation 5-26
modem operation 6-48
network operation 6-3, 6-48
overview 6-2
physical configuration view 6-4
platforms supported 5-1
remote operation 6-3, 6-48
running 6-3
SCO UNIX installation 5-14
system requirements 6-2
Windows installation 5-13
striping 7-13
sub-agent, SNMP D-10
SX405x
connectors 2-9
features 2-8
installation 3-9
types 2-8
synchronous SCSI 7-29
system requirements 1-5
T
temperature 6-10, C-3
termination 7-33
configuration 3-4
enabling 4-14
power 4-14, 7-34
TERMPWR 4-13, 4-14, 7-34
transfer padding and residue reporting 7-28
U
Ultra2 SCSI 7-28, 7-32
upgrading firmware 4-25, 6-15
V
Verify 6-42
virtual cache 7-10
W
wide SCSI
IDs 3-6
termination 3-4
with Narrow (8-bit) devices 3-3
Windows 95/98 5-7
Windows NT 5-2
X
XOR 7-19
I-7
Download PDF

advertising