Administration guide for WLB Xenserver 6.5
Citrix XenServer Workload Balancing 6.5 Administrator's
Guide
Published February 2015
1.0 Edition
Citrix XenServer Workload Balancing 6.5 Administrator's Guide
Copyright © 2015 Citrix Systems. Inc. All Rights Reserved.
Version: 6.5
Citrix, Inc.
851 West Cypress Creek Road
Fort Lauderdale, FL 33309
United States of America
Disclaimers
This document is furnished "AS IS." Citrix, Inc. disclaims all warranties regarding the contents of this document,
including, but not limited to, implied warranties of merchantability and fitness for any particular purpose. This
document may contain technical or other inaccuracies or typographical errors. Citrix, Inc. reserves the right to
revise the information in this document at any time without notice. This document and the software described
in this document constitute confidential information of Citrix, Inc. and its licensors, and are furnished under a
license from Citrix, Inc.
Citrix Systems, Inc., the Citrix logo, Citrix XenServer and Citrix XenCenter, are trademarks of Citrix Systems, Inc.
and/or one or more of its subsidiaries, and may be registered in the United States Patent and Trademark Office
and in other countries. All other trademarks and registered trademarks are property of their respective owners.
Trademarks
Citrix®
XenServer ®
XenCenter ®
Contents
1. About ............................................................................................................. 1
1.1. What's New in 6.5? ..................................................................................................... 1
1.1.1. New Features in 6.5 .......................................................................................... 1
1.2. Finding Workload Balancing Documentation .................................................................. 1
2. Introducing Workload Balancing ................................................................... 2
2.1. Workload Balancing Overview ...................................................................................... 2
2.1.1. Reports on Workloads ....................................................................................... 3
2.1.2. Workload Balancing Basic Concepts ................................................................... 3
2.1.3. Pool Requirements ............................................................................................ 3
3. Using Workload Balancing ............................................................................ 4
3.1. Introduction to Basic Tasks ........................................................................................... 4
3.2. Choosing the Best Host for a VM .................................................................................. 4
3.2.1. To start a virtual machine on the optimal server ................................................. 5
3.2.1.1. To resume a virtual machine on the optimal server .................................. 6
3.3. Accepting Optimization Recommendations .................................................................... 7
3.3.1. To accept an optimization recommendation ....................................................... 7
3.3.2. Understanding WLB Recommendations under High Availability ............................ 7
4. Generating Workload Reports ....................................................................... 9
4.1. Overview of Workload Reports ..................................................................................... 9
4.1.1. Generating and Managing Workload Balancing Reports ....................................... 9
4.1.1.1. To generate a Workload Balancing report ................................................. 9
4.1.1.2. To navigate in a Workload Balancing report ............................................ 10
4.1.1.3. To print a Workload Balancing report ..................................................... 10
4.1.1.3.1. To export a Workload Balancing report ........................................ 10
4.1.2. Workload Balancing Report Glossary ................................................................ 11
4.1.2.1. Chargeback Utilization Analysis .............................................................. 11
4.1.2.2. Host Health History .............................................................................. 13
4.1.2.3. Pool Optimization Performance History .................................................. 13
iii
4.1.2.4. Pool Audit Trail ..................................................................................... 14
4.1.2.4.1. Report Contents ........................................................................ 14
4.1.2.4.2. Audit Log Event Names .............................................................. 14
4.1.2.5. Pool Health .......................................................................................... 15
4.1.2.6. Pool Health History ............................................................................... 16
4.1.2.7. Pool Optimization History ..................................................................... 16
4.1.2.8. Virtual Machine Motion History ............................................................ 17
4.1.2.9. Virtual Machine Performance History .................................................... 17
5. Managing Workload Balancing Features and Settings ................................. 18
5.1. Changing Workload Balancing Settings ........................................................................ 18
5.1.1. Adjusting the Optimization Mode ..................................................................... 19
5.1.1.1. Optimizing and Managing Power Automatically ...................................... 20
5.1.1.1.1. Applying Recommendations Automatically .................................. 21
5.1.1.1.2. Enabling Workload Balancing Power Management ....................... 21
5.1.1.1.3. Designing Environments for Power Management and VM
Consolidation ........................................................................................... 22
5.1.1.1.4. To apply optimization recommendations automatically ................. 23
5.1.1.1.5. To select hosts for power management ....................................... 23
5.1.1.2. Understanding When Workload Balancing Makes Recommendations....... 24
5.1.1.2.1. Optimization and Consolidation Process ...................................... 25
5.1.1.2.2. Changing the Critical Thresholds ................................................. 27
5.1.1.2.3. Tuning Metric Weightings ........................................................... 28
5.1.1.3. Excluding Hosts from Recommendations ................................................ 29
5.1.1.3.1. To exclude hosts from Workload Balancing .................................. 30
5.1.1.4. Controlling Automated Recommendations ............................................. 30
5.1.1.4.1. Setting Conservative or Aggressive Automated
Recommendations .................................................................................... 30
5.2. To Modify the Pool Audit Trail Granularity Setting ........................................................ 35
6. Administering Workload Balancing ............................................................. 36
6.1. Administering and Maintaining Workload Balancing ..................................................... 36
6.1.1. Connecting to the Workload Balancing Virtual Appliance ................................... 36
iv
6.1.1.1. To connect your pool to the Workload Balancing virtual appliance ........... 37
6.1.1.2. To obtain the IP address for the WLB virtual appliance ............................ 39
6.1.1.3. Workload Balancing Access Control Permissions ..................................... 39
6.1.2. Determining the Status of the Workload Balancing Virtual Appliance .................. 39
6.1.3. Reconfiguring a Pool to Use Another WLB Appliance ......................................... 39
6.1.4. Updating Workload Balancing Credentials ......................................................... 40
6.1.5. Changing the Workload Balancing IP Address .................................................... 41
6.1.6. Stopping Workload Balancing ........................................................................... 41
6.1.7. Entering Maintenance Mode with Workload Balancing Enabled .......................... 42
6.1.7.1. To enter maintenance mode with Workload Balancing enabled ................ 42
6.1.8. Increasing the Workload Balancing Disk Size ..................................................... 42
6.1.9. Removing the Workload Balancing Virtual Appliance ......................................... 43
6.1.10. Managing the Workload Balancing Database ................................................... 43
6.1.10.1. Accessing the Database ....................................................................... 44
6.1.10.2. Controlling Database Grooming ........................................................... 44
6.1.10.2.1. Database Grooming Parameters ................................................ 44
6.1.10.3. Changing the Database Password ........................................................ 45
6.1.10.4. Archiving Database Data ..................................................................... 45
6.1.10.4.1. Enabling client authentication to the database ........................... 45
6.1.10.4.2. Modifying the Database Configuration Files ............................... 45
6.1.10.5. Changing the Database Maintenance Window ...................................... 46
6.1.11. Customizing Workload Balancing .................................................................... 47
6.2. Upgrading Workload Balancing ................................................................................... 48
6.2.1. Upgrading Workload Balancing with Internet Access .......................................... 49
6.2.2. Upgrading Workload Balancing without Direct Internet Access ........................... 49
6.2.3. Checking Workload Balancing and CentOS Version Numbers .............................. 52
7. Appendix A: XenServer Workload Balancing Troubleshooting .................... 53
7.1. Troubleshooting Workload Balancing .......................................................................... 53
7.1.1. General Troubleshooting Tips ........................................................................... 53
7.1.2. Error Messages ............................................................................................... 53
7.1.3. Issues Entering Workload Balancing Credentials ................................................ 53
v
7.1.4. Issues with Firewalls ........................................................................................ 53
7.1.5. Losing the Connection to Workload Balancing ................................................... 54
7.1.6. Workload Balancing Connection Errors ............................................................. 54
7.1.7. Workload Balancing Stops Working .................................................................. 54
7.1.8. Issues Changing Workload Balancing Servers ..................................................... 55
8. Appendix B: Workload Balancing Commands ............................................. 56
8.1. Service Commands ..................................................................................................... 56
8.1.1. Logging in to the Workload Balancing Virtual Appliance ..................................... 56
8.1.2. wlb restart ...................................................................................................... 56
8.1.3. wlb start ......................................................................................................... 56
8.1.4. wlb stop ......................................................................................................... 56
8.1.5. wlb status ....................................................................................................... 56
8.1.6. Modifying the Workload Balancing Configuration Options .................................. 57
8.1.7. Editing the Workload Balancing Configuration File ............................................. 57
8.1.8. Increasing the Detail in the Workload Balancing Log .......................................... 58
8.2. Workload Balancing XE Commands ............................................................................. 59
8.2.1. pool-initialize-wlb ............................................................................................ 59
8.2.2. pool-param-set other-config ............................................................................ 59
8.2.3. pool-retrieve-wlb-diagnostics ........................................................................... 60
8.2.4. host-retrieve-wlb-evacuate-recommendations .................................................. 60
8.2.5. vm-retrieve-wlb-recommendations ................................................................... 60
8.2.6. pool-certificate-list .......................................................................................... 60
8.2.7. pool-certificate-install ...................................................................................... 60
8.2.8. pool-certificate-sync ........................................................................................ 60
8.2.9. pool-param-set ................................................................................................ 60
8.2.10. pool-deconfigure-wlb ..................................................................................... 61
8.2.11. pool-retrieve-wlb-configuration ...................................................................... 61
8.2.12. pool-retrieve-wlb-recommendations ............................................................... 61
8.2.13. pool-retrieve-wlb-report ................................................................................ 61
8.2.14. pool-send-wlb-configuration ........................................................................... 62
9. Appendix C: Configuring Certificates for Workload Balancing ..................... 64
vi
9.1. Overview ................................................................................................................... 64
9.2. Configuring XenServer to Verify the Self-Signed Certificate ........................................... 65
9.3. Configuring XenServer to Verify a Certificate-authority Certificate ................................. 66
9.3.1. Task 1: Obtaining a Certificate-authority Certificate ........................................... 66
9.3.2. Task 2: Specifying the New Certificate .............................................................. 68
9.3.3. Task 3: Importing the Certificate Chain into the Pool ......................................... 68
Index ................................................................................................................ 70
vii
Chapter 1. About
This document explains how to run the XenServer Workload Balancing virtual appliance. To get started and import
the Workload Balancing virtual appliance, see the XenServer Workload Balancing 6.5 Quick Start Guide.
Note:
Workload Balancing is available for XenServer Enterprise Edition customers or those who have
access to XenServer through their XenDesktop entitlement. To learn more about XenServer
editions, and to find out how to upgrade, visit the Citrix website here. For detailed information
on Licensing - refer to CTX141511 - XenServer 6.5 Licensing FAQ.
1.1. What's New in 6.5?
Previously removed in XenServer 6.2, Workload Balancing (WLB) is reintroduced and enhanced with XenServer
6.5.
1.1.1. New Features in 6.5
Workload Balancing 6.5 includes the following enhanced features.
• Pool Audit Trail. The enhanced Pool Audit Trail feature allows you to specify the granularity of the audit log
report. Customers can also search and filter the audit trail logs by specific users, objects, and by time
• Online Upgrade. Customers can now apply upgrades directly from the live SLB Server by the yum update
mechanism, or instead download the upgraded RPM packages from Citrix.com
1.2. Finding Workload Balancing Documentation
The Workload Balancing documentation also includes:
• XenServer Workload Balancing 6.5 Quick Start provides information about downloading and importing the
Workload Balancing virtual appliance.
• In XenCenter, press F1 to access the comprehensive online help.
1
Chapter 2. Introducing Workload Balancing
This chapter provides introductions to the following topics:
• Workload Balancing at a high level
• Workload Reports
• Basic concepts of Workload Balancing
2.1. Workload Balancing Overview
Note:
Workload Balancing is available for XenServer Enterprise Edition customers or those who have
access to XenServer through their XenDesktop entitlement. To learn more about XenServer
editions, and to find out how to upgrade, visit the Citrix website here. For detailed information
on Licensing - refer to CTX141511 - XenServer 6.5 Licensing FAQ.
Workload Balancing is a XenServer component, packaged as a virtual appliance, that:
• Creates reports about VM performance in your XenServer environment
• Evaluates resource utilization and locates virtual machines on the best possible hosts in the pool for their
workload's needs
Even if you do not want to use Workload Balancing to balance your virtual machines, you may want to run
it anyway for the workload reporting functionality. When deployed to manage virtual machine workloads,
Workload Balancing can:
• Balance virtual-machine workloads across hosts in a XenServer resource pool
• Determine the best host on which to start a virtual machine
• Determine the best host on which to resume a virtual machine that you powered off
• Determine the best host to move a virtual machine to when a host fails
• Determine the optimal server for each of the host's virtual machines when you put a host into or take a host
out of Maintenance Mode
Depending on your preference, Workload Balancing can accomplish these tasks automatically or prompt you to
accept its rebalancing and placement recommendations. You can also configure Workload Balancing to power
off hosts automatically at specific times of day (for example, to save power at night).
Workload Balancing functions by evaluating the utilization of virtual machines across a pool and, when a host
exceeds a performance threshold, relocating the VM to a less-taxed host in the pool. To rebalance workloads,
Workload Balancing moves VMs to balance the resource utilization on hosts.
To ensure the rebalancing and placement recommendations align with your environment's needs, you can
configure Workload Balancing to optimize workloads for either resource performance or to maximize the number
of virtual machines that fit on hosts. These optimization modes can be configured to change automatically at
predefined times or stay the same at all times. For additional granularity, you can fine-tune the weighting of
individual resource metrics (CPU, network, and memory).
To help you perform capacity planning, Workload Balancing provides historical reports about host and pool
health, optimization and virtual-machine performance, and virtual-machine motion history.
2
2.1.1. Reports on Workloads
Because Workload Balancing captures performance data, you can also use this component to generate reports,
known as Workload Reports, about your virtualized environment.
Workload Reports provide data for a pool or host's health, for auditing, optimizations, and placement (or motion)
history. Also, you can run a chargeback report that shows virtual machine usage and can help you measure and
assign costs.
To run reports, you do not need to configure for Workload Balancing to make placement recommendations or
move virtual machines. However, you do need to configure the Workload Balancing component and, ideally, set
Critical thresholds to values that reflect the point at which the performance of the hosts in your pool degrades.
For more information, see Chapter 4, Generating Workload Reports.
2.1.2. Workload Balancing Basic Concepts
When virtual machines are running, they consume computing resources on the physical host, such as CPU,
Memory, Network Reads, and Network Writes. For example, some virtual machines, depending on their workload
(that is, the applications running on a virtual machine and their user transactions), might consume more CPU
resources than other virtual machines on the same host. Naturally, the combined resource consumption of all
virtual machines on a host reduces the host's available resources.
Workload Balancing captures data for resource performance on virtual machines and physical hosts and stores it
in a database. Workload Balancing uses this data, combined with the preferences you set, to provide optimization
and placement recommendations.
Optimizations are a way in which hosts are "improved" to align with your goals: Workload Balancing makes
recommendations to redistribute the virtual machines across hosts in the pool to increase either performance
or density. When Workload Balancing is making recommendations, it makes them in light of its ultimate
goal: to create balance or harmony across the hosts in the pool. When Workload Balancing acts on these
recommendations, the action is known as an optimization.
Within a Workload Balancing context:
• Performance refers to the usage of physical resources on a host (for example, the CPU, memory, network, and
disk utilization on a host). When you set Workload Balancing to maximize performance, it recommends placing
virtual machines to ensure the maximum amount of resources are available for each virtual machine.
• Density refers to the number of virtual machines on a host. When you set Workload Balancing to maximize
density, it recommends placing virtual machines to ensure they have adequate computing power so you can
reduce the number of hosts powered on in a pool.
Workload Balancing does not conflict with settings you already specified for High Availability: Citrix designed
these features to work together.
2.1.3. Pool Requirements
To balance a pool with Workload Balancing, the hosts in the pool must meet the requirements for XenMotion
(live migration), including:
• Shared remote storage
• Similar processor configurations
• Gigabit Ethernet
If the hosts do not meet these requirements, Workload Balancing cannot migrate the virtual machines in the pool.
3
Chapter 3. Using Workload Balancing
This chapter provides information about the following topics:
• Using Workload Balancing to start virtual machines on the best possible host
• Accepting the recommendations Workload Balancing issues to move virtual machines to different hosts
Note:
Workload Balancing is available for XenServer Enterprise Edition customers or those who have
access to XenServer through their XenDesktop entitlement. To learn more about XenServer
editions, and to find out how to upgrade, visit the Citrix website here. For detailed information
on Licensing - refer to CTX141511 - XenServer 6.5 Licensing FAQ.
3.1. Introduction to Basic Tasks
Workload Balancing is a powerful XenServer component that includes many features designed to optimize the
workloads in your environment, such as host power management, the ability to schedule optimization-mode
changes, and running reports. In addition, you can fine-tune the criteria Workload Balancing uses to make
optimization recommendations.
However, when you first begin using Workload Balancing, there are two main tasks you will probably use
Workload Balancing for on a daily (or regular) basis:
• Determining the best host on which to run a virtual machine
• Accepting Workload Balancing optimization recommendations
Running reports about the workloads in your environment, which is described in Chapter 4, Generating Workload
Reports, is another frequently used task.
Determining the best host on which to run a virtual machine
Determining the host on which to start and run a virtual machine, which is known as virtual-machine placement,
is handy when you want to restart a powered off virtual machine and when you want to migrate a virtual machine
to a different host. Placement recommendations may also be useful in XenDesktop environments.
Accepting Workload Balancing recommendations
After Workload Balancing is running for a while, it begins to make recommendations about ways in which you
can improve your environment. For example, if your goal is to improve virtual-machine density on hosts, at some
point, Workload Balancing may issue a recommendation to consolidate virtual machines on a host. Assuming
you are not running in automated mode, you can choose to either accept this recommendation (and apply it)
or just simply ignore it.
Both of these tasks, and the way in which you perform them in XenCenter, are explained in more depth in the
sections that follow.
Important:
After Workload Balancing is running for a period of time, if you do not receive optimal
placement recommendations, Citrix strongly recommends you evaluate your performance
thresholds as described in Section 5.1.1.2, “Understanding When Workload Balancing Makes
Recommendations”. It is critical to set Workload Balancing to the correct thresholds for your
environment or its recommendations might not be appropriate.
3.2. Choosing the Best Host for a VM
When Workload Balancing is enabled and you restart a virtual machine that is offline, XenCenter provides
recommendations to help you determine the optimal physical host in the resource pool on which to start a virtual
machine. The recommendations are also known as star ratings since stars are used to indicate the best host.
4
This illustration shows a screen capture of the Start On Server feature. More stars appear beside
host17 since this is the optimal host on which to start the VM. host16 does not have any stars
beside it, which indicates that host is not recommended. However, since host16 is enabled the
user can select it. host18 is greyed out due to insufficient memory, so the user cannot select it.
The term optimal refers to the physical server best suited to hosting your workload. There are several factors
Workload Balancing uses when determining which host is optimal for a workload:
• The amount of resources available on each host in the pool. When a pool is running in Maximum Performance
mode, Workload Balancing tries to balance the virtual machines across the hosts in the pool so that all virtual
machines have good performance. When a pool is running in Maximum Density mode, Workload Balancing
tries to place virtual machines onto hosts as densely as possible while ensuring the virtual machines have
sufficient resources.
• The optimization mode in which the pool is running (Maximum Performance or Maximum Density). When a
pool is running in Maximum Performance mode, Workload Balancing tries to place virtual machines on hosts
with the most resources available of the type the VM requires. When a pool is running in Maximum Density
mode, Workload Balancing tries to place virtual machines on hosts that already have virtual machines running
so that virtual machines are running on as few hosts as possible.
• The amount and type of resources the VM requires. After Workload Balancing has monitored a virtual machine
for a while, it uses the VM metrics it gathered to make placement recommendations according to the type of
resources the virtual machine requires. For example, Workload Balancing may select a host with less available
CPU but more available memory if it is what the VM requires (based on its past performance history).
When Workload Balancing is enabled, XenCenter provides ratings to indicate the optimal hosts for starting a
virtual machine, including when you want to start the virtual machine when it is powered off or suspended and
when you want to migrate the virtual machine to a different host (Migrate and Maintenance Mode).
When you use these features with Workload Balancing enabled, host recommendations appear as star ratings
beside the name of the physical host. Five empty stars indicates the lowest-rated (least optimal) server. When it
is not possible to start or migrate a virtual machine to a host, when you use the menu command for one of the
placement features, the host name is greyed out and the reason it cannot accept the VM appears beside it.
In general, Workload Balancing functions more effectively and makes better, less frequent optimization
recommendations if you start virtual machines on the hosts it recommends (that is, by using one of the placement
features to select the host with the most stars beside it).
3.2.1. To start a virtual machine on the optimal server
1. In the Resources pane of XenCenter, select the virtual machine you want to start.
2. From the VM menu, select Start on Server and then select one of the following:
• Optimal Server. The optimal server is the physical host that is best suited to the resource demands of the
virtual machine you are starting. Workload Balancing determines the optimal server based on its historical
records of performance metrics and your placement strategy. The optimal server is the server with the most
stars.
5
• One of the servers with star ratings listed under the Optimal Server command. Five stars indicates the
most-recommended (optimal) server and five empty stars indicates the least-recommended server.
Tip:
You can also select Start on Server by right-clicking the virtual machine you want to start in
the Resources pane.
3.2.1.1. To resume a virtual machine on the optimal server
1. In the Resources pane of XenCenter, select the suspended virtual machine you want to resume.
2. From the VM menu, select Resume on Server and then select one of the following:
• Optimal Server. The optimal server is the physical host that is best suited to the resource demands of the
virtual machine you are starting. Workload Balancing determines the optimal server based on its historical
records of performance metrics and your placement strategy. The optimal server is the server with the most
stars.
• One of the servers with star ratings listed under the Optimal Server command. Five stars indicates the
most-recommended (optimal) server and five empty stars indicates the least-recommended server.
Tip:
You can also select Resume on Server by right-clicking the suspended virtual machine in the
Resources pane.
6
3.3. Accepting Optimization Recommendations
Workload Balancing provides recommendations about ways you can migrate virtual machines to optimize your
environment. Optimization recommendations appear in the WLB tab in XenCenter.
This illustration shows a screen capture of the Optimization Recommendations list, which appears on the WLB
tab. The Operation column displays the behavior change suggested for that optimization recommendation.
The Reason column displays the purpose of the recommendation. This screen capture shows an optimization
recommendation for a virtual machine, HA-prot-VM-7, and a host, host17.domain4.bedford4.ctx4.
Optimization recommendations are based on the:
• Placement strategy you select (that is, the optimization mode), as described in Section 5.1.1, “Adjusting the
Optimization Mode”.
• Performance metrics for resources such as a physical host's CPU, memory, network, and disk utilization.
• The role of the host in the resource pool. When making placement recommendations, Workload Balancing
considers the pool master for virtual machine placement only if no other host can accept the workload.
(Likewise, when a pool is operating in Maximum Density mode, Workload Balancing considers the pool master
last when determining the order in which to fill hosts with virtual machines.)
The optimization recommendations display the name of the virtual machine that Workload Balancing
recommends relocating, the host it currently resides on, and the host Workload Balancing recommends as
the machine's new location. The optimization recommendations also display the reason Workload Balancing
recommends moving the virtual machine (for example, "CPU" to improve CPU utilization). When Workload
Balancing power management is enabled, Workload Balancing also displays optimization recommendations
(specifically consolidation recommendation) for hosts it recommends powering on or off.
After you click Apply Recommendations, XenServer performs all operations listed in the Optimization
Recommendations list.
Tip:
You can find out the optimization mode for a resource pool by selecting the pool in XenCenter
and checking the Configuration section of the WLB tab.
3.3.1. To accept an optimization recommendation
1. In the Resources pane of XenCenter, select the resource pool for which you want to display recommendations.
2. Click the WLB tab. If there are any recommended optimizations for any virtual machines on the selected
resource pool, they display in the Optimization Recommendations section of the WLB tab.
3. To accept the recommendations, click Apply Recommendations. XenServer begins performing all the
operations listed in the Operations column of the Optimization Recommendations section.
After you click Apply Recommendations, XenCenter automatically displays the Logs tab so you can see the
progress of the virtual machine migration.
3.3.2. Understanding WLB Recommendations under High Availability
If you have Workload Balancing and the XenServer High Availability feature enabled in the same pool, it is
helpful to understand how the two features interact. Workload Balancing is designed not to interfere with High
7
Availability. If there is a conflict between a Workload Balancing recommendation and a High Availability setting,
the High Availability setting always takes precedence. In practice, this means:
• When you attempt to start a virtual machine on a host, Workload Balancing will not give you star ratings if the
virtual machine cannot be started on the host without violating the High Availability plan.
• Workload Balancing will not automatically power off any hosts beyond the number specified in the Failures
allowed box in the Configure HA dialog.
• However, Workload Balancing may still make recommendations to power off more hosts than the number
of host failures to tolerate. (For example, Workload Balancing will still make a recommendation to power off
two hosts when High Availability is only configured to tolerate one host failure.) However, when you attempt
to apply the recommendation, XenCenter may display an error message stating that High Availability will
no longer be guaranteed.
• When Workload Balancing is running in automated mode and has power management enabled, any
recommendations that exceed the number of host failures to tolerate are ignored. In this situation, if you
look in the Workload Balancing log, you will see a message that says a power-management recommendation
was not applied because High Availability is enabled.
8
Chapter 4. Generating Workload Reports
This chapter provides information about using the Workload Balancing component to generate reports about
your virtualized environment, including reports about hosts and virtual machines. Specifically, this chapter
provides information about the following:
• How to generate reports
• What workload reports are available
Note:
Workload Balancing is available for XenServer Enterprise Edition customers or those who have
access to XenServer through their XenDesktop entitlement. To learn more about XenServer
editions, and to find out how to upgrade, visit the Citrix website here. For detailed information
on Licensing - refer to CTX141511 - XenServer 6.5 Licensing FAQ.
4.1. Overview of Workload Reports
The Workload Balancing reports can help you perform capacity planning, determine virtual server health, and
evaluate the effectiveness of your configured threshold levels.
Workload Balancing lets you generate reports on three types of objects: physical hosts, resource pools, and virtual
machines. At a high level, Workload Balancing provides two types of reports:
• Historical reports that display information by date
• "Roll up" style reports, which provide a summarizing overview of an area
Workload Balancing provides some reports for auditing purposes, so you can determine, for example, the number
of times a virtual machine moved.
You can use the Pool Health report to evaluate the effectiveness of your optimization thresholds. While Workload
Balancing provides default threshold settings, you might need to adjust these defaults for them to provide
value in your environment. If you do not have the optimization thresholds adjusted to the correct level for your
environment, Workload Balancing recommendations might not be appropriate for your environment.
To generate a Workload Balancing report, the pool must be running Workload Balancing. Ideally, the pool should
have been running Workload Balancing for a couple of hours or long enough to generate the type of data you
want to display in the reports.
4.1.1. Generating and Managing Workload Balancing Reports
This topic explains how to generate, navigate, print, and export Workload Balancing reports.
4.1.1.1. To generate a Workload Balancing report
1. In XenCenter, from the Pool menu, select View Workload Reports.
Tip:
You can also display the Workload Reports screen from the WLB tab by clicking the Reports
button.
2. From the Workload Reports screen, select a report from the Reports pane.
3. Select the Start Date and the End Date for the reporting period. Depending on the report you select, you
might need to specify a host in the Host list box.
4. Click Run Report. The report displays in the report window. For information about the meaning of the reports,
see Section 4.1.2, “Workload Balancing Report Glossary”.
9
4.1.1.2. To navigate in a Workload Balancing report
After generating a report, you can use the toolbar buttons in the report to navigate and perform certain tasks. To
display the name of a toolbar button, pause your mouse over toolbar icon.
Table 4.1. Report Toolbar Buttons
Document Map. Lets you display a document map
that helps you navigate through long reports.
Page Forward/Back. Lets you move one page ahead
or back in the report.
Back to Parent Report. Lets you return to the parent
report when working with drill-through reports. Note:
This button is only available in drill-through reports,
such as the Pool Health report.
Stop Rendering. Cancels the report generation.
Print. Lets you print a report and specify general
printing options, such as the printer, the number of
pages, and the number of copies.
Print Layout. Lets you display a preview of the report
before you print it. To exit Print Layout, click the Print
Layout button again.
Page Setup. Lets you specify printing options such as
the paper size, page orientation, and margins.
Export. Lets you export the report as an Acrobat
(.PDF) file or as an Excel file with a .XLS extension.
Find. Lets you search for a word in a report, such as
the name of a virtual machine.
4.1.1.3. To print a Workload Balancing report
Before you can print a report, you must first generate it.
1. (Optional.) Preview the printed document by clicking the following Print Layout button:
2. (Optional.) Change the paper size and source, page orientation, or margins by clicking the following Page Setup
button:
3. Click the following Print button:
4.1.1.3.1. To export a Workload Balancing report
You can export a report in either Microsoft Excel or Adobe Acrobat (PDF) formats.
1. After generating the report, click the following Export button:
2. Select one of the following items from the Export button menu:
10
• Excel
• Acrobat (PDF) file
Note:
The data that appears between report export formats may be inconsistent, depending on the
export format you select. Reports exported to Excel include all the data available for reports,
including "drilldown" data, whereas reports exported to PDF and displayed in XenCenter only
contain the data that you selected when you generated the report.
4.1.2. Workload Balancing Report Glossary
This topic provides information about the following Workload Balancing reports:
4.1.2.1. Chargeback Utilization Analysis
You can use the Chargeback Utilization Analysis report ("chargeback report") to determine how much of a
resource (such as a physical server) a specific department in your organization used. Specifically, the report shows
information about all the virtual machines in your pool, including their availability and resource utilization. Since
this report shows virtual machine availability ("up time"), it can help you demonstrate Service Level Agreements
compliance and availability.
The chargeback report can help you implement a simple chargeback solution and facilitate billing. To bill
customers for usage of a specific resource, generate the report, save it as Excel, and then customize the
spreadsheet data to include your price per unit. Alternatively, you can import the Excel data into your billing
system.
If you know that you will want to bill internal or external customers for virtual machine usage, consider
incorporating department or customer names in your virtual-machine naming conventions. This will make reading
chargeback reports easier.
The resource reporting in the chargeback report is, in some cases, based on the allocation of physical resources
to individual virtual machines.
Likewise, because XenServer lets you allocate fixed or automatic allocations of memory, the average memory
data in this report is based on the amount of memory currently allocated to the virtual machine, whether it is
through a fixed memory allocation or an automatically adjusting memory allocation (Dynamic Memory Control).
The chargeback report contains the following columns of data:
• VM Name. The name of the virtual machine to which the data in the columns in that row applies.
• VM Uptime. The number of minutes the virtual machine was powered on (or, more specifically, appears with
a green icon beside it in XenCenter).
• VCPU Allocation. The number of virtual CPUs configured on the virtual machine. Each virtual CPU receives an
equal share of the physical CPU(s) on the host. For example, if you configured eight virtual CPUs on a host that
contains two physical CPUs and this column had "1" in it, then this value is equal to 2/16 of the total processing
power on the host.
• Minimum CPU Usage (%). The lowest recorded value for virtual CPU utilization in the reporting period. This
value is expressed as a percentage of the VM's virtual CPU capacity, and the capacity is based on the number
of virtual CPUs allocated to the VM. For example, if, when you created the VM, you allocated one virtual CPU
to it, Minimum CPU Usage represents the lowest percentage of virtual CPU usage XenServer recorded, even
if it was only for a second. If you allocated two virtual CPUs to the VM, the value in this column represents the
lowest usage of the combined capacity of both virtual CPUs, expressed as a percentage.
Ultimately, the percentage of CPU usage represents the lowest recorded workload that virtual CPU handled. For
example, if you allocate one virtual CPU to a VM and the physical CPU on the host is 2.4 GHz, you are allocating
one-eighth of 2.4 GHz to the VM. This means that if the VM's allocated capacity is 0.3GHz, or 300MHz, and
11
the Minimum CPU Usage for the VM was 20%, the VM's lowest usage of the physical host's CPU during the
reporting period was 60MHz.
• Maximum CPU Usage (%). The highest percentage of the VM's virtual CPU capacity that the VM consumed
during the reporting period. The CPU capacity consumed is a percentage of the virtual CPU capacity you
allocated to the VM. For example, if, when you created the VM, you allocated one virtual CPU to it, the
Maximum CPU Usage represents the highest recorded percentage of virtual CPU usage during the time
reported. If you allocated two virtual CPUs to the VM, the value in this column represents the highest utilization
from the combined capacity of both virtual CPUs.
• Average CPU Usage (%). The average amount, expressed as a percentage, of the VM's virtual CPU capacity that
was in use during the reporting period. The CPU capacity is the virtual CPU capacity you allocated to the VM.
If you allocated two virtual CPUs to the VM, the value in this column represents the average utilization from
the combined capacity of both virtual CPUs.
• Total Storage Allocation (GB). The amount of disk space that is currently allocated to the VM at the time the
report was run. Frequently, unless you modified it, this disk space is the amount of disk space you allocated
to the VM when you created it.
• Virtual NIC Allocation. The number of virtual interfaces (VIFs) allocated to the VM.
• Current Minimum Dynamic Memory (MB).
• Fixed memory allocation. If you assigned a VM a fixed amount of memory (for example, 1024MB), the same
amount of memory appears in the following columns: Current Minimum Dynamic Memory (MB), Current
Maximum Dynamic Memory (MB), Current Assigned Memory (MB), and Average Assigned Memory (MB).
• Dynamic memory allocation. If you configured XenServer to adjust a VM's memory automatically based on
a range (that is, you enabled Dynamic Memory Control), the minimum amount of memory specified in the
range appears in this column. For example, if in the Memory Settings dialog box in XenCenter, you selected
the Automatically allocate memory within this range option for this VM and then specified the range values
as 1024MB as the minimum memory and 2048MB as the maximum memory, then 1024MB will appear in
the Current Minimum Dynamic Memory (MB) column.
• Current Maximum Dynamic Memory (MB).
• Dynamic memory allocation. If XenServer is set to adjust a VM's memory automatically based on a range,
the maximum amount of memory specified in the range appears in this column. For example, if the memory
range you provided was 1024MB minimum and 2048MB maximum, then 2048MB will appear in the Current
Maximum Dynamic Memory (MB) column.
• Fixed memory allocation. If you assign a VM a fixed amount of memory (for example, 1024MB), the same
amount of memory appears in the following columns: Current Minimum Dynamic Memory (MB), Current
Maximum Dynamic Memory (MB), Current Assigned Memory (MB), and Average Assigned Memory (MB).
• Current Assigned Memory (MB).
• Dynamic memory allocation. When Dynamic Memory Control is configured, this value indicates the amount
of memory XenServer is currently allocating to the VM at the time at which the report was run.
• Fixed memory allocation. If you assign a VM a fixed amount of memory (for example, 1024MB), the same
amount of memory appears in the following columns: Current Minimum Dynamic Memory (MB), Current
Maximum Dynamic Memory (MB), Current Assigned Memory (MB), and Average Assigned Memory (MB).
Note:
If you change the VM's memory allocation immediately before running this report, the value
reflected in this column reflects the new memory allocation you configured.
• Average Assigned Memory (MB).
• Dynamic memory allocation. When Dynamic Memory Control is configured, this value indicates the average
amount of memory XenServer allocated to the VM over the reporting period.
• Fixed memory allocation. If you assign a VM a fixed amount of memory (for example, 1024MB), the same
amount of memory appears in the following columns: Current Minimum Dynamic Memory (MB), Current
Maximum Dynamic Memory (MB), Current Assigned Memory (MB), and Average Assigned Memory (MB).
12
Note:
If you change the VM's memory allocation immediately before running this report, the value
displayed in this column may not change from what it would have previously displayed. The
value in this column reflects the average over the time period.
• Average Network Reads (BPS). The average amount of data (in bits per second) the VM received during the
reporting period.
• Average Network Writes (BPS). The average amount of data (in bits per second) the VM sent during the
reporting period.
• Average Network Usage (BPS). The combined total (in bits per second) of the Average Network Reads and
Average Network Writes. For example, if a VM sent, on average, 1,027 bits per second and received, on average,
23,831 bits per second over the reporting period, then the Average Network Usage would be the combined
total of these two values: 24,858 bits per second.
• Total Network Usage (BPS). The total of all network read and write transactions in bits per second over the
reporting period.
4.1.2.2. Host Health History
This report displays the performance of resources (CPU, memory, network reads, and network writes) on specific
host in relation to threshold values.
The colored lines (red, green, yellow) represent your threshold values. You can use this report with the Pool
Health report for a host to determine how a particular host's performance might be affecting overall pool health.
When you are editing the performance thresholds, you can use this report for insight into host performance.
You can display resource utilization as a daily or hourly average. The hourly average lets you see the busiest hours
of the day, averaged, for the time period.
To view report data grouped by hour, expand + Click to view report data grouped by house for the time period
under the Host Health History title bar.
Workload Balancing displays the average for each hour for the time period you set. The data point is based on a
utilization average for that hour for all days in the time period. For example, in a report for May 1, 2009 to May
15, 2009, the Average CPU Usage data point represents the resource utilization of all fifteen days at 12:00 hours
combined together as an average. That is, if CPU utilization was 82% at 12PM on May 1st, 88% at 12PM on May
2nd, and 75% on all other days, the average displayed for 12PM is 76.3%.
Note:
Workload Balancing smooths spikes and peaks so data does not appear artificially high.
4.1.2.3. Pool Optimization Performance History
The optimization performance report displays optimization events (that is, when you optimized a resource pool)
against that pool's average resource usage. Specifically, it displays resource usage for CPU, memory, network
reads, and network writes.
The dotted line represents the average usage across the pool over the period of days you select. A blue bar
indicates the day on which you optimized the pool.
This report can help you determine if Workload Balancing is working successfully in your environment. You can
use this report to see what led up to optimization events (that is, the resource usage before Workload Balancing
recommended optimizing).
This report displays average resource usage for the day; it does not display the peak utilization, such as when the
system is stressed. You can also use this report to see how a resource pool is performing if Workload Balancing
is not making optimization recommendations.
13
In general, resource usage should decline or be steady after an optimization event. If you do not see improved
resource usage after optimization, consider readjusting threshold values. Also, consider whether or not the
resource pool has too many virtual machines and whether or not new virtual machines were added or removed
during the time frame you specified.
4.1.2.4. Pool Audit Trail
This report displays the contents of the XenServer Audit Log, a XenServer feature designed to log attempts to
perform unauthorized actions and select authorized actions, including import/export, host and pool backups, and
guest and host console access. The report gives more meaningful information when XenServer administrators are
given their own user accounts with distinct roles assigned to them using the Role-Based Access Control feature.
Important:
To run the audit log report, the Audit Logging feature must be enabled. By default, Audit Log
is always enabled in the Workload Balancing virtual appliance.
The enhanced Pool Audit Trail feature allows you to specify the granularity of the audit log report. You can also
search and filter the audit trail logs by specific users, objects, and by time. The Pool Audit Trail Granularity is set
to Minimum by default. This option captures limited amount of data for specific users and object types. You can
modify the setting at any time based on the level of detail you would require in your report. For example, set
the granularity to Medium for a user-friendly report of the audit log. If you require a detailed report, set the
option to Maximum.
4.1.2.4.1. Report Contents
The Pool Audit Trail report contains the following:
• Time. The time XenServer recorded the user's action.
• User Name. The name of the person who created the session in which the action was performed. In some
cases, this may be the User ID
• Event Object. The object that was the subject of the action (for example, a virtual machine).
• Event Action. The action that occurred. For definitions of these actions, see Audit Log Event Names.
• Access. Whether or not the user had permission to perform the action.
• Object Name. The name of the object (for example, the name of the virtual machine).
• Object UUID. The UUID of the object (for example, the UUID of the virtual machine).
• Succeeded. This provides the status of the action (that is, whether or not it was successful).
4.1.2.4.2. Audit Log Event Names
The Audit Log report logs XenServer events, event objects and actions, including import/export, host and pool
backups, and guest and host console access. The following table defines some of the typical events that appear
frequently in the XenServer Audit Log and Pool Audit Trail report, and it also specifies the granularity of these
events.
In the Pool Audit Trail report, the events listed in the Event Action column apply to a pool, VM, or host. To
determine what the events apply to, see the Event Object and Object Name columns in the report. For additional
event definitions, see the events section of the Citrix XenServer Management API.
Pool
Audit
Granularity
Trail
Event Action
User Action
Minimum
VM.start
Started a VM
Minimum
VM.copy
Copied the specified VM, creating a new VM
Minimum
host.reboot
Restarted the XenServer host
14
Pool
Audit
Granularity
Trail
Event Action
User Action
Minimum
host-disable
Put the host into a state where no new VMs can be started
Minimum
pool.join
Instructed the host to join a new pool
Minimum
pool.join_force
Instructed (forced) the host to join a pool
Medium
SR.destroy
Destroyed the Storage Repository
Medium
SR.create
Created a new Storage Repository
Medium
VDI.snapshot
Took a read-only snapshot of the VDI, returning a
reference to the snapshot
Medium
VDI.clone
Took an exact copy of the VDI returning a reference to the
new disk
Medium
VIF.plug
Hot-plugged the specified VIF, dynamically attaching it to
the running VM
Medium
VIF.unplug
Hot-unplugged the specified VIF, dynamically detaching it
from the running VM
Maximum
auth.get_subject_identifierQueried the external directory service to obtain the
subject identifier as a string from the human-readable
subject name
Maximum
task.cancel
Requested that a task be cancelled
Maximum
VBD.insert
Inserted new media into the device
Maximum
VIF.get_by_uuid
Obtained a reference to the VIF instance with the
specified UUID
Maximum
VIF.get_shareable
Obtained the shareable field of the given VDI
Maximum
SR.get_all
Returns a list of all of the SRs known to the system
Maximum
pool.create_new_blob
Created a placeholder for a named binary piece of data
that is associated with this pool
Maximum
host.send_debug_keys
Injected the given string as debugging keys into Xen
Maximum
VM.get_boot_record
Returned a record describing the VMs dynamic state,
initialised when the VM boots and updated to reflect
runtime configuration changes, for example, CPU hotplug
4.1.2.5. Pool Health
The pool health report displays the percentage of time a resource pool and its hosts spent in four different
threshold ranges: Critical, High, Medium, and Low. You can use the Pool Health report to evaluate the
effectiveness of your performance thresholds.
A few points about interpreting this report:
• Resource utilization in the Average Medium Threshold (blue) is the optimum resource utilization regardless of
the placement strategy you selected. Likewise, the blue section on the pie chart indicates the amount of time
that host used resources optimally.
15
• Resource utilization in the Average Low Threshold Percent (green) is not necessarily positive. Whether Low
resource utilization is positive depends on your placement strategy. For example, if your placement strategy
is Maximum Density and most of the time your resource usage was green, Workload Balancing might not be
fitting the maximum number of virtual machines possible on that host or pool. If this is the case, you should
adjust your performance threshold values until the majority of your resource utilization falls into the Average
Medium (blue) threshold range.
• Resource utilization in the Average Critical Threshold Percent (red) indicates the amount of time average
resource utilization met or exceeded the Critical threshold value.
If you double-click on a pie chart for a host's resource usage, XenCenter displays the Host Health History report
for that resource (for example, CPU) on that host. Clicking the Back to Parent Report toolbar button returns you
to the Pool Health history report.
If you find the majority of your report results are not in the Average Medium Threshold range, you probably
need to adjust the Critical threshold for this pool. While Workload Balancing provides default threshold settings,
these defaults are not effective in all environments. If you do not have the thresholds adjusted to the correct
level for your environment, the Workload Balancing optimization and placement recommendations might not be
appropriate. For more information, see Section 5.1.1.2.2, “Changing the Critical Thresholds”.
4.1.2.6. Pool Health History
This report provides a line graph of resource utilization on all physical hosts in a pool over time. It lets you see the
trend of resource utilization—if it tends to be increasing in relation to your thresholds (Critical, High, Medium,
and Low). You can evaluate the effectiveness of your performance thresholds by monitoring trends of the data
points in this report.
Workload Balancing extrapolates the threshold ranges from the values you set for the Critical thresholds when
you connected the pool to Workload Balancing. Although similar to the Pool Health report, the Pool Health History
report displays the average utilization for a resource on a specific date rather than the amount of time overall
the resource spent in a threshold.
With the exception of the Average Free Memory graph, the data points should never average above the Critical
threshold line (red). For the Average Free Memory graph, the data points should never average below the Critical
threshold line (which is at the bottom of the graph). Because this graph displays free memory, the Critical
threshold is a low value, unlike the other resources.
A few points about interpreting this report:
• When the Average Usage line in the chart approaches the Average Medium Threshold (blue) line, it indicates
the pool's resource utilization is optimum regardless of the placement strategy configured.
• Resource utilization approaching the Average Low Threshold (green) is not necessarily positive. Whether Low
resource utilization is positive depends on your placement strategy. For example, if your placement strategy
is Maximum Density and most days the Average Usage line is at or below the green line, Workload Balancing
might not be placing virtual machines as densely as possible on that pool. If this is the case, you should adjust
the pool's Critical threshold values until the majority of its resource utilization falls into the Average Medium
(blue) threshold range.
• When the Average Usage line intersects with the Average Critical Threshold Percent (red), this indicates the
days when the average resource utilization met or exceeded the Critical threshold value for that resource.
If you find the data points in the majority of your graphs are not in the Average Medium Threshold range, but
you are satisfied with the performance of this pool, you might need to adjust the Critical threshold for this pool.
For more information, see Section 5.1.1.2.2, “Changing the Critical Thresholds”.
4.1.2.7. Pool Optimization History
The Pool Optimization History report provides chronological visibility into Workload Balancing optimization
activity.
16
Optimization activity is summarized graphically and in a table. Drilling into a date field within the table displays
detailed information for each pool optimization performed for that day.
This report lets you see the following information:
• VM Name. The name of the virtual machine that Workload Balancing optimized.
• Reason. The reason for the optimization.
• Status. If the optimization was successful.
• From Host. The physical server where the virtual machine was originally hosted.
• To Host. The physical server where the virtual machine was migrated.
• Time. The time when the optimization occurred.
Tip:
You can also generate a Pool Optimization History report from the WLB tab by clicking the
View History link.
4.1.2.8. Virtual Machine Motion History
This line graph displays the number of times virtual machines migrated on a resource pool over a period of time.
It indicates if a migration resulted from an optimization recommendation and to which host the virtual machine
moved. This report also indicates the reason for the optimization. You can use this report to audit the number
of migrations on a pool.
Some points about interpreting this report:
• The numbers on the left side of the chart correspond with the number of migrations possible, which is based
on how many virtual machines are in a resource pool.
• You can look at details of the migrations on a specific date by expanding the + sign in the Date section of the
report.
4.1.2.9. Virtual Machine Performance History
This report displays performance data for each virtual machine on a specific host for a time period you specify.
Workload Balancing bases the performance data on the amount of virtual resources allocated for the virtual
machine. For example, if the Average CPU Usage for your virtual machine is 67%, this means that your virtual
machine was using, on average, 67% of its virtual CPU for the period you specified.
The initial view of the report displays an average value for resource utilization over the period you specified.
Expanding the + sign displays line graphs for individual resources. You can use these graphs to see trends in
resource utilization over time.
This report displays data for CPU Usage, Free Memory, Network Reads/Writes, and Disk Reads/Writes.
17
Chapter 5. Managing Workload Balancing
Features and Settings
This chapter provides information about how to perform optional changes to Workload Balancing settings,
including how to:
• Adjust the optimization mode
• Optimize and manage power automatically
• Change the critical thresholds
• Tune metric weightings
• Exclude hosts from recommendations
• Configure advanced automation settings and data storage
• Adjust the Pool Audit Trail granularity settings
This chapter assumes that you already connected your pool to a Workload Balancing virtual appliance. For
information about downloading, importing, and configuring a Workload Balancing virtual appliance, see the Citrix
Workload Balancing 6.5 Quick Start. To connect to the virtual appliance, see Section 6.1.1, “Connecting to the
Workload Balancing Virtual Appliance”.
5.1. Changing Workload Balancing Settings
After connecting to the Workload Balancing virtual appliance, you can, if desired, edit the settings Workload
Balancing uses to calculate placement and recommendations.
Placement and optimization settings you can modify include changing the placement strategy, configuring
automatic optimizations and power management, editing performance thresholds and metric weightings, and
excluding hosts.
Workload Balancing settings apply collectively to all virtual machines and hosts in the pool.
Provided the network and disk thresholds align with the hardware in your environment, you might want to
consider using most of the defaults in Workload Balancing initially.
After Workload Balancing is enabled for a while, Citrix recommends evaluating your performance thresholds and
determining if you need to edit them. For example, consider if you are:
• Getting recommendations when they are not yet required. If this is the case, try adjusting the thresholds until
Workload Balancing begins providing suitable recommendations.
• Not getting recommendations when you think you should receive them. For example, if your network has
insufficient bandwidth and you do not receive recommendations, you might need to tweak your settings.
If this is the case, try lowering the network critical thresholds until Workload Balancing begins providing
recommendations.
Before you edit your thresholds, you might find it handy to generate a Pool Health report and the Pool Health
History report for each physical host in the pool.
You can use the Workload Balancing Configuration properties in XenCenter to modify the configuration settings.
To update the credentials XenServer and the Workload Balancing server use to communicate, see Section 8.1.7,
“Editing the Workload Balancing Configuration File”.
To display the Workload Balancing Settings dialog box
1.
In the Infrastructure pane of XenCenter, select XenCenter > your-pool.
18
2.
In the Properties pane, click the WLB tab.
3.
In the WLB tab, click Settings.
5.1.1. Adjusting the Optimization Mode
Workload Balancing makes recommendations to rebalance, or optimize, the virtual-machine workload in your
environment based on a strategy for placement you select known as the optimization mode.
Workload Balancing lets you chose from two optimization modes:
• Maximize Performance. (Default.) Workload Balancing attempts to spread workload evenly across all physical
hosts in a resource pool. The goal is to minimize CPU, memory, and network pressure for all hosts. When
Maximize Performance is your placement strategy, Workload Balancing recommends optimization when a host
reaches the High threshold.
• Maximize Density. Workload Balancing attempts to minimize the number of physical hosts that must be online
by consolidating the active virtual machines.
When you select Maximize Density as your placement strategy, you can specify parameters similar to the
ones in Maximize Performance. However, Workload Balancing uses these parameters to determine how it can
pack virtual machines onto a host. When Maximize Density is your placement strategy, Workload Balancing
recommends consolidation optimizations when a virtual machine reaches the Low threshold.
Workload Balancing also lets you apply these optimization modes all of the time, Fixed, or switch between modes
for specified time periods, Scheduled:
• Fixed optimization modes set Workload Balancing to a specific optimization behavior—either to try to create
the best performance or the highest density—at all times.
• Scheduled optimization modes let you schedule for Workload Balancing to apply different optimization modes
depending on the time of day. For example, you might want to configure Workload Balancing to optimize for
performance during the day when you have users connected and then, to save energy, specify for Workload
Balancing to optimize for Maximum Density at night.
When you configure Scheduled optimization modes, Workload Balancing automatically changes to the
optimization mode at the beginning of the time period you specified. You can configure Everyday, Weekdays,
Weekends, or individual days. For the hour, you select a time of day.
To set an optimization mode for all time periods
1.
In the Resources pane of XenCenter, select XenCenter > your-pool.
2.
In the Properties pane, click the WLB tab.
3.
On the WLB tab, click Settings.
4.
In the left pane, click Optimization Mode.
5.
In the Fixed section of the Optimization Mode page, select one of these optimization modes:
• Maximize Performance. (Default.) Attempts to spread workload evenly across all physical hosts in a
resource pool. The goal is to minimize CPU, memory, and network pressure for all hosts.
• Maximize Density. Attempts to fit as many virtual machines as possible onto a physical host. The goal is to
minimize the number of physical hosts that must be online. (Note that Workload Balancing considers the
performance of consolidated VMs and issues a recommendation to improve performance if a resource
on a host reaches a Critical threshold.)
To specify times when the optimization mode will change automatically
1.
In the Infrastructure pane of XenCenter, select XenCenter > your-pool.
2.
In the Properties pane, click the WLB tab.
19
3.
On the WLB tab, click Settings.
4.
In the left pane, click Optimization Mode
5.
In the Optimization Mode pane, select Scheduled. The Scheduled section becomes available.
6.
Click Add New.
7.
In the Change to box, select one of the following modes:
• Maximize Performance. Attempts to spread workload evenly across all physical hosts in a resource pool.
The goal is to minimize CPU, memory, and network pressure for all hosts.
• Maximize Density. Attempts to fit as many virtual machines as possible onto a physical host. The goal is
to minimize the number of physical hosts that must be online.
8.
Select the day of the week and the time when you want Workload Balancing to begin operating in this mode.
9.
Create more scheduled mode changes (that is, "tasks") until you have the number you need. If you only
schedule one task, Workload Balancing will switch to that mode as scheduled, but then it will never switch
back.
10. Click OK.
To delete or pause a scheduled optimization mode task
1.
Display the Optimization Mode dialog box by following steps 1 to 4 in the previous procedure.
2.
Select the task you want to delete or disable from the Scheduled Mode Changes list.
3.
Do one of the following:
• Delete the task permanently. Click the Delete button.
• Stop the task from running temporarily. Right-click the task and click Disable.
Tip:
You can also disable or enable tasks by selecting the task, clicking Edit, and selecting the
Enable Task check box in the Optimization Mode Scheduler dialog.
Note:
To re-enable a task, right click the task in the Scheduled Mode Changes list and click Enable.
.
To edit a scheduled optimization mode task
1.
Do one of the following:
• Double-click the task you want to edit.
• Select the task you want to edit, and click Edit.
2.
In the Change to box, select a different mode or make other changes as desired.
Note:
Clicking Cancel, before clicking OK, will undo any changes you made in the Optimization tab,
including deleting a task.
5.1.1.1. Optimizing and Managing Power Automatically
You can configure Workload Balancing to apply recommendations automatically (Automation) and turn hosts on
or off automatically (Power Management). In order to power down hosts automatically (for example, during lowusage periods), you must configure Workload Balancing to apply recommendations automatically and enable
power management. Both power management and automation are described in the sections that follow.
20
5.1.1.1.1. Applying Recommendations Automatically
Workload Balancing lets you configure for it to apply recommendations on your behalf and perform the
optimization actions it recommends automatically. You can use this feature, which is known as Automation,
to apply any recommendations automatically, including ones to improve performance or power down hosts.
However, to power down hosts as virtual-machines usage drops, you must configure automation, power
management, and Maximum Density mode.
By default, Workload Balancing does not apply recommendations automatically. You must enable Automation
if you want Workload Balancing to apply recommendations automatically. If you do not, you must apply
recommendations manually by clicking Apply Recommendations.
Workload Balancing will not automatically apply recommendations to hosts or virtual machines if the
recommendations conflict with High Availability settings. If a pool will become overcommitted by applying
Workload Balancing optimization recommendations, XenCenter will prompt you whether or not you want to
continue applying the recommendation. When Automation is enabled, Workload Balancing will not apply any
power-management recommendations that exceed the number of host failures to tolerate in the High Availability
plan.
When Workload Balancing is running with the Automation feature enabled, this is sometimes referred to as
running in automated mode.
It is possible to tune how Workload Balancing applies recommendations in automated mode. For information,
see Section 5.1.1.4.1, “Setting Conservative or Aggressive Automated Recommendations”
5.1.1.1.2. Enabling Workload Balancing Power Management
The term power management refers to the ability to the turn the power on or off for physical hosts. In a Workload
Balancing context, this term refers to powering hosts in a pool on or off based on the pool's total workload.
Configuring Workload Balancing power management on a host requires that:
• The hardware for the host has remote power on/off capabilities
• The Host Power On feature is configured for the host
• The host has been explicitly selected as a host to participate in (Workload Balancing) Power Management
In addition, if you want Workload Balancing to power off hosts automatically, you also need to configure Workload
Balancing to:
• Apply recommendations automatically
• Apply Power Management recommendations automatically
When enabled and the pool is in Maximum Density mode, if Workload Balancing detects unused resources, it
recommends powering off hosts until it eliminates all excess capacity in the pool. If Workload Balancing detects
there is not sufficient host capacity in the pool to turn off hosts, it recommends leaving the hosts on until the
pool's workload decreases enough to power off hosts. When you configure Workload Balancing to power off
extra hosts automatically, it applies these recommendations automatically and, consequently, behaves in the
same way.
When a host is set to participate in Power Management, Workload Balancing makes power-on/off
recommendations as needed. If you configure Workload Balancing so that it can automatically power on hosts
when you are running in Maximum Performance mode, Workload Balancing will power on one or more hosts if
a resource's utilization on a host exceeds the High threshold. However, when running in Maximum Performance
mode, Workload Balancing never powers off hosts after it has powered them on.
If you turn on the option to apply Power Management recommendations automatically, you do so at the pool
level. However, you can specify which hosts from the pool you want to participate in Power Management.
Understanding Power Management Behavior
21
Before Workload Balancing recommends powering hosts on or off, it selects the hosts to transfer virtual machines
to (that is, to "fill"). It does so in the following order:
1. Filling the pool master since it is the host that cannot be powered off.
2. Filling the host with the most virtual machines.
3. Filling subsequent hosts according to which hosts have the most virtual machines running.
When Workload Balancing fills the pool master, it does so assuming artificially low (internal) thresholds for
the master. Workload Balancing uses these low thresholds as a buffer to prevent the pool master from being
overloaded.
Workload Balancing fills hosts in this order to encourage density.
This illustration shows how, when consolidating VMs on hosts in Maximum Density mode, XenServer attempts to
fill the pool master first, the most loaded host second, and the least loaded host third.
If Workload Balancing detects a performance issue while the pool is in Maximum Density mode, it attempts to
address the issue by recommending migrating workloads among the powered-on hosts. If Workload Balancing
cannot resolve the issue using this method, it attempts to power-on a host. (Workload Balancing determines
which host(s) to power-on by applying the same criteria it would if the optimization mode was set to Maximum
Performance.)
When Workload Balancing is running in Maximum Performance mode, Workload Balancing recommends
powering on hosts until the resource utilization on all hosts in the pool falls below the High threshold.
If, while migrating one or more virtual machines, Workload Balancing determines that increasing capacity would
benefit the pool's overall performance, it powers on hosts automatically or recommends doing so.
Important:
Workload Balancing never recommends powering on a host unless Workload Balancing
powered it off.
5.1.1.1.3. Designing Environments for Power Management and VM Consolidation
When you are planning a XenServer implementation and you intend to configure automatic VM consolidation
and power management, consider your workload design. For example, you may want to:
22
• Place Different Types of Workloads in Separate Pools. If you have an environment with distinct types of
workloads (for example, user applications versus domain controllers) or types of applications that perform
better with certain types of hardware, consider if you need to locate the virtual machines hosting these
workloads in different pools.
Because power management and VM consolidation are managed at the pool level, you should design pools so
they contain workloads that you want consolidated at the same rate, factoring in considerations such as those
discussed in Section 5.1.1.4, “Controlling Automated Recommendations”.
• Exclude Hosts from Workload Balancing. Some hosts may need to be on at all times. See Section 5.1.1.3,
“Excluding Hosts from Recommendations” for more information.
5.1.1.1.4. To apply optimization recommendations automatically
1. In the Infrastructure pane of XenCenter, select XenCenter > your-pool.
2. In the Properties pane, click the WLB tab.
3. In the WLB tab, click Settings.
4. In the left pane, click Automation.
5. Select one or more of the following check boxes:
• Automatically apply Optimization recommendations. When you select this option, you do not need to
accept optimization recommendations manually. Workload Balancing automatically accepts optimization
and placement recommendations it makes.
• Automatically apply Power Management recommendations. The behavior of this option varies according
to the pool's optimization mode:
• Maximum Performance Mode. When Automatically apply Power Management recommendations is
enabled, Workload Balancing automatically powers on hosts when doing so improves host performance.
• Maximum Density Mode. When Automatically apply Power Management recommendations is
enabled, Workload Balancing automatically powers off hosts when resource utilization drops below the
Low threshold (that is, Workload Balancing powers hosts off automatically during low usage periods).
6. (Optional.) Fine-tune optimization recommendations by clicking Advanced in the left pane of the Settings
dialog and doing one or more of the following:
• Specifying the number of times Workload Balancing must make an optimization recommendation before
the recommendation is applied automatically. The default is three times, which means the recommendation
is applied on the third time it is made.
• Selecting the lowest level of optimization recommendation that you want Workload Balancing to apply
automatically. The default is High.
• Changing the aggressiveness with which Workload Balancing applies its optimization recommendations.
You may also want to specify the number of minutes Workload Balancing has to wait before applying an
optimization recommendation to a recently moved VM.
All of these settings are explained in more depth in Section 5.1.1.4.1, “Setting Conservative or Aggressive
Automated Recommendations”.
7. Do one of the following:
• If you want to configure power management, click Automation/Power Management and proceed to the
Section 5.1.1.1.5, “To select hosts for power management”.
• If you do not want to configure power management and you are finished configuring automation, click OK.
5.1.1.1.5. To select hosts for power management
1. In the Power Management section, select the hosts that you want Workload Balancing to recommend
powering on and off.
Note:
23
Selecting hosts for power management recommendations without selecting the
Automatically apply Power Management recommendations check box results in Workload
Balancing suggesting power-management recommendations but not applying them
automatically for you.
2. Click OK. If none of the hosts in the resource pool support remote power management, Workload Balancing
displays the message, "No hosts support Power Management."
5.1.1.2. Understanding When Workload Balancing Makes Recommendations
Workload Balancing continuously evaluates the resource metrics of physical hosts and VMs across the pool(s)
it is managing against thresholds. Thresholds are preset values that function like boundaries that a host must
exceed before Workload Balancing can make an optimization recommendation. At a very high level. the Workload
Balancing process is as follows:
1. Workload Balancing detects that the threshold for a resource was violated.
2. Workload Balancing evaluates if it will make an optimization recommendation.
3. Workload Balancing determines which host(s) it will recommend function as the destination host(s), which is
the host where Workload Balancing will recommend relocating one or more VMs.
4. Workload Balancing makes the recommendation.
After Workload Balancing determines that a host could benefit from optimization, before it makes the
recommendation, it evaluates other hosts on the pool to decide (a) the order in which it should make the
optimization (what hosts, what VMs) and (b) where it should recommend placing a virtual machine when it does
make a recommendation. To accomplish these two tasks, Workload Balancing uses thresholds and weightings
as follows:
• Thresholds, which are the boundary values Workload Balancing compares your pool's resource metrics against,
are used to determine whether or not to make a recommendation and what hosts are a suitable candidate for
hosting relocated virtual machines.
• Weightings, which are a way of ranking resources according to how much you want them to be considered,
are used to determine the processing order. That is, after Workload Balancing determines it needs to make
a recommendation, it uses your specifications of what resources are important to determine which hosts'
performance to address first and which virtual machines to recommend migrating first.
For each resource Workload Balancing monitors, it has four levels of thresholds (Critical, High, Medium, and
Low), which are discussed in the sections that follow. Workload Balancing evaluates whether to make a
recommendation when a resource metric on a host:
• Exceeds the High threshold when the pool is running in Maximum Performance mode (improve performance)
• Drops below the Low threshold when the pool is running in Maximum Density mode (consolidate VMs on hosts)
• Exceeds the Critical threshold when the pool is running in Maximum Density mode (improve performance)
For example, if the High threshold for a pool running in Maximum Performance mode is set to 80%, when the CPU
utilization on a host reaches 80.1%, Workload Balancing will evaluate whether or not to issue a recommendation.
When a resource violates its threshold, Workload Balancing evaluates the resource's metric against its historical
performance to prevent issuing an optimization recommendation based on a temporary resource spike. To do
so, Workload Balancing creates an historically averaged utilization metric by evaluating the data for resource
utilization captured at the following times:
Data Captured
Weight
Immediately, at the time threshold was exceeded (that is, real-time data)
70%
Thirty minutes before the threshold was exceeded
25%
24
Data Captured
Weight
Twenty-four hours before the threshold was exceeded
5%
For example, if the CPU utilization on the host exceeds the High threshold at 12:02 PM on Thursday, Workload
Balancing checks the data utilization at 11:32 AM on Thursday, and at 12:02PM on Wednesday (the previous
day). If the CPU utilization is 80.1% at 12:02 PM, but only 50% at 11:32AM, and 78% at 12:32 PM on Wednesday,
Workload Balancing does not make a recommendation. This is because the historically averaged utilization is
72.47%, so Workload Balancing assumes the utilization is a temporary spike. However, if the CPU utilization was
78% at 11:32AM, Workload Balancing would make a recommendation since the historically averaged utilization
is 80.1%.
5.1.1.2.1. Optimization and Consolidation Process
The Workload Balancing process for determining potential optimizations varies according to the optimization
mode (Maximum Performance or Maximum Density). However, regardless of the optimization mode,
optimization and placement recommendations are made using a two-stage process:
1. Determine potential optimizations. (That is, what VMs to migrate off hosts.)
2. Determine placement recommendations. (That is, what hosts would be suitable candidates for new hosts.)
Note:
Workload Balancing only recommends migrating VMs that meet the XenServer criteria for live
migration, including the destination host must have the storage the VM requires. Likewise,
the destination host must have sufficient resources to accommodate adding the VM without
exceeding the thresholds of the optimization mode configured on the pool (for example,
the High threshold in Maximum Performance mode and the Critical threshold for Maximum
Density mode).
When Workload Balancing is running in automated mode, you can tune the way it applies recommendations. See
Section 5.1.1.4.1, “Setting Conservative or Aggressive Automated Recommendations” for more information.
Optimization Recommendation Process in Maximum Performance Mode
When running in Maximum Performance mode, Workload Balancing uses the following process to determine
potential optimizations:
1. Every two minutes Workload Balancing evaluates the resource utilization for each host in the pool. It does
so by monitoring on each host and determining if each resource's utilization exceeds its High threshold. See
Section 5.1.1.2.2, “Changing the Critical Thresholds” for more information about the High threshold.
If, in Maximum Performance mode, a resource's utilization exceeds its High threshold, then Workload
Balancing starts the process to determine if it wants to make an optimization recommendation. Workload
Balancing determines if it should make an optimization recommendation based on whether or not doing so
can ease performance constraints, such as ones revealed by the High threshold.
For example, if Workload Balancing sees that the performance of the virtual machines on Host A is being
negatively affected by insufficient CPU resources, if Workload Balancing can find another host with less CPU
utilization, it will recommend moving one or more virtual machines to another host.
2. If a resource's utilization on a host exceeds the relevant threshold, Workload Balancing combines the
resource's current utilization with the historical data from thirty minutes ago and twenty-four hours ago
to form the historically averaged utilization. If the historically averaged utilization exceeds the resource's
threshold, Workload Balancing determines it will make an optimization recommendation.
3. Workload Balancing uses metric weightings to determine what hosts it should optimize first. The resource to
which you have assigned the most weight is the one that Workload Balancing attempts to address first. See
Section 5.1.1.2.3, “Tuning Metric Weightings” for information about metric weightings.
4. Workload Balancing determines which hosts can support the VM(s) it wants to migrate off of host(s).
25
Workload Balancing makes this determination by calculating the projected effect on resource utilization of
placing different combinations of VMs on hosts. (Workload Balancing uses a method of performing these
calculations that in mathematics is known as permutation.)
To do so, Workload Balancing creates a single metric or score to forecast the impact of migrating a VM to the
host. The score indicates the suitability of a host as a home for additional VMs.
To score the host's performance, Workload Balancing combines the host's current metrics, the host's metrics
from the last thirty minutes, the host's metrics from 24 hours ago, and the VM's metrics.
5. After scoring the hosts and VMs, Workload Balancing attempts to build (in essence) virtual models of what
the hosts would look like with different combinations of VMs so that it can determine the best host to place
the VM.
In Maximum Performance mode, Workload Balancing uses metric weightings to determine what hosts it
should optimize first and what VMs on those hosts it should migrate first. Workload Balancing bases its
models on the metric weightings. For example, if CPU utilization is assigned the highest importance, Workload
Balancing sorts hosts and VMs to optimize first according to what hosts are most affected by CPU utilization
(that is, are running closest to the High threshold for CPU utilization) and what VMs have the highest CPU
utilization (or are running the closest to its High threshold).
6. Workload Balancing continues calculating optimizations. It perceives hosts as potential candidates for
optimization and views VMs as candidates for migration until the predicted resource utilization on the VM's
host drops below the configured High threshold. The term predicted resource utilization refers to the resource
utilization Workload Balancing forecasts a host will have after Workload Balancing has added or removed a
VM from the host.
Consolidation Process in Maximum Density Mode
Workload Balancing determines if it should make a recommendation (for consolidation) based on whether it can
migrate a virtual machine off a host and still run the destination host below the Critical threshold.
1. When a resource's utilization drops below its Low threshold, Workload Balancing begins calculating potential
consolidation scenarios.
2. When Workload Balancing discovers a potential way in which it can consolidate VMs on a host, it evaluates
whether or not the destination host is a suitable home for the VM.
3. Like in Maximum Performance mode, Workload Balancing scores the host to determine the suitability of a
host as a home for new VMs.
For example, if Workload Balancing determines that it can consolidate virtual machines on fewer hosts,
before it makes an optimization recommendation, Workload Balancing checks to make sure that the resource
utilization on the destination hosts will not exceed the Critical thresholds after the VMs are relocated to them.
Note:
Workload Balancing does not consider metric weightings when it makes a consolidation
recommendation; it only considers metric weightings to ensure performance on hosts.
4. After scoring the hosts and VMs, Workload Balancing attempts to build (in essence) virtual models of what
the hosts would look like with different combinations of VMs so that it can determine the best host to place
the VM.
5. Workload Balancing calculates the effect of adding more VMs to a host until it forecasts that adding another
VM to the host would cause one of its resources to exceed the Critical threshold.
6. Workload Balancing recommendations always suggest filling the pool master first since it is the host that
cannot be powered off. However, Workload Balancing applies a buffer to the pool master so that it cannot
be over-allocated.
7. Workload Balancing continues to recommend migrating VMs on to hosts until no hosts remain that will not
exceed a Critical threshold if a VM is migrated on to them.
26
5.1.1.2.2. Changing the Critical Thresholds
You might want to change Critical thresholds as a way of controlling when optimization recommendations are
triggered. This section provides guidance about how to modify the default Critical thresholds on hosts in the pool
and how values set for Critical threshold alter High, Medium, and Low thresholds.
Workload Balancing determines whether or not to produce recommendations based on if the averaged
historical utilization for a resource on a host violates its threshold. As previously discussed in Section 5.1.1.2.1,
“Optimization and Consolidation Process”, Workload Balancing recommendations are triggered when the High
threshold in Maximum Performance mode or Low and Critical thresholds for Maximum Density mode are
violated. After you specify a new Critical threshold for a resource, Workload Balancing resets the resource's other
thresholds relative to the new Critical threshold. (To simplify the user interface, the Critical threshold is the only
threshold you can change through XenCenter.)
The following table shows the default values for the Workload Balancing thresholds:
Metric
Critical
High
Medium
Low
CPU Utilization
90%
76.5%
45%
22.5%
Free Memory
51MB
63.75MB
510MB
1020MB
Network Reads
25MB/sec
21.25MB/sec
12.5MB/sec
6.25MB/sec
Network Writes
25MB/sec
21.25MB/sec
12.5MB/sec
6.25MB/sec
To calculate the values for all metrics except memory (that is, the High, Medium, and Low CPU, network, and disk
metrics), Workload Balancing multiplies the new value for the Critical threshold with the following factors:
• High Threshold Factor: 0.85
• Medium Threshold Factor: 0.50
• Low Threshold Factor: 0.25
This means that if you increase, for example, the pool's Critical threshold for CPU Utilization to 95%, Workload
Balancing automatically resets the High, Medium, and Low thresholds to 80.75%, 47.5%, and 23.75% respectively.
To calculate the threshold values for free memory, Workload Balancing multiplies the Critical threshold with these
factors:
• High Threshold Factor: 1.25
• Medium Threshold Factor: 10.0
• Low Threshold Factor: 20.0
To perform this calculation for a specific threshold, multiply the factor for the threshold with the value you entered
for the critical threshold for that resource:
High, Medium, or Low Threshold = Critical Threshold * Threshold Factor
For example, if you change the Critical threshold for Network Reads to 40MB/sec and you want to know its Low
threshold, you multiply 40 by 0.25, which equals 10MB/sec. To obtain the Medium threshold, you multiple 40
by 0.50, and so on.
While the Critical threshold triggers many optimization recommendations, other thresholds can also trigger
optimization recommendations, as follows:
• High threshold.
27
• Maximum Performance. Exceeding the High threshold triggers optimization recommendations to relocate
a VM to a host with lower resource utilization.
• Maximum Density. Workload Balancing will not recommend placing a virtual machine on host if moving that
VM to the host will cause the utilization of any of the host's resources to exceed the High threshold value.
• Low threshold.
• Maximum Performance. Workload Balancing does not trigger recommendations from the Low threshold.
• Maximum Density. When a metric value drops below the Low threshold, it signals Workload Balancing that
hosts are being underutilized and triggers an optimization recommendation to consolidate virtual machines
on fewer hosts. Workload Balancing continues to recommend moving VMs onto a host until the metric
values for one of the host's resource reaches its High threshold.
However, if after a VM is relocated, a resource's utilization on the VM's new host exceeds its Critical
threshold, Workload Balancing will temporarily use an algorithm similar to the Maximum Performance loadbalancing algorithm to find a new host for the virtual machines. Workload Balancing continues to use this
algorithm to recommend moving virtual machines until resource utilization on hosts across the pool falls
below the High threshold.
5.1.1.2.2.1. To change the critical thresholds
1. In the Infrastructure pane of XenCenter, select XenCenter > your-resource-pool.
2. In the Properties pane, click the WLB tab.
3. In the WLB tab, click Settings.
4. In the left pane, select Critical Thresholds. These critical thresholds are used to evaluate host resource
utilization.
5. In Critical Thresholds page, type one or more new values in the Critical Thresholds boxes. The values represent
resource utilization on the host.
Workload Balancing uses these thresholds when making virtual-machine placement and pool-optimization
recommendations. Workload Balancing strives to keep resource utilization on a host below the critical values
set.
5.1.1.2.3. Tuning Metric Weightings
The way in which Workload Balancing uses metric weightings when determining which hosts and VMs to process
first varies according to the optimization mode, Maximum Density or Maximum Performance.
When Workload Balancing is processing optimization recommendations, it creates an optimization order. To
determine this, Workload Balancing ranks the hosts to address first according to which hosts have the highest
metric values for whatever resource is ranked as the most important in the metric weightings page.
In general, metric weightings are mainly used when a pool is in Maximum Performance mode. However, when
Workload Balancing is in Maximum Density mode, it does use metric weightings if a resource exceeds its Critical
threshold.
Maximum Performance mode
In Maximum Performance mode, Workload Balancing uses metric weightings to determine (a) which hosts'
performance to address first and (b) which VMs to recommend migrating first.
For example, if you rank Network Writes as the most important resource for Workload Balancing to consider—
that is, you move the Metric Weighting slider all the way to the right (More Important) and you move all the
other sliders to the middle—then Workload Balancing will begin addressing performance issues (that is, making
optimization recommendations) on the host with the highest number of Network Writes per second first.
If you configure all resources to be equally important (that is, the Metric Weighting slider is in the same place
for all resources), Workload Balancing addresses CPU utilization first and memory second, as these are typically
the most constrained resources.
28
Maximum Density mode
In Maximum Density mode, Workload Balancing only uses metric weightings when a host reaches the
Critical threshold—then Workload Balancing applies a Maximum Performance-like algorithm until no Hosts are
exceeding the Critical thresholds. When using the Maximum Performance-like algorithm, Workload Balancing
uses metric weightings to determine the optimization order in the same way as it does for Maximum Performance
mode.
If two or more hosts have resource(s) exceeding their Critical thresholds, Workload Balancing verifies the
importance you set for each resource before determining which host to optimize first and which VMs on that
host to relocate first.
For example, your pool contains Host A and Host B, which are in the following state:
• The CPU utilization on Host A exceeds the Critical threshold for CPU, and the metric weighting for CPU utilization
is set to the far right of the slider (More Important).
• The memory utilization on Host B exceeds the Critical threshold for memory, and the metric weighting for
memory utilization is set to the far left of the slider (Less Important).
Workload Balancing will recommend optimizing Host A first because the resource on it that reached the Critical
threshold is the resource assigned the highest weight. After Workload Balancing determines that it needs to
address the performance on Host A, Workload Balancing then begins recommending placements for VMs on that
host beginning with the VM that has the highest CPU utilization, since that CPU utilization is the resource with
the highest weight.
After Workload Balancing has recommended optimizing Host A, it makes optimization recommendations for Host
B. When it recommends placements for the VMs on Host B, it does so by addressing CPU utilization first, since
CPU utilization was assigned the highest weight.
If there are additional hosts that need optimization, Workload Balancing addresses the performance on those
hosts according to what host has the third highest CPU utilization.
By default, all metric weightings are set to the farthest point on the slider (More Important).
Note:
The weighting of metrics is relative. This means that if all of the metrics are set to the same
level, even if that level is Less Important, they will all be weighted the same. The relation of
the metrics to each other is more important than the actual weight at which you set each
metric.
5.1.1.2.3.1. To edit metric weighting factors
1. Stopping
2. In the Infrastructure pane of XenCenter, select XenCenter > your-resource-pool.
3. Click the WLB tab, and then click Settings.
4. In the left pane, select Metric Weighting.
5. In Metric Weighting page, as desired, adjust the sliders beside the individual resources.
Moving the slider towards Less Important indicates that ensuring virtual machines always have the highest
amount of this resource available is not as vital on this resource pool.
5.1.1.3. Excluding Hosts from Recommendations
When configuring Workload Balancing, you can specify that specific physical hosts are excluded from Workload
Balancing optimization and placement recommendations, including Start On placement recommendations.
Situations when you may want to exclude hosts from recommendations include when:
29
• You want to run the pool in Maximum Density mode and consolidate and shut down hosts, but there are
specific hosts you want to exclude from this behavior.
• When two virtual-machine workloads always need to run on the same host (for example, if they have
complementary applications or workloads).
• You have workloads that you do not want moved (for example, a domain controller or database server).
• You want to perform maintenance on a host and you do not want VMs placed on the host.
• The performance of the workload is so critical that the cost of dedicated hardware is irrelevant.
• Specific hosts are running high-priority workloads (virtual machines), and you do not want to use the High
Availability feature to prioritize these virtual machines.
• The hardware in the host is not the optimum for the other workloads in the pool.
Regardless of whether you specify a fixed or scheduled optimization mode, hosts excluded remain excluded
even when the optimization mode changes. Therefore, if you only want to prevent Workload Balancing from
shutting off a host automatically, consider not enabling (or deselecting) Power Management for that host instead
as described in Section 5.1.1.1, “Optimizing and Managing Power Automatically ”.
When you exclude a host from recommendations, you are specifying for Workload Balancing not to manage that
host at all. This means that Workload Balancing will not make any optimization recommendations for an excluded
host. In contrast, when you do not select a host to participate in Power Management, Workload Balancing still
manages the host, but it will not issue power-management recommendations for that host.
5.1.1.3.1. To exclude hosts from Workload Balancing
Use this procedure to exclude a host in a pool that Workload Balancing is managing from power management,
host evacuation, placement, and optimization recommendations.
1. In the Resources pane of XenCenter, select XenCenter > your-resource-pool.
2. In the Properties pane, click the WLB tab.
3. In the WLB tab, click Settings.
4. In the left pane, select Excluded Hosts.
5. In Excluded Hosts page, select the hosts for which you do not want WLB to recommend alternate placements
and optimizations.
5.1.1.4. Controlling Automated Recommendations
Workload Balancing supplies some advanced settings that let you to control how Workload Balancing applies
automated recommendations. These settings appear on the Advanced page of the Workload Balancing
Configuration dialog.
To display the Advanced page of the Workload Balancing Configuration dialog
1.
In the Resources pane of XenCenter, select XenCenter > your-resource-pool.
2.
In the Properties pane, click the WLB tab.
3.
In the WLB tab, click Settings.
4.
In the left pane, select Advanced.
5.1.1.4.1. Setting Conservative or Aggressive Automated Recommendations
When running in automated mode, the frequency of optimization and consolidation recommendations and how
soon they are automatically applied is a product of multiple factors, including:
• How long you specify Workload Balancing waits after moving a VM before making another recommendation
• The number of recommendations Workload Balancing must make before applying a recommendation
automatically
30
• The severity level a recommendation must achieve before the optimization is applied automatically
• The level of consistency in recommendations (recommended VMs to move, destination hosts) Workload
Balancing requires before applying recommendations automatically
Important:
In general, Citrix recommends that you do not adjust the settings for factors without either
guidance from Citrix Technical Support or significant observation and testing of your pool's
behavior with Workload Balancing enabled. Incorrectly configuring these settings can result
in Workload Balancing not making any recommendations.
5.1.1.4.1.1. VM Migration Interval
You can specify the number of minutes Workload Balancing must wait after the last time a particular virtual
machine was moved, regardless of the cause, before Workload Balancing can generate another recommendation
that includes that particular recently moved virtual machine.
The recommendation interval is designed to prevent Workload Balancing from generating recommendations for
artificial reasons (for example, if there was a temporary utilization spike).
When Automation is configured, it is especially important to be careful when modifying the recommendation
interval. If an issue occurs that leads to continuous, recurring spikes, increasing the frequency (that is, setting a
lower number) can generate many recommendations and, consequently, relocations.
Note:
Setting a recommendation interval does not affect how long Workload Balancing waits to
factor recently rebalanced hosts into recommendations for Start-On Placement, Resume, and
Maintenance Mode.
5.1.1.4.1.2. Recommendation Count
Every two minutes, Workload Balancing checks to see if it can generate recommendations for the pool it is
monitoring. When you enable Automation, you can specify the number of times a consistent recommendation
must be made before Workload Balancing can automatically apply the recommendation. To do so, you configure a
setting known as the Recommendation Count. The Recommendation Count and the Optimization Aggressiveness
setting let you fine-tune the automated application of recommendations in your environment.
As described in the aggressiveness section, Workload Balancing uses the similarity of recommendations to a) vet
if the recommendation is truly needed and b) determine if the destination host has stable enough performance
over a prolonged period of time to accept a relocated VM (without needing to move it off the host again shortly).
Workload Balancing uses the Recommendation Count value to determine a recommendation must be repeated
before Workload Balancing automatically applies the recommendation.
Workload Balancing uses this setting as follows:
1. Every time Workload Balancing generates a recommendation that meets its consistency requirements, as
indicated by the Optimization Aggressiveness setting, Workload Balancing increments the Recommendation
Count. If the recommendation does not meet the consistency requirements, Workload Balancing may reset
the Recommendation Count to zero, depending on the factors described in Section 5.1.1.4.1.4, “Optimization
Aggressiveness”.
2. When Workload Balancing generates enough consistent recommendations to meet the value for the
Recommendation Count, as specified in the Recommendations text box, it automatically applies the
recommendation.
If you choose to modify this setting, the value you should set varies according to your environment. Consider
these scenarios:
• If host loads and activity increase extremely quickly in your environment, you may want to increase value
for the Recommendation Count. Workload Balancing generates recommendations every two minutes. For
31
example, if you set this interval to "3", then six minutes later Workload Balancing applies the recommendation
automatically.
• If host loads and activity increase gradually in your environment, you may want to decrease the value for the
Recommendation Count.
Accepting recommendations uses system resources and affects performance when Workload Balancing is
relocating the virtual machines. Increasing the Recommendation Count increases the number of matching
recommendations that must occur before Workload Balancing applies the recommendation, which encourages
Workload Balancing to apply more conservative, stable recommendations and may decrease the potential for
spurious VM moves. It should be noted, however, that the Recommendation Count is set to a conservative value
by default.
Because of the potential impact adjusting this setting may have on your environment, Citrix only recommends
changing it with extreme caution, preferably by testing and iteratively changing the value or under the guidance
of Citrix Technical Support.
5.1.1.4.1.3. Recommendation Severity
All optimization recommendations include a severity rating (Critical, High, Medium, Low) that indicates the
importance of the recommendation. Workload Balancing bases this rating on a combination of factors including
configuration options you set, such as thresholds and metric tunings, resources available for the workload, and
resource-usage history. (The severity rating for a recommendation appears in the Optimization Recommendations
pane on the WLB tab.)
When you configure Workload Balancing to apply recommendations automatically, you can set the minimum
severity level that should be associated with a recommendation before Workload Balancing automatically applies
it.
5.1.1.4.1.4. Optimization Aggressiveness
To provide additional assurance when running in automated mode, Workload Balancing has consistency criteria
for accepting optimizations automatically so as to prevent moving VMs due to spikes and anomalies. In
automated mode, Workload Balancing does not accept the first recommendation it produces. Instead, Workload
Balancing waits to automatically apply a recommendation until a host or VM exhibits consistent behavior over
time. The phrase consistent behavior over time refers to factors such as whether a host continues to trigger
recommendations and whether the same VMs on that host continue to trigger recommendations.
Workload Balancing determines if behavior is consistent by using criteria for consistency and by having criteria for
the number of times the same recommendation is made (that is, the Recommendation Count). You can configure
how strictly you want Workload Balancing to apply the consistency criteria using the Optimization Aggressiveness
setting.
While Citrix primarily designed the Optimization Aggressiveness setting for demonstration purposes, you can use
this setting to control the amount of stability you want in your environment before Workload Balancing applies
an optimization recommendation. The most stable setting (Low aggressiveness) is configured by default. In this
context, the term stable refers to the similarity of the recommended changes over time, as explained throughout
this section. In most environments, aggressiveness is not desirable. Consequently, Low is the default setting.
Workload Balancing uses up to four criteria to ascertain consistency. The number of criteria that must be met
varies according to the level you set in the Optimization Aggressiveness setting. The lower the level (for example,
Low or Medium) the less aggressively Workload Balancing is in accepting a recommendation. In other words,
Workload Balancing is stricter about requiring criteria to match (or less cavalier or aggressive) about consistency
when aggressiveness is set to Low.
For example, if the aggressiveness level is set to Low, Workload Balancing requires that each criterion for Low is
met the number of times specified in the Recommendations box (where you specify the Recommendation Count
value) before automatically applying the recommendation.
For example, if you set the Recommendation Count in the Recommendations box to "3", you are making
Workload Balancing wait until it sees all the criteria listed below for Low met and repeated in three consecutive
recommendations. This helps ensure that the VM actually needs to be moved and that the destination host
32
Workload Balancing is recommending has consistently stable resource utilization over a longer period of time. It
reduces the potential for a recently moved VM to be moved off a host due to host performance changes after
the move. By default, this setting is set to a conservative setting (Low) to encourage stability.
Citrix does not recommend increasing the Optimization Aggressiveness setting to increase the frequency with
which your hosts are being optimized. If you feel that your hosts are not being optimized quickly or frequently
enough, try adjusting the Critical thresholds and comparing them against the Pool Health report.
The consistency criteria associated with the different levels of aggressiveness is the following:
Low
• All VMs in subsequent recommendations must be the same (as demonstrated by matching UUIDs in each
recommendation).
• All destination hosts must be the same in subsequent recommendations
• The recommendation that immediately follows the initial recommendation must match or else the
Recommendation Count reverts to 1
Medium
• All VMs in subsequent recommendations must be from the same host same; however, they can be different
VMs from the ones in the first recommendation.
• All destination hosts must be the same in subsequent recommendations
• One of the next two recommendation that immediately follows the first recommendation must match or else
the Recommendation Count reverts to 1
High
• All VMs in the recommendations must be from the same host same; however, the recommendations do not
have to immediately follow each other.
• The host from which Workload Balancing recommended the VM move must be the same in each
recommendation
• The Recommendation Count does not revert to 1 if the two recommendations that follow the first
recommendation do not match
Example
The following example illustrates the way in which Workload Balancing uses the Optimization Aggressiveness
setting and the Recommendation Count to determine whether or not to automatically accept a recommendation.
The first column represents the recommendation number. The second column, "Placement Recommendations,"
represents the placement recommendations made when Workload Balancing issued the optimization
recommendation: each recommendation proposes three VM placements (moves). The third, fourth, and
fifth columns represent the effect of the Optimization Aggressiveness setting on a group of placement
recommendations (the group is denoted by the row, for example "Recommendation #1"). The number in
the aggressiveness columns is the number of times there have been consecutive recommendation at that
Optimization Aggressiveness setting. For example, 1 in the medium column for Recommendation #2 indicates
that the recommendation was not consistent enough at that Optimization Aggressiveness setting and the counter
was reset to 1.
For example, in the table below, when the Optimization Aggressiveness setting is set to High, the
Recommendation Count continues to increase after Recommendation #1, #2, and #3—even though the same
VMs are not recommended for new placements in each recommendation. Workload Balancing applies the
placement recommendation with Recommendation #3 because it has seen the same behavior from that host for
three consecutive recommendations.
In contrast, when set to Low aggressiveness, the consecutive recommendations count does not increase for the
first four recommendations. In fact, the Recommendation Count resets to 1 with each recommendation because
33
the same VMs were not recommended for placements. The Recommendation Count does not start to increase
until the same recommendation is made in Recommendation #5. Finally, Workload Balancing automatically
applies the recommendation made in Recommendation #6 after the third time it issues the same placement
recommendations.
Recommendation #1
Placement Recommendations
High
Medium
Low
Aggressiveness,Aggressiveness,
Aggressiveness,
Recommendation
Recommendation
Recommendation
Count
Count
Count
• Move VM1 from Host A to Host B
1
1
1
2
1
1
3 (Apply)
1
1
2
1
3 (Apply)
2
• Move VM3 from Host A to Host B
• Move VM5 from Host A to Host C
Recommendation #2
• Move VM1 from Host A to Host B
• Move VM3 from Host A to Host C
• Move VM7 from Host A to Host C
Recommendation #3
• Move VM1 from Host A to Host B
• Move VM3 from Host A to Host C
• Move VM5 from Host A to Host C
Recommendation #4
• Move VM1 from Host A to Host B
• Move VM3 from Host A to Host B
• Move VM5 from Host A to Host C
Recommendation #5
• Move VM1 from Host A to Host B
• Move VM3 from Host A to Host B
• Move VM5 from Host A to Host C
Recommendation #6
• Move VM1 from Host A to Host B
3 (Apply)
• Move VM3 from Host A to Host B
• Move VM5 from Host A to Host C
To configure VM Recommendation intervals
1. In the Resources pane of XenCenter, select XenCenter>your-pool.
2. In the Properties pane, click the WLB tab.
3. In the WLB tab, click Settings.
4. In the left pane, click Advanced.
5. In the VM Recommendation Interval section, do one or more of the following:
• In the Minutes box, type a value for the number of minutes you want Workload Balancing to wait before
making another optimization recommendation on a newly rebalanced host.
• In the Recommendations box, type a value for the number of recommendations you want Workload
Balancing to make before it applies a recommendation automatically.
• Select a minimum severity level before optimizations are applied automatically.
• Modify how aggressively Workload Balancing applies optimization recommendations when it is running
in automated mode. Increasing the aggressiveness level reduces constraints on the consistency of
34
recommendations before automatically applying them. The Optimization Aggressiveness setting directly
complements the Recommendation Count setting (that is, the Recommendations box).
Note:
If you type "1" for the value in the Recommendations setting, the Optimization
Aggressiveness setting is not relevant.
5.2. To Modify the Pool Audit Trail Granularity Setting
Follow this procedure to modify the granularity settings:
1. Select the pool in the Infrastructure view, click the WLB tab, and then click Settings.
2. In the left pane, click Advanced.
3. On the Advance page, click the Pool Audit Trail Report Granularity drop-down list, and select an option from
the list.
Important:
Select the granularity based on the your audit log requirements. For example, if you set your
audit log report granularity to Minimum, the audit report only captures limited amount of
data for specific users and object types. If you set the granularity to Medium, the report
provides a user-friendly report of the audit log. If you choose to set the granularity to
Maximum, the report will contain detailed information about the audit log report. Setting
the audit log report to Maximum can cause the Workload Balancing server to use more disk
space and memory.
4. To confirm your changes, click OK.
35
Chapter 6. Administering Workload
Balancing
This chapter provides information about the following topics:
• How to reconfigure a pool to use a different Workload Balancing virtual appliance
• How to disconnect a pool from Workload Balancing or temporarily stop Workload Balancing
• Database grooming
• How to change configuration options
Note:
Workload Balancing is available for XenServer Enterprise Edition customers or those who have
access to XenServer through their XenDesktop entitlement. To learn more about XenServer
editions, and to find out how to upgrade, visit the Citrix website here. For detailed information
on Licensing - refer to CTX141511 - XenServer 6.5 Licensing FAQ.
6.1. Administering and Maintaining Workload Balancing
After Workload Balancing has been up and running for a while, there are routine tasks that you may need to
perform in order to keep Workload Balancing running optimally. These tasks may arise either as the result of
changes to your environment (such as different IP addresses or credentials), hardware upgrades, or routine
maintenance.
Some administrative tasks you may want to perform on Workload Balancing include:
• Connecting or reconnecting a pool to a Workload Balancing virtual appliance
• Reconfiguring a pool to use another Workload Balancing virtual appliance
• Renaming the Workload Balancing user account
• Disconnecting the Workload Balancing virtual appliance from a pool
• Removing the Workload Balancing virtual appliance
• Understanding the Role Based Access Control permissions Workload Balancing requires
Workload Balancing lets you fine-tune some aspects of its behavior through a configuration file known as the
wlb.conf file.
This section also discusses some database administration tasks for those interested in additional ways to manage
the Workload Balancing database.
6.1.1. Connecting to the Workload Balancing Virtual Appliance
After Workload Balancing configuration, you must connect the pool you want managed to the Workload Balancing
virtual appliance using either the XE commands or the Connect to WLB Server dialog box in XenCenter. Likewise,
you might need to reconnect to the same virtual appliance at some point.
To complete the XenCenter procedure that follows, you need the:
• Host name (or IP address) and port of the Workload Balancing virtual appliance.
• Credentials for the resource pool you want Workload Balancing to monitor.
• Credentials for the account you created on the Workload Balancing virtual appliance. This is often known as the
Workload Balancing user account. XenServer uses this account to communicates with Workload Balancing. You
created this account on the Workload Balancing virtual appliance during Workload Balancing Configuration.
36
This illustration shows how (1) XenServer communicates with Workload Balancing using
an account you created during Workload Balancing Configuration and (2) the Workload
Balancing virtual appliance authenticates to XenServer using the credentials for the pool.
If you want to specify the Workload Balancing virtual appliance's host name for use when connecting to the
Workload Balancing virtual appliance (that is, in the Connect to WLB Server dialog), you must first add its host
name and IP address to your DNS server.
If you want to configure certificates from a certificate authority, Citrix recommends specifying an FQDN or an IP
address that does not expire.
When you first connect to Workload Balancing, it uses the default thresholds and settings for balancing
workloads. Automatic features, such as Automated Optimization Mode, Power Management, and Automation,
are disabled by default.
Note:
Workload Balancing is available for XenServer Enterprise Edition customers or those who have
access to XenServer through their XenDesktop entitlement. To learn more about XenServer
editions, and to find out how to upgrade, visit the Citrix website here. For detailed information
on Licensing - refer to CTX141511 - XenServer 6.5 Licensing FAQ.
6.1.1.1. To connect your pool to the Workload Balancing virtual appliance
1. In the Resources pane of XenCenter, select XenCenter > your-resource-pool.
2. In the Properties pane, click the WLB tab.
The WLB tab displays the Connect button.
3. In the WLB tab, click Connect.
The Connect to WLB Server dialog box appears.
37
4. In the Server Address section, enter the following:
a. In the Address box, type the IP address or FQDN of the Workload Balancing virtual appliance (for example,
your WLB-appliance-computername.yourdomain.net).
Tip:
To obtain the IP address for the WLB virtual appliance, see Section 6.1.1.2, “To obtain the IP
address for the WLB virtual appliance”.
b. Enter the port number in the Port box. XenServer uses this port to communicate with Workload Balancing.
By default, XenServer connects to Workload Balancing (specifically the Web Service Host service) on port
8012. If you changed the port number during Workload Balancing Configuration, you must enter that port
number here.
Note:
Use the default port number unless you changed it during Workload Balancing Configuration.
The port number specified during Workload Balancing Configuration, in any firewall rules,
and in the Connect to WLB Server dialog must match.
5. In the WLB Server Credentials section, enter the user name (for example, wlbuser) and password the
XenServer pool will use to connect to the Workload Balancing virtual appliance.
This must be the account you created during Workload Balancing Configuration. By default, the user name
for this account is wlbuser.
6. In the XenServer Credentials section, enter the user name and password for the pool you are configuring.
Workload Balancing will use these credentials to connect to the XenServer hosts in that pool.
38
To use the credentials with which you are currently logged into XenServer, select the Use the current
XenCenter credentials check box. If you have assigned a role to this account using the Access Control feature
(RBAC), be sure the role has sufficient permissions to configure Workload Balancing. For more information,
see Section 6.1.1.3, “Workload Balancing Access Control Permissions”.
7. After connecting the pool to the Workload Balancing virtual appliance, Workload Balancing automatically
begins monitoring the pool with the default optimization settings. If you want to modify these settings or
change the priority given to specific resources, wait at least sixty seconds (until the XenCenter Log shows
discovery is finished) before proceeding and see Section 5.1, “Changing Workload Balancing Settings”.
6.1.1.2. To obtain the IP address for the WLB virtual appliance
1. Select the WLB virtual appliance in the Resource pane in XenCenter, and click on the Console tab.
2. Log in to the appliance. Enter the VM user name (typically "root") and the root password you created when
you imported the appliance.
3. Enter the following command at the prompt:
ifconfig
6.1.1.3. Workload Balancing Access Control Permissions
When Role Based Access Control (RBAC) is implemented in your environment, all user roles can display the WLB
tab. However, not all roles can perform all operations. The following table lists the minimum role administrators
require to use Workload Balancing features:
Task
Minimum Required Role
Configure, Initialize, Enable, Disable WLB
Pool Operator
Apply WLB Optimization Recommendations (in WLB
tab)
Pool Operator
Accept WLB Placement Recommendations ("star"
recommendations)
VM Power Admin
Generate WLB Reports, including the Pool Audit Trail
report
Read Only
Display WLB Configuration
Read Only
If a user tries to use Workload Balancing and he or she does not have sufficient permissions, a role elevation
dialog appears. For more information about RBAC, see the XenServer Administrator's Guide.
6.1.2. Determining the Status of the Workload Balancing Virtual Appliance
Run the service workloadbalancing status command, as described in the Chapter 8, Appendix B: Workload
Balancing Commands.
6.1.3. Reconfiguring a Pool to Use Another WLB Appliance
You can reconfigure a resource pool to use a different Workload Balancing virtual appliance.
39
However, to prevent the old Workload Balancing virtual appliance from remaining inadvertently configured and
running against a pool, you must disconnect the pool from the old Workload Balancing virtual appliance before
connecting the pool to the new (or different) Workload Balancing virtual appliance.
After disconnecting a pool from the old Workload Balancing virtual appliance, you can connect the pool by
specifying the name of the new Workload Balancing virtual appliance. Perform the steps in the procedure that
follows on the pool you want to connect to a different Workload Balancing virtual appliance.
To use a different Workload Balancing virtual appliance
1. From the Pool menu, select Disconnect Workload Balancing Server and click Disconnect when prompted.
2. In the WLB tab, click Connect. The Connect to WLB Server dialog appears.
3. In the Address box, type the IP address or FQDN of the new Workload Balancing server.
4. In the WLB Server Credentials section, enter the user name (for example, wlbuser) and password the
XenServer pool (master) will use to connect to the Workload Balancing virtual appliance.
This must be the account you created during Workload Balancing Configuration for the new virtual appliance.
By default, the user name for this account is wlbuser.
5. In the XenServer Credentials section, enter the user name and password for the pool you are configuring
(typically the password for the pool master). Workload Balancing will use these credentials to connect to the
hosts in the pool.
To use the credentials with which you are currently logged into XenServer, select the Use the current
XenCenter credentials check box. If you have assigned a role to this account using the Access Control feature
(RBAC), be sure the role has sufficient permissions to configure Workload Balancing. For more information,
see Section 6.1.1.3, “Workload Balancing Access Control Permissions”.
6.1.4. Updating Workload Balancing Credentials
Situations when you may want to update the Workload Balancing credentials include if you need to change the
name of the user account XenServer uses to communicate with Workload Balancing.
After initial configuration, if you need to update the credentials XenServer and the Workload Balancing appliance
use to communicate, you can do so using a three-step process:
1. Pause Workload Balancing by clicking Pause in the WLB tab.
2. Change the WLB credentials by running the wlbconfig command. For more information, see Chapter 8,
Appendix B: Workload Balancing Commands.
3. Re-enable Workload Balancing and specify the new credentials, as described in To re-enable Workload
Balancing and specify the new credentials.
To re-enable Workload Balancing and specify the new credentials
1.
After the progress bar completes, click Connect.
The Connect to WLB Server dialog box appears.
2.
Click Update Credentials.
3.
In the Server Address section, modify the following as desired:
40
• In the Address box, type the IP address or FQDN of the Workload Balancing appliance.
• (Optional.) If you changed the port number during Workload Balancing Configuration, enter that port
number. The port number you specify in this box and during Workload Balancing Configuration is the port
number XenServer uses to connect to Workload Balancing.
By default, XenServer connects to Workload Balancing on port 8012.
Note:
Do not edit this port number unless you have changed it when you ran the Workload
Balancing Configuration wizard. The port number value specified when you ran the Workload
Balancing Configuration wizard and the Connect to WLB Server dialog must match.
4.
In the WLB Server Credentials section, enter the user name (for example, wlbuser) and password the
computers running XenServer will use to connect to the Workload Balancing server.
5.
In the XenServer Credentials section, enter the user name and password for the pool you are configuring
(typically the password for the pool master). Workload Balancing will use these credentials to connect to
the computers running XenServer in that pool.
6.
In the XenServer Credentials section, enter the user name and password for the pool you are configuring.
Workload Balancing will use these credentials to connect to the computers running XenServer in that pool.
To use the credentials with which you are currently logged into XenServer, select the Use the current
XenCenter credentials check box.
6.1.5. Changing the Workload Balancing IP Address
Situations when you may want to update the
To change the Workload Balancing IP address, do the following:
1. Stop the Workload Balancing services by running the service workloadbalancing stop command on the virtual
appliance.
2. Change the Workload Balancing IP address by running the ifconfig command on the virtual appliance.
3. Re-enable Workload Balancing and specify the new IP address, as described in To re-enable Workload
Balancing and specify the new credentials.
4. Start the Workload Balancing services by running the service workloadbalancing start command on the virtual
appliance.
6.1.6. Stopping Workload Balancing
Because Workload Balancing is configured at the pool level, when you want it to stop managing a pool, you must
do one of the following:
• Pause Workload Balancing. Pausing Workload Balancing stops XenCenter from displaying recommendations
for the specified resource pool and managing the pool. Pausing is designed for a short period of time and lets
you resume monitoring without having to reconfigure. When you pause Workload Balancing, data collection
stops for that resource pool until you enable Workload Balancing again.
• Disconnect the pool from Workload Balancing. Disconnecting from the Workload Balancing virtual appliance
breaks the connection between the pool and if possible, deletes the pool data from the Workload Balancing
database. When you disconnect from Workload Balancing, Workload Balancing stops collecting data on the
pool.
To pause Workload Balancing temporarily
1.
In the Resource pane of XenCenter, select the resource pool for which you want to disable Workload
Balancing.
41
2.
In the WLB tab, click Pause. A message appears on the WLB tab indicating that Workload Balancing is
currently paused.
Tip:
To resume monitoring, click the Resume button in the WLB tab.
To disconnect a pool from Workload Balancing
1.
In the Infrastructure pane of XenCenter, select the resource pool on which you want to stop Workload
Balancing.
2.
From the Infrastructure menu, select Disconnect Workload Balancing Server. The Disconnect Workload
Balancing dialog box appears.
3.
Click Disconnect to permanently stop Workload Balancing from monitoring the pool.
Tip:
If you disconnected the pool from the Workload Balancing virtual appliance, to re-enable
Workload Balancing on that pool, you must reconnect to a Workload Balancing appliance. For
information, see the Section 6.1.1, “Connecting to the Workload Balancing Virtual Appliance”.
6.1.7. Entering Maintenance Mode with Workload Balancing Enabled
When Workload Balancing is enabled, if you take a physical host off-line for maintenance (that is, suspend a host
by entering Maintenance Mode), XenServer automatically migrates the virtual machines running on that host to
their optimal hosts when available. XenServer migrates them based on Workload Balancing recommendations
(performance data, your placement strategy, and performance thresholds).
If an optimal host is not available, the words Click here to suspend the VM appear in the Enter Maintenance
Mode dialog box. In this case, because there is not a host with sufficient resources to run the virtual machine,
Workload Balancing does not recommend a placement. You can either suspend this virtual machine or exit
Maintenance Mode and suspend a virtual machine on another host in the same pool. Then, if you reenter the
Enter Maintenance Mode dialog box, Workload Balancing might be able to list a host that is a suitable candidate
for migration.
Note:
When you take a host off-line for maintenance and Workload Balancing is enabled, the words
"Workload Balancing" appear in the upper-right corner of the Enter Maintenance Mode
dialog box.
6.1.7.1. To enter maintenance mode with Workload Balancing enabled
1. In the Resources pane of XenCenter, select the physical host that you want to take off-line. From the Server
menu, select Enter Maintenance Mode.
2. In the Enter Maintenance Mode dialog box, click Enter maintenance mode. The virtual machines running on
the host are automatically migrated to the optimal host based on the Workload Balancing performance data,
your placement strategy, and performance thresholds.
To take the host out of maintenance mode, right-click the host and select Exit Maintenance Mode. When you
remove a host from maintenance mode, XenServer automatically restores that host's original virtual machines
to that host.
6.1.8. Increasing the Workload Balancing Disk Size
This procedure explains how to resize the virtual disk of the Workload Balancing virtual appliance. You must shut
down the virtual appliance before performing these steps. Consequently, Workload Balancing will be unavailable
for approximately five minutes.
42
Warning:
Citrix recommends taking a snapshot of your data before performing this procedure.
Incorrectly performing these steps can result in corrupting the Workload Balancing virtual
appliance.
To extend the size of the Workload Balancing virtual-appliance disk
1.
Shut down the Workload Balancing virtual appliance.
2.
In the XenCenter resource pane, select the Workload Balancing virtual appliance (typically “Citrix WLB Virtual
Appliance”).
3.
Click the Storage tab.
4.
Select the “vdi_xvda” disk, and click the Properties button.
5.
In the “vdi_xvda” Properties dialog, select Size and Location.
6.
Increase the disk size as needed, and click OK.
7.
Start the Workload Balancing virtual appliance and log in to it.
8.
Run the following command on the Workload Balancing virtual appliance:
resize2fs /dev/xvda
9.
Run the df –h command to confirm the new disk size.
6.1.9. Removing the Workload Balancing Virtual Appliance
Citrix recommends removing the Workload Balancing virtual appliance by using the standard procedure to delete
virtual machines from XenCenter.
When you delete the Workload Balancing virtual appliance, the PostgreSQL database containing the Workload
Balancing is deleted. To save this data, you must migrate it from the database before deleting the Workload
Balancing virtual appliance.
6.1.10. Managing the Workload Balancing Database
The Workload Balancing database is a PostgreSQL database. PostgreSQL is an open-source enterprise-class
relational database. You can find documentation for PostgreSQL by searching the web.
The following information is intended for database administrators and advanced users of PostgreSQL who are
comfortable with database administration tasks. If you are not experienced with PostgreSQL, Citrix recommends
that you become familiar with it before you attempt the database tasks in the sections that follow.
By default, the PostgreSQL user name is postgres. You set the password for this account during Workload
Balancing Configuration.
The amount of historical data you can store is based on the size of the virtual disk allocated to Workload Balancing
(by default 8GB) and the minimum required space, which is discussed in Section 6.1.10.2.1, “Database Grooming
Parameters”.
If you want to store a lot of historical data, for example if you want to enable the Pool Audit trail Report, you
can do either of the following:
• Make the virtual disk size assigned to the Workload Balancing virtual appliance larger. To do so, import the
virtual appliance, and increase the size of the virtual disk by following the steps in Section 6.1.8, “Increasing
the Workload Balancing Disk Size”.
• Create periodic duplicate back-up copies of the data by enabling remote client access to the database and
using a third-party database administration tool, such as pgAdmin.
In terms of managing the database, you can control the space that database data consumes by configuring
database grooming.
43
6.1.10.1. Accessing the Database
The Workload Balancing 6.5 virtual appliance has firewall configured in it. Before you can access database, you
must add the postgresql server port to the iptables.
To open the database port in the firewall
1.
From the Workload Balancing virtual appliance console, run the following command:
iptables -A INPUT -i eth0 -p tcp -m tcp --dport 5432 -m state --state NEW,ESTABLISHED -j ACCEP
2.
(Optional.) To make this configuration persist after the virtual appliance is rebooted, run the following
command:
iptables-save > /etc/sysconfig/potables
6.1.10.2. Controlling Database Grooming
The Workload Balancing database is set up to automatically delete the oldest data whenever the VPX reaches
the minimum amount of disk space Workload Balancing requires to run. By default, the minimum amount of
required disk space is set to 1024MB.
The Workload Balancing database grooming options are controlled through the wlb.conf file.
When there is not enough disk space left on the Workload Balancing virtual appliance, Workload Balancing
automatically starts grooming historical data. The process is as follows:
1. At a predefined grooming interval, the Workload Balancing data collector checks if grooming is required.
Grooming is required if the database data has grown to the point where the only space that remains
unused is the minimum required disk space (which is set by GroomingRequiredMinimumDiskSizeInMB) . (You
can change the grooming interval if desired using GroomingIntervalInHour. However, by default Workload
Balancing checks to see if grooming is required once per hour.)
2. If grooming is required, Workload Balancing begins by grooming the data from the oldest day. Workload
Balancing then checks to see if there is now enough disk space for it to meet the minimum disk-space
requirement.
3. If the first grooming did not free enough disk space, then Workload Balancing repeats grooming up to
GroomingRetryCounter times without waiting for GroomingIntervalInHour hour .
4. If the first or repeated grooming freed enough disk space, then Workload Balancing waits for
GroomingIntervalInHour hour and returns to Step 1.
5. If the grooming initiated by the GroomingRetryCounter did not free enough disk space, then Workload
Balancing waits for GroomingIntervalInHour hour and returns to Step 1.
6.1.10.2.1. Database Grooming Parameters
There are five parameters in the wlb.conf file that control various aspects of database grooming. They are as
follows:
• GroomingIntervalInHour. Controls how many hours elapse before the next grooming check is done. For
example, if you enter "1", Workload Balancing checks the disk space hourly. If you enter "2", Workload
Balancing checks disk space every two hours to determine if grooming needs to occur.
• GroomingRetryCounter. Controls the number of times Workload balancing tries rerunning the grooming
database query.
• GroomingDBDataTrimDays. Controls the number of days worth of data Workload Balancing will delete from
the database each time it tries to groom data. The default value is one day.
• GroomingDBTimeoutInMinute. Controls the number of minutes that the database grooming will take before
it times out and is canceled. If the grooming query takes longer than is expected and does not finish running
within the timeout period, the grooming task is canceled. The default value is 0 minutes, which means that
database grooming never times out.
44
• GroomingRequiredMinimumDiskSizeInMB. Controls the minimum amount of free space left in the virtual disk
assigned to the Workload Balancing virtual appliance. When the data in the virtual disk grows to the point
where there is only minimum disk size left on the virtual disk, Workload Balancing begins to trigger database
grooming. The default value is 2048MB.
To edit these values, see the Section 8.1.7, “Editing the Workload Balancing Configuration File”.
6.1.10.3. Changing the Database Password
While it is possible to change the database password by using the wlb.conf file, Citrix recommends running the
wlbconfig command instead, as described in Section 8.1.6, “Modifying the Workload Balancing Configuration
Options”.
6.1.10.4. Archiving Database Data
To avoid having older historical data deleted, you can, optionally, copy data from the database for archiving. To
do so, you must perform the following tasks:
1. Enable client authentication on the database.
2. Set up archiving using the PostgreSQL database administration tool of your choice.
6.1.10.4.1. Enabling client authentication to the database
While you can connect directly to the database through the Workload Balancing console, you can also use a
PostgreSQL database management tool, such as the free utility pgAdmin, which is available on the web. After
downloading a database management tool, install it on the computer from which you want to connect to the
database (for example, the same laptop where you run XenCenter).
Before you can enable remote client authentication to the database, you must:
1. Modify the database configuration files, including pg_hba.conf file and the postgresql.conf, to allow
connections.
2. Stop the Workload Balancing services, restart the database, and then restart the Workload Balancing services.
3. In the database-management tool, configure the IP address of the database (that is, the IP address of the
Workload Balancing VPX) and the database password.
6.1.10.4.2. Modifying the Database Configuration Files
To enable client authentication on the database, you must modify two files on the Workload Balancing virtual
appliance: the pg_hba.conf file and the postgresql.conf file.
To edit the pg_hba.conf file
1.
Modify the pg_hba.conf file. From the Workload Balancing virtual appliance console, open the pg_hba.conf
file with an editor, such as VI. For example:
vi /var/lib/pgsql/9.0/data/pg_hba.conf
2.
If your network uses IPv4, add the IP address from the connecting computer to this file. For example:
In the configuration section, enter the following under #IPv4 local connections:
TYPE
DATABASE
USER
CIDR-ADDRESS
METHOD
host
all
all
0.0.0.0/0
trust
Enter your IP address in the CIDR-ADDRESS field.
Note:
45
Instead of entering 0.0.0.0/0, you can enter your IP address and replace the last three digits
with 0/24. The trailing "24" after the / defines the subnet mask and only allows connections
from IP addresses within that subnet mask.
When you enter trust for the Method field, it enables the connection to authenticate without requiring a
password. If you enter password for the Method field, you must supply a password when connecting to
the database.
3.
If your network uses IPv6, add the IP address from the connecting computer to this file. For example:
Enter the following under #IPv6 local connections:
TYPE
DATABASE
USER
CIDR-ADDRESS
METHOD
host
all
all
::0/0
trust
Enter the IPv6 addresses in the CIDR-ADDRESS field. In this example, the ::0/0 opens the database up to
connections from any IPv6 addresses.
4.
Save the file and quit the editor.
5.
After changing any database configurations, you must restart the database to apply the changes. Run the
following command:
service postgresql-9.0 restart
To edit the postgresql.conf file
1.
Modify the postgresql.conf file. From the Workload Balancing virtual appliance console, open the
postgresql.conf file with an editor, such as VI. For example:
vi /var/lib/pgsql/9.0/data/postgresql.conf
2.
Edit the file so that it listens on any port and not just the local host. For example:
a.
Find the following line:
# listen_addresses='localhost'
b.
Remove the comment symbol (#) and edit the line to read as follows:
listen_addresses='*'
3.
Save the file and quit the editor.
4.
After changing any database configurations, you must restart the database to apply the changes. Run the
following command:
service postgresql-9.0 restart
6.1.10.5. Changing the Database Maintenance Window
Workload Balancing automatically performs routine database maintenance daily at 12:05AM GMT (00:05), by
default. During this maintenance window, data collection occurs but the recording of data may be delayed.
However, the Workload Balancing user interface controls are available during this period and Workload Balancing
still makes optimization recommendations.
Database maintenance includes releasing allocated unused disk space and re-indexing the database.
Maintenance lasts for approximately 6 to 8 minutes. In larger pools, maintenance may last longer, depending on
how long Workload Balancing takes to perform discovery.
Depending on your time zone, you may want to change the time when maintenance occurs. For example, in the
Japan Standard Time (JST) time zone, Workload Balancing maintenance will occur at 9:05 AM (09:05), which may
46
conflict with peak usage in some organizations. If you want to specify a seasonal time change, such as Daylight
Saving Time or summer time, you must build the change into value you enter.
To change the maintenance time
1.
In the Workload Balancing console, run the following command from any directory:
crontab -e
Workload Balancing displays the following:
05 0 * * * /opt/citrix/wlb/wlbmaintenance.sh
05 0 represents the default time for Workload Balancing to perform maintenance in minutes (05) and then
hours (0). (The asterisks represent the day, month, and year the job will run: do not edit these fields.) 05 0
indicates database maintenance occurs at 12:05 AM, or 00:05, Greenwich Mean Time (GMT) every night.
This means, for example, that if you live in New York city in the United States, by default, the maintenance
would run at 7:05 PM (19:05) during winter months and 8:05 PM in summer months.
Important:
Do not edit the day, month, and year the job will run (as represented by asterisks). Database
maintenance must run daily.
2.
3.
Enter the time at which you want maintenance to occur in GMT. For example, assuming that you want the
maintenance to run at midnight:
If your time zone is...
UTC Offset
Value for Maintenance
to Run at 12:05 AM
Local Time
Value
in
Saving Time
Pacific Time Zone (PST)
in the United States (for
example, California)
UTC-08
05 8
05 7
Japan Standard Time
(JST)
UTC+09
05 15
N/A
Chinese Standard Time
UTC +08
04 15
N/A
Daylight
Save the file and quit the editor.
6.1.11. Customizing Workload Balancing
Workload Balancing provides several methods of customization:
• Command lines for scripting. See the commands in Chapter 8, Appendix B: Workload Balancing Commands.
• Host Power On scripting support. You can also customize Workload Balancing (indirectly) through the Host
Power On scripting. How to create a custom script for Host Power On is documented in the XenServer
Administrator's Guide.
47
6.2. Upgrading Workload Balancing
Note:
Workload Balancing is available for XenServer Enterprise Edition customers or those who have
access to XenServer through their XenDesktop entitlement. To learn more about XenServer
editions, and to find out how to upgrade, visit the Citrix website here. For detailed information
on Licensing - refer to CTX141511 - XenServer 6.5 Licensing FAQ.
There are two ways in which you can upgrade to Workload Balancing 6.5. You can either (a) install a new appliance
or (b) use Workload Balancing functionality to upgrade the existing virtual appliance.
This section provides information about upgrading to Workload Balancing 6.5 from any version of Workload
Balancing since Workload Balancing 6.1. You cannot use this procedure to upgrade Workload Balancing when it
is not in a virtual appliance format (for example, Workload Balancing 2.1). Using the upgrade feature lets you
retain your Workload Balancing settings and historical data.
In the virtual appliance, Workload Balancing runs on the CentOS operating system. When running the Linux
upgrade command yum update, you have the choice of upgrading (a) both Workload Balancing and CentOS, (b)
Workload Balancing only, or (c) CentOS only. These choices are specified using parameters. For the XenServer 6.5
release, Citrix recommends, if possible, upgrading both Workload Balancing and CentOS.
It is easier to upgrade Workload Balancing when you have Internet access. However, it is still possible to upgrade
without Internet access, as explained in the section that follows.
Perform the following procedures on the existing Workload Balancing virtual appliance you want to upgrade. You
do not need to download the latest virtual appliance from Citrix.com. This procedure applies the same updates
contained in the Workload Balancing 6.5 to your existing virtual appliance.
Important:
Citrix strongly recommends that you take a snapshot of the Workload Balancing virtual
appliance before you upgrade. For information about creating snapshots, see the XenCenter
Help.
Virtual Disk Space Requirements
Upgrading the virtual appliance requires at least 582k for the GNU wget package plus the size of the download
package, as follows:
• Upgrading both Workload Balancing and CentOS requires at least 145 MB
• Upgrading only Workload Balancing requires at least 55MB
• Upgrading only CentOS requires at least 100MB
When to Upgrade
Citrix suggests upgrading during a time when Workload Balancing would not typically make a lot of
recommendations (low activity). Workload Balancing will be off-line until after the upgrade is complete and
Workload Balancing has finished discovery.
The amount of time it takes for the appliance to begin making optimization recommendations after you upgrade
it depends on the size of your pool and how long it takes for Workload Balancing to finish discovery. After
discovery is complete, it takes about five minutes before Workload Balancing begins making recommendations.
Discovery time varies from approximately 20 minutes to discover 200 virtual machines to 40 minutes for 800
virtual machines. This time varies according to the size of the virtual machines, hardware, and other factors. You
can tell discovery is finished when the star ratings become available for the Start On server feature.
Upgrades and Certificates
48
If you configured certificates in Workload Balancing 6.1 or Workload Balancing 6.0.1 and you are upgrading to
XenServer 6.5, you do not need to perform any additional steps to retain your certificate configuration. The
upgrade scripts will upgrade your certificate configuration for you.
However, if you upgrade your virtual appliance and then b) want to configure certificates for the first time, you
must use the certificate-configuration instructions in the Workload Balancing 6.1 Administrator's Guide.
6.2.1. Upgrading Workload Balancing with Internet Access
Upgrading with the Internet requires that you download GNU wget, an HTTP retrieval utility. You also need to
download a Python script that configures a repository (add-repo.py) on your virtual appliance.
To upgrade Workload Balancing in environments with Internet access
1.
If you have not done so already, log in to the Workload Balancing appliance you want to upgrade as described
in Section 8.1.1, “Logging in to the Workload Balancing Virtual Appliance”
2.
Install GNU wget so you can retrieve the upgrade repository installation script using HTTP:
a.
From the bash prompt, run the following command:
yum install wget
b.
3.
During installation, wget Setup asks you to accept various prompts, such as the size of the download
package and the CentOS key. Type y when prompted.
When the wget installation is complete, download "add-repo" script by running the following command:
wget http://updates.xensource.com/XenServer/WLB/6.5/add-repo.py
When this command finishes running, a message appears stating the 'add-repo.py' script is saved.
4.
At the bash prompt, run the following command to create the upgrade repository on the Workload Balancing
appliance:
python add-repo.py
After the script finishes, the output states "Done."
5.
Do one of the following to upgrade your Workload Balancing virtual appliance:
• To upgrade both CentOS and Workload Balancing, run:
yum update
• To upgrade Workload Balancing only, run:
yum update citrix-wlb
• To upgrade CentOS only, run:
yum update --disablerepo=citrix-wlb
6.
After upgrading either CentOS or Workload Balancing or both, restart the Workload Balancing virtual
appliance.
Note:
After upgrading CentOS, the operating-system time changes from Coordinated Universal Time
(UTC) to the local time zone. One side effect of this change is that the timestamps in the
Workload Balancing log file (LogFile.log) also change from UTC time to local time. If you want
to change the system time back to UTC time, run the following command in the Workload
Balancing virtual appliance: rm -rf /etc/localtime.
6.2.2. Upgrading Workload Balancing without Direct Internet Access
49
Before you start this procedure, you must use a computer with Internet access to download the Workload
Balancing upgrade ISO files. Depending on your environment, you may need to copy the files to a USB flash drive
or a shared directory. During the procedure that follows you are instructed to copy the files to the Workload
Balancing virtual appliance.
Note:
To obtain the MD5 for the ISO files, see the checksums.txt in the ISO folder (http://
updates.xensource.com/XenServer/WLB/6.5/iso/checksums.txt).
To upgrade Workload Balancing in environments without Internet access
1.
Download the Workload Balancing ISO files (wlb-6.5-upgrade.iso and CentOS-5.7-upgrade.iso) from the
updates website (http://updates.xensource.com/XenServer/WLB/6.5/iso).
2.
Using a copying utility, such as WinSCP, copy both ISO files to the /tmp/ directory on the Workload Balancing
virtual appliance. The ISO files require approximately 200MB of free space on the Workload Balancing virtual
appliance.
3.
Create a directory to mount the ISO files:
mkdir /media/cdrom
4.
Upgrade Workload Balancing:
a.
Mount the wlb-6.5-upgrade.iso to the /media/cdrom directory:
mount -o loop /tmp/wlb-6.5-upgrade.iso /media/cdrom
b.
Create the upgrade repository:
python /media/cdrom/add-repo.py
c.
Run the yum upgrade utility to upgrade Workload Balancing:
yum update --disablerepo=* --enablerepo=citrix-wlb-media
d.
Unmount the Workload Balancing ISO file:
umount /media/cdrom
e.
Clean up the yum cache:
rm -r /var/cache/yum/
5.
Upgrade the CentOS operating system:
a.
Mount the CentOS-5.7-upgrade.iso to the /media/cdrom directory:
mount -o loop /tmp/CentOS-5.7-upgrade.iso /media/cdrom
b.
Install the CentOS update:
yum update --disablerepo=* --enablerepo=citrix-centos-media
c.
Unmount the CentOS ISO file:
umount /media/cdrom
6.
After upgrading either CentOS or Workload Balancing or both, restart the Workload Balancing virtual
appliance.
Note:
After upgrading CentOS, the operating-system time changes from Coordinated Universal Time
(UTC) to the local time zone. One side effect of this change is that the timestamps in the
Workload Balancing log file (LogFile.log) also change from UTC time to local time. If you want
to change the system time back to UTC time, run the following command in the Workload
Balancing virtual appliance: rm -rf /etc/localtime.
50
51
6.2.3. Checking Workload Balancing and CentOS Version Numbers
• To check the Workload Balancing virtual appliance build number, run the following command on a XenServer
host in a pool being monitored by that VPX:
xe pool-retrieve-wlb-diagnostics | grep -i " version=" |
grep -i "DwmDataColSvc.exe"
The Workload Balancing version number appears at the top of the output.
• To check the CentOS version number, run the following command at the bash prompt of the Workload
Balancing appliance:
cat /etc/redhat-release
52
Chapter 7. Appendix A: XenServer
Workload Balancing Troubleshooting
7.1. Troubleshooting Workload Balancing
While Workload Balancing usually runs smoothly, this series of topics provides guidance in case you encounter
issues.
7.1.1. General Troubleshooting Tips
• Start troubleshooting by reviewing the Workload Balancing log files (LogFile.log and wlb_install_log.log). You
can find these logs in Workload Balancing virtual appliance in this location (by default):
• /var/log/wlb
• Check the logs in the XenCenter Logs tab for more (different) information.
• To check the Workload Balancing virtual appliance build number, run the following command on a XenServer
host in a pool being monitored by that VPX:
xe pool-retrieve-wlb-diagnostics | more
The Workload Balancing version number appears at the top of the output.
7.1.2. Error Messages
Workload Balancing displays errors on screen as dialog boxes and as error messages in the Logs tab in XenCenter.
If an error message appears, review the XenCenter event log for additional information. For information about
the location of this log, see the XenCenter Help.
7.1.3. Issues Entering Workload Balancing Credentials
If you cannot successfully enter the virtual appliance user account and password while configuring the Connect
to WLB Server dialog, try the following:
• Make sure that Workload Balancing virtual appliance imported and was configured correctly and all of its
services are running. See Section 8.1.3, “wlb start”.
• Check to make sure you are entering the correct credentials. The default credentials appear in the Workload
Balancing Quick Start.
• You can enter a host name in the Address box, but it must be the fully qualified domain name (FQDN) of the
Workload Balancing virtual appliance (and not the host name of the physical server hosting the appliance). For
example, yourcomputername.yourdomain.net. If you are having trouble entering a computer name,
try using the Workload Balancing appliance's IP address instead.
• Verify that the host is using the correct DNS server and the XenServer host can contact Workload Balancing
server using its FQDN. To do this, ping the Workload Balancing appliance using its FQDN from the XenServer
host. For example, enter the following in the XenServer host console:
ping wlb-vpx-1.mydomain.net
7.1.4. Issues with Firewalls
The following error appears if the Workload Balancing virtual appliance is behind a (hardware) firewall, and you
did not configure the appropriate firewall settings: "There was an error connecting to the Workload Balancing
53
server: <pool name> Click Initialize WLB to reinitialize the connection settings." This error may also appear if the
Workload Balancing appliance is otherwise unreachable.
Resolution
If the Workload Balancing virtual appliance is behind a firewall, you must open port 8012.
Likewise, the port XenServer uses to contact Workload Balancing (8012 by default), must match the port number
specified when you ran the Workload Balancing Configuration wizard.
7.1.5. Losing the Connection to Workload Balancing
If, after configuring and connecting to Workload Balancing, you receive a connection error, the credentials may
no longer be valid. To isolate this issue, try:
• Verifying the credentials you entered in the Connect to WLB Server dialog box match the credentials:
• You created during Workload Balancing Configuration
• On XenServer (that is, the pool master credentials)
• Verifying the IP address or FQDN for the Workload Balancing virtual appliance that you entered in the Connect
to WLB Server dialog box is correct.
• Verifying the user name you created during Workload Balancing Configuration matches the credentials you
entered in the Connect to WLB Server dialog box.
7.1.6. Workload Balancing Connection Errors
If you receive a connection error in the Workload Balancing Status line on the WLB tab, you might need to
reconfigure Workload Balancing on that resource pool.
Click the Connect button on the WLB tab and reenter the server credentials.
7.1.7. Workload Balancing Stops Working
If Workload Balancing does not work (for example, it does not let you save any changes to settings, or star ratings
and/or optimization recommendations do not appear), check the Workload Balancing log file for the following
error message:
"dwmdatacolsvc.exe: Don't have a valid pool. Trying again in 10 minutes."
Cause
This typically occurs in pools that have one or more problematic VMs. When VMs are problematic, this may
become evident as follows:
• Windows. The Windows VM crashes due to a stop error ("blue screen").
• Linux. The Linux VM may be unresponsive in the console and typically does not shut down.
Workaround
1. Force the VM to shut down. To do so, you can do one of the following on the host with the problematic VM:
• In XenCenter, select the VM, and then from the VM menu, click Force Shutdown.
• Run the vm-shutdown xe command with the <force> parameter set to true as described in the XenServer
Administrator's Guide. For example:
xe vm-shutdown \
force=<true> \
uuid=<vm_uuid> \
54
You can find the host UUID on the General tab for that host (in XenCenter) or by running the host-list xe
command. You can find the VM UUID on the General tab for the VM or by running the vm-list xe command
as described in the XenServer Administrator's Guide .
2. In the xsconsole of the XenServer hosting the crashed VM or in XenCenter, migrate all of the VMs to another
host, then run the xe-toolstack-restart command.
7.1.8. Issues Changing Workload Balancing Servers
If you connect a resource pool to a different Workload Balancing server without disconnecting from Workload
Balancing, both the old and new Workload Balancing server will monitor the pool.
To solve this problem, you can either shutdown and delete the old Workload Balancing virtual appliance or
manually stop the Workload Balancing services (analysis, data collector and Web service) so that it will no longer
monitor the pool.
Note:
Citrix does not recommend using the pool-deconfigure-wlb xe command to disconnect a pool
from the Workload Balancing virtual appliance or using the pool-initialize-wlb xe command
to specify a different appliance.
55
Chapter 8. Appendix B: Workload Balancing
Commands
This appendix provides a reference for the Workload Balancing commands that you can perform from a XenServer
host or console to control Workload Balancing or configure Workload Balancing settings on the XenServer host.
This appendix includes xe commands and service commands.
8.1. Service Commands
Run the following service commands on the Workload Balancing appliance. To do so, you must log in to the
Workload Balancing virtual appliance.
8.1.1. Logging in to the Workload Balancing Virtual Appliance
Before you can run any service commands or edit the wlb.conf file, you must log in to the Workload Balancing
virtual appliance. To do so, you must enter a user name and password. Unless you created additional user
accounts on the virtual appliance, log in using the root user account. You specified this account when you ran
Workload Balancing Configuration wizard (before you connected your pool to Workload Balancing). You can,
optionally, use the Console tab in XenCenter to log in to the appliance.
To log in to the Workload Balancing virtual appliance
1.
At the <name-of-your-WLB-VPX> login prompt, enter the account user name. For example, where wlb-vpxpos-pool is the name of your Workload Balancing appliance:
wlb-vpx-pos-pool login: root
2.
At the Password prompt, enter the password for the account:
wlb-vpx-pos-pool login: root
Note:
To log off the Workload Balancing virtual appliance, simply type logout at the command
prompt.
8.1.2. wlb restart
Run the wlb restart command from anywhere in the Workload Balancing appliance to stop and then restart the
Workload Balancing Data Collection, Web Service, and Data Analysis services.
8.1.3. wlb start
Run the wlb start command from anywhere in the Workload Balancing appliance to start the Workload Balancing
Data Collection, Web Service, and Data Analysis services.
8.1.4. wlb stop
Run the wlb stop command from anywhere in the Workload Balancing appliance to stop the Workload Balancing
Data Collection, Web Service, and Data Analysis services.
8.1.5. wlb status
Run the wlb status command from anywhere in the Workload Balancing appliance to determine the status of the
Workload Balancing server. After you execute this command, the status of the three Workload Balancing services
(the Web Service, Data Collection Service, and Data Analysis Service) is displayed.
56
8.1.6. Modifying the Workload Balancing Configuration Options
Many Workload Balancing configurations, such as the database and web-service configuration options, are stored
in the wlb.conf file, a configuration file on the Workload Balancing virtual appliance.
To make it easier to modify the most commonly used options, Citrix provides a command, wlb config. Running
the wlb config command on the Workload Balancing virtual appliance lets you rename the Workload Balancing
user account, change its password, or change the PostgreSQL password. After you execute this command, the
Workload Balancing services are restarted.
To run the wlb config command
•
Run the following from the command prompt:
wlb config
The screen displays a series of questions guiding you through changing your Workload Balancing user name
and password and the PostgreSQL password. Follow the questions on the screen to change these items.
Important:
Double-check any values you enter in the wlb.conf file: Workload Balancing does not validate
values in the wlb.conf file. Consequently, if the configuration parameters you specify are not
within the required range, Workload Balancing does not generate an error log.
8.1.7. Editing the Workload Balancing Configuration File
You can modify Workload Balancing configuration options by editing the wlb.conf file, which is stored in /opt/
citrix/wlb directory on the Workload Balancing virtual appliance. In general, Citrix does not recommend changing
the settings in this file without guidance from Citrix; however, there are three categories of settings you can
change if desired:
• Workload Balancing account name and password. It is easier to modify these by running the wlb config
command.
• Database password. This can be modified using the wlb.conf file. However, Citrix recommends modifying it
through the wlb config command since this command modifies the wlb.conf file and automatically updates the
password in the database. If you choose to modify the wlb.conf file instead, you must run a query to update
the database with the new password.
• Database grooming parameters. You can modify database grooming parameters, such as the database
grooming interval, using this file by following the instructions in the database-management section of the
Workload Balancing Administrator's Guide. However, Citrix recommends using caution if you do so.
For all other settings in the wlb.conf file, Citrix currently recommends leaving them at their default, unless Citrix
instructed you to modify them.
To edit the wlb.conf file
1.
Run the following from the command prompt on the Workload Balancing virtual appliance (using VI as an
example):
vi /opt/citrix/wlb/wlb.conf
The screen displays several different sections of configuration options.
2.
Modify the configuration options, and exit the editor.
You do not need to restart Workload Balancing services after editing the wlb.conf file. The changes go into
effect immediately after exiting the editor.
Important:
57
Double-check any values you enter in the wlb.conf file: Workload Balancing does not validate
values in the wlb.conf file. Consequently, if the configuration parameters you specify are not
within the required range, Workload Balancing does not generate an error log.
8.1.8. Increasing the Detail in the Workload Balancing Log
The Workload Balancing log provides a list of events on the Workload Balancing virtual appliance, including
actions for the analysis engine, database, and audit log. This log file is found in this location: /var/log/wlb/
LogFile.log.
You can, if desired, increase the level of detail the Workload Balancing log provides. To do so, modify the Trace
flags section of the Workload Balancing configuration file (wlb.conf), which is found in the following location: /
opt/citrix/wlb/wlb.conf. Enter a 1 or true to enable logging for a specific trace and a 0 or false to disable logging.
For example, to enable logging for the Analysis Engine trace, enter:
AnalEngTrace=1
You may want to increase logging detail before reporting an issue to Citrix Technical Support or when
troubleshooting.
Logging
Option
Trace Flag
Benefit or Purpose
Analysis
Engine
Trace
AnalEngTrace
Logs details of the analysis engine calculations.
Database
Trace
DatabaseTrace
Data
Collection
Trace
DataCollectionTrace
Shows details of the decisions the analysis engine is making and
potentially gain insight into the reasons Workload Balancing is
not making recommendations.
Logs details about database reads/writes.
However, leaving this trace on will increase the log file size
quickly.
Logs the actions of retrieving metrics.
This lets you see the metrics Workload Balancing is retrieving
and inserting into the Workload Balancing data store. However,
leaving this trace on will increase the log file size quickly.
Data
DataCompactionTrace
Compaction
Trace
Logs details about how many milliseconds it took to compact
the metric data.
Data Event
Trace
DataEventTrace
This trace provides details about events Workload Balancing
catches from XenServer.
Data
Grooming
Trace
DataGroomingTrace
This trace provides details about the database grooming.
Data
Metrics
Trace
DataMetricsTrace
Logs details about the parsing of metric data.
Queue
QueueManagementTrace
Management
Trace
Leaving this trace on will increase the log-file size quickly.
Logs details about data collection queue management
processing.
(This is for internal use.)
58
Logging
Option
Trace Flag
Benefit or Purpose
Data Save
Trace
DataSaveTrace
Logs details about the pool being saved to the database.
Score Host
Trace
ScoreHostTrace
Logs details about how Workload Balancing is arriving at a score
for a host.
This trace shows the detailed scores generated by Workload
Balancing when it calculates the star ratings for selecting
optimal servers for VM placement.
Audit
Trace
Log
AuditLogTrace
Shows the action of the audit log data being captured and
written.
(This is generally only for internal use and does not provide
information that is captured in the audit log.)
However, leaving this trace on will increase the log file size
quickly.
Scheduled
Task Trace
ScheduledTaskTrace
Web
Service
Trace
WlbWebServiceTrace
Logs details about scheduled tasks.
For example, if your scheduled mode changes are not working,
you might want to enable this trace to investigate the cause.
Logs details about the communication with the web-service
interface.
8.2. Workload Balancing XE Commands
Commands for controlling the Workload Balancing feature.
8.2.1. pool-initialize-wlb
pool-initialize-wlb wlb_url=<wlb_server_address> \
wlb_username=<wlb_server_username> \
wlb_password=<wlb_server_password> \
xenserver_username=<pool_master_username> \
xenserver_password=<pool_master_password>
Initializes the connection between the Workload Balancing server and the XenServer pool.
Note:
Initializing a pool requires running two commands. After executing the pool-initialize-wlb
command, execute xe pool-param-set wlb-enabled=true uuid=<pool-uuid>[].
8.2.2. pool-param-set other-config
Use the pool-param-set other-config command to specify the timeout when communicating with the Workload
Balancing server. All requests are serialized, and the timeout covers the time from a request being queued to its
response being completed. In other words, slow calls cause subsequent ones to be slow. Defaults to 30 seconds
if unspecified or if the setting cannot be parsed.
xe pool-param-set other-config:wlb_timeout=<85> \
uuid=<315688af-5741-cc4d-9046-3b9cea716f69>
59
8.2.3. pool-retrieve-wlb-diagnostics
Use the pool-retrieve-wlb-diagnostics command to return the contents of the Workload Balancing Log file for the
Workload Balancing server that the pool is currently using. This command also returns some version information.
xe pool-retrieve-wlb-diagnostics \
8.2.4. host-retrieve-wlb-evacuate-recommendations
host-retrieve-wlb-evacuate-recommendations uuid=<host_uuid>
Returns the evacuation recommendations for a host, and a reference to the UUID of the recommendations object.
8.2.5. vm-retrieve-wlb-recommendations
Returns the workload balancing recommendations for the selected VM. The simplest way to select the VM on
which the operation is to be performed is by supplying the argument vm=<name_or_uuid>. VMs can also be
specified by filtering the full list of VMs on the values of fields. For example, specifying power-state=halted
selects all VMs whose power-state is halted. Where multiple VMs are matching, specify the option --multiple
to perform the operation. The full list of fields that can be matched can be obtained by the command xe vm-list
params=all. If no parameters to select VMs are given, the operation will be performed on all VMs.
8.2.6. pool-certificate-list
XenServer Workload Balancing component lets you use certificates to secure communication between XenServer
pools and the Workload Balancing server. You can either use the default test certificate, which is created
automatically during Workload Balancing configuration, or you can specify a certificate from a Trusted Authority.
To use a Trusted Authority certificate, the certificate must be in X.509 format and comply with the guidelines
provided earlier in the Workload Balancing Administrator's Guide. If you want to verify the presence of a
certificate from a Trusted Authority, you must:
pool-certificate-list
Lists all installed SSL certificates.
8.2.7. pool-certificate-install
pool-certificate-install filename=<certificatefilename>
Run this command on the pool to install the certificate on the Workload Balancing virtual appliance on the pool
master.
8.2.8. pool-certificate-sync
pool-certificate-sync
Run this command on the pool, after running the pool-certificate-install command, to make sure the certificate
and certificate revocation lists are synchronized from the pool master to all slaves on the pool.
8.2.9. pool-param-set
pool-param-set wlb-verify-cert=<true> uuid=<uuid_of_pool>
Run this command on the pool, after running the pool-certificate-sync command, to make XenServer always
verify the certificate when communicating with the Workload Balancing server.
Tip:
Pressing the Tab key automatically populates the UUID of the pool.
60
8.2.10. pool-deconfigure-wlb
Permanently deletes all workload balancing configuration.
8.2.11. pool-retrieve-wlb-configuration
Prints all Workload Balancing configuration information to standard out.
8.2.12. pool-retrieve-wlb-recommendations
Prints all Workload Balancing recommendations to standard out.
8.2.13. pool-retrieve-wlb-report
Gets a Workload Balancing report of the specified type and saves it to the specified file. XenCenter uses this
command when executing Workload Balancing reports. However, running this command from the CLI does not
output report data in a graphical format. Outputting reports requires using the XenCenter Pool > View Workload
Reports menu command. The available reports are:
• pool_health
• host_health_history
• optimization_performance_history
• pool_optimization_history
• pool_health_history
• vm_chargeback_history
• pool_audit_history
• vm_movement_history
• vm_performance_history
Example usage for each report type is shown below. The utcoffset parameter specifies the number of hours
the pool's time zone is ahead or behind Coordinated Universal Time (UTC). For example, <-5> represents the
Eastern Standard Time zone in North America, which is five hours behind UTC. The start parameter and end
parameters specify the number of hours to report. For example specifying start=-3 and end=0 will cause
Workload Balancing to report on the last three hours of activity. LocaleCode represents the language for the
report output. You can enter either <en> for English or <ja> for Japanese.
xe pool-retrieve-wlb-report report=pool_health \
LocaleCode=<en> \
Start=<-3> \
End=<0> \
PoolID=<c75f9ee9-422f-9cde-4781-24f6cbd9279c> \
UTCOffset=<-5> \
filename=/<pool_health>.txt \
xe pool-retrieve-wlb-report report=host_health_history \
Start=<-3> \
End=<0> \
PoolID=<c75f9ee9-422f-9cde-4781-24f6cbd9279c> \
HostID=<471626e5-8e8b-429b-9990-6013726d7e86> \
UTCOffset=<-5> \
filename=/<host_health_history.txt> \
xe pool-retrieve-wlb-report report=optimization_performance_history \
LocaleCode=<en> \
Start=-3 \
End=0 \
PoolID=c75f9ee9-422f-9cde-4781-24f6cbd9279c \
UTCOffset=-5 \
filename=</optimization_performance_history>.txt \
61
xe pool-retrieve-wlb-report report=pool_health_history \
LocaleCode=<en> \
Start=<-3> \
End=<0> \
PoolID=<c75f9ee9-422f-9cde-4781-24f6cbd9279c> \
UTCOffset=<-5> \
filename=</pool_health_history>.txt \
xe pool-retrieve-wlb-report report=vm_movement_history \
LocaleCode=<en> \
Start=<-3> \
End=<0> \
PoolID=<c75f9ee9-422f-9cde-4781-24f6cbd9279c> \
UTCOffset=<-5> \
filename=</vm_movement_history>.txt \
xe pool-retrieve-wlb-report report=vm_performance_history \
LocaleCode=<en> \
Start=<-3> \
End=<0> \
PoolID=<c75f9ee9-422f-9cde-4781-24f6cbd9279c> \
HostID=<471626e5-8e8b-429b-9990-6013726d7e86> \
UTCOffset=<-5> \
filename=</vm_performance_history>.txt \
xe pool-retrieve-wlb-report report=pool_audit_history \
LocaleCode=<en> \
Start=<-3> \
End=<0> \
PoolID=<c75f9ee9-422f-9cde-4781-24f6cbd9279c> \
ReportVersion=<Creedence> \
AuditUsers=<ALL> \
AuditObject=<ALL> \
StartLine=<1/> \
EndLine=<10000> \
UTCOffset=<-5> \
filename=</pool_audit_history>.txt \
xe pool-retrieve-wlb-report report=pool_optimization_history
LocaleCode=<en>
Start=<-3>
End=<0>
PoolID=<c75f9ee9-422f-9cde-4781-24f6cbd9279c>
UTCOffset=<-5>
filename=</pool_optimization_history>.txt
xe pool-retrieve-wlb-report report=vm_chargeback_history
LocaleCode=<en>
Start=<-3>
End=<0>
PoolID=<c75f9ee9-422f-9cde-4781-24f6cbd9279c>
UTCOffset=<-5>
filename=</vm_chargeback_history>.txt
8.2.14. pool-send-wlb-configuration
Modifies Workload Balancing configuration settings, including thresholds, Workload Balancing powermanagement settings, and weightings. It is not mandatory to configure all settings with the command. You can
configure only some parameters, but not all, if desired.
Before using the pool-send-wlb-configuration command, learn about the default values on your system by
running pool-retrieve-wlb-configuration.
When you run the pool-retrieve-wlb-configuration command, additional parameters appear that are not
documented in this section. Citrix does not recommend editing these parameters.
62
pool-send-wlb-configuration
[config:HostMemoryThresholdCritical=<HostCpuThresholdCritical=value>\
config:HostMemoryThresholdHigh=<HostMemoryThresholdHigh=value>\
config:HostPifReadThresholdCritical=<HostPifReadThresholdCritical=value>\
config:HostPifReadThresholdHigh=<HostPifReadThresholdHigh=value>\ config:set_host_configuration=<true |
false> ...\ ]
Use the pool-send-wlb-configuration command with the arguments <ParticipatesInPowerManagement> and
<set_host_configuration> to configure Workload Balancing's Host Power Management feature.
xe pool-send-wlb-configuration \
config:<host_21_> \
ParticipatesInPowerManagement=<true> \
config:set_host_configuration=<true>
63
Chapter 9. Appendix C: Configuring
Certificates for Workload Balancing
This chapter provides information about two optional tasks for securing certificates:
• Configuring XenServer to verify a certificate from a Trusted Authority
• Configuring XenServer to verify the default Citrix WLB self-signed certificate
Important:
Certificate configuration has changed for this release. If you upgraded a Workload Balancing
Balancing 6.x virtual appliance to Workload Balancing 6.5 and you did not have certificates
configured before the upgrade, you must use the certificate instructions in the Workload
Balancing 6.0 Administrator's Guide (and not this guide) if you configure certificates.
9.1. Overview
XenServer and Workload Balancing communicate over HTTPS. Consequently during Workload Balancing
Configuration, the wizard automatically creates a self-signed test certificate. This self-signed test certificate lets
Workload Balancing establish an SSL connection to XenServer.
Note:
The self-signed certificate is a placeholder to facilitate HTTPS communication and is not from a
trusted certificate authority. For added security, Citrix recommends using a certificate signed
from a trusted certificate authority.
By default, Workload Balancing creates this SSL connection with XenServer automatically. You do not need
to perform any certificate configurations during or after configuration for Workload Balancing to create this
SSL connection. However, to use a certificate from another certificate authority, such as a signed one from a
commercial authority, you must configure Workload Balancing and XenServer to use it.
Regardless of what certificate Workload Balancing uses, by default, XenServer does not validate the identity of the
certificate before it establishes connection to Workload Balancing. To configure XenServer to check for a specific
certificate, you must export the root certificate that was used to sign the certificate, copy it to XenServer, and
configure XenServer to check for it when a connection to Workload Balancing is made. In this scenario, XenServer
acts as the client and Workload Balancing acts as the server.
This illustration shows how XenServer verifies that a specific certificate is present before it lets the Workload
Balancing virtual appliance establish a connection to it over SSL. In this case, the real certificate (the certificate
with the private key) is on the Workload Balancing server and the certificate that was used to sign it, is on the
XenServer pool master.
Depending on your security goals, you can either:
64
• Configure XenServer to verify the self-signed test certificate. See Section 9.2, “Configuring XenServer to Verify
the Self-Signed Certificate”.
• Configure XenServer to verify a certificate from a trusted certificate authority. See Section 9.3, “Configuring
XenServer to Verify a Certificate-authority Certificate”.
9.2. Configuring XenServer to Verify the Self-Signed Certificate
You can configure XenServer to verify that the Citrix WLB self-signed certificate is authentic before XenServer
permits Workload Balancing to connect.
Important:
To verify the Citrix WLB self-signed certificate, you must connect to Workload Balancing using
its host name. To find the Workload Balancing host name, run hostname command on the
virtual appliance.
If you want to configure Workload Balancing to verify the Citrix WLB self-signed certificate, perform the steps
in the procedure that follows.
To configure XenServer to verify the self-signed certificate
1.
Copy the self-signed certificate from the Workload Balancing virtual appliance to the pool master. The Citrix
WLB self-signed certificate is stored at /etc/ssl/certs/server.pem. Run the following on the pool master to
copy the certificate:
scp [email protected]<wlb-ip>:/etc/ssl/certs/server.pem .
2.
If you receive a message stating that the authenticity of <wlb-ip> cannot be established, type yes to continue.
3.
Enter Workload Balancing virtual appliance root password when prompted, the certificate will be copied to
current directory.
4.
Install the certificate. Run the pool-certificate-install command from the directory where you copied the
certificate. For example:
xe pool-certificate-install filename=server.pem
5.
Verify the certificate was installed correctly by running the pool-certificate-list command on the pool
master:
xe pool-certificate-list
If you installed the certificate correctly, the output of this command includes the exported root certificate
(for example, server.pem). Running this command lists all installed SSL certificates, including the certificate
you just installed.
6.
Synchronize the certificate from the master to all hosts in the pool by running the pool-certificate-sync
command on the pool master:
xe pool-certificate-sync
Running the pool-certificate-sync command on the master synchronizes the certificate and certificate
revocation lists on all the pool servers with the master. This ensures all hosts in the pool use the same
certificate(s).
There is no output from this command. However, the next step does not work if this one did not work
successfully.
7.
Instruct XenServer to verify the certificate before connecting to the Workload Balancing virtual appliance.
Run the following command on the pool master:
xe pool-param-set wlb-verify-cert=<true> uuid=<uuid_of_pool>
Tip:
65
Pressing the Tab key automatically populates the UUID of the pool.
8.
(Optional.) To verify this procedure worked successfully, perform the following steps:
a.
To test if the certificate synchronized to the other hosts in pool, run the pool-certificate-list command
on those hosts.
b.
To test if XenServer was set to verify the certificate, run the pool-param-get command with the
param-name=<wlb-verify-cert> parameter. For example:
xe pool-param-get param-name=<wlb-verify-cert> uuid=<uuid_of_pool>
9.3. Configuring XenServer to Verify a Certificate-authority
Certificate
You can configure XenServer to verify a certificate signed by a trusted certificate authority.
For trusted authority certificates, XenServer requires an exported certificate or certificate chain (the intermediate
and root certificates) in .pem format that contains the public key.
If you want Workload Balancing to use a trusted authority certificate, do the following:
1. Obtain a signed certificate from the certificate authority. See Section 9.3.1, “Task 1: Obtaining a Certificateauthority Certificate”.
2. Follow the instructions in Section 9.3.2, “Task 2: Specifying the New Certificate” to specify and apply the new
certificate.
3. Install the obtained certificates and enable certificate verification on the pool master. See Section 9.3.3, “Task
3: Importing the Certificate Chain into the Pool”.
Before beginning these tasks, make sure:
• You know the IP address for the XenServer pool master.
• XenServer can resolve the Workload Balancing host name. (For example, you can try pinging the Workload
Balancing FQDN from the XenServer console for the pool master.)
Important:
If you want to use an IP address to connect to Workload Balancing, you must specify that IP
address as the Subject Alternative Name (SAN) when you create the certificate.
9.3.1. Task 1: Obtaining a Certificate-authority Certificate
To obtain a certificate from a certificate authority, you must generate a Certificate Signing Request (CSR).
Generating a CSR for the Workload Balancing virtual appliance is two-task process. You must (1) create a private
key and (2) use that private key to generate the CSR. You must perform both of these procedures on the Workload
Balancing virtual appliance.
Guidelines for Specifying the Common Name
The Common Name (CN) you specify when creating a CSR must exactly match the FQDN of your Workload
Balancing virtual appliance and the FQDN or IP address you specified in the Address box in the Connect to WLB
Server dialog box.
To ensure the name matches, specify the Common Name using one of these guidelines:
• Specify the same information for the certificate's Common Name as you specified in the Connect to WLB Server
dialog. For example, if your Workload Balancing virtual appliance is named “wlb-vpx.yourdomain,” specify
66
“wlb-vpx.yourdomain” in the Connect to WLB Server and provide “wlb-vpx.yourdomain” as the Common Name
when creating the CSR.
• If you connected your pool to Workload Balancing using an IP address, use the FQDN as the Common Name
and specify the IP address as a Subject Alternative Name (SAN). However, this may not work in all situations.
Note:
Certificate verification is a security measure designed to prevent unwanted connections. As
a result, Workload Balancing certificates must meet strict requirements or the certificate
verification will not succeed and XenServer will not allow the connection. Likewise, for
certificate verification to succeed, you must store the certificates in the specific locations in
which XenServer expects to find the certificates.
To create a private key file
1.
Create a private key file:
openssl genrsa -des3 -out <privatekey.pem> 2048
2.
Remove the password:
openssl rsa -in <privatekey.pem> -out <privatekey.nop.pem>
Note:
If you enter the password incorrectly or inconsistently, you may receive some messages
indicating the there is a user interface error. You can ignore the message and just rerun the
command to create the private key file.
To generate the CSR
1.
Generate the CSR:
a.
Create the CSR using the private key:
openssl req -new -key <privatekey.nop.pem> -out <csr>
b.
Follow the prompts to provide the information necessary to generate the CSR:
Country Name. Enter the SSL Certificate country codes for your country. For example, CA for Canada
or JM for Jamaica. You can find a list of SSL Certificate country codes on the web.
State or Province Name (full name). Enter the state or province where the pool is located. For example,
Massachusetts or Ontario.
Locality Name. The name of the city where the pool is located.
Organization Name. The name of your company or organization.
Organizational Unit Name. Enter the department name. This field is optional.
Common Name. Enter the FQDN of your Workload Balancing server. This must match the name the
pool uses to connect to Workload Balancing.
Email Address. This email address is included in the certificate when you generate it.
c.
Provide optional attributes or click Enter to skip providing this information.
The CSR request is saved in the current directory and is named csr.
2.
Display the CSR in the console window by running the following commands in the Workload Balancing
appliance console:
cat csr
67
3.
Copy the entire Certificate Request and use the CSR to request the certificate from the certificate authority.
9.3.2. Task 2: Specifying the New Certificate
Use this procedure to specify Workload Balancing use a certificate from a certificate authority. This procedure
installs the root and (if available) intermediate certificates.
To specify a new certificate
1.
Download the signed certificate, root certificate and, if the certificate authority has one, the intermediate
certificate from the certificate authority.
2.
If you did not download the certificates to the Workload Balancing virtual appliance. Do one of the following:
a.
If you are copying the certificates from a Windows computer to the Workload Balancing appliance, use
WinSCP or another copying utility, to copy the files.
For the host name, you can enter the IP address and leave the port at the default. The user name and
password are typically root and whatever password you set during configuration.
b.
If you are copying the certificates from a Linux computer to the Workload Balancing appliance, use SCP
or another copying utility, to copy the files to the directory of your choice on the Workload Balancing
appliance. For example:
scp <root_ca.pem> [email protected]<wlb-ip>:/<path_on_your_WLB>
3.
On the Workload Balancing virtual appliance, merge contents of all the certificates (root certificate,
intermediate certificate (if it exists), and signed certificate) into one file. For example:
cat <signed_cert.pem> <intermediate_ca.pem> <root_ca.pem> > <server.pem>
4.
Rename the existing certificate and key using the move command:
mv /etc/ssl/certs/server.pem /etc/ssl/certs/server.pem_orig
mv /etc/ssl/certs/server.key /etc/ssl/certs/server.key_orig
5.
Copy the merged certificate:
mv <server.pem> /etc/ssl/certs/server.pem
6.
Copy the private key created in this step:
mv <privatekey.nop.pem> /etc/ssl/certs/server.key
7.
Make the private key readable only by root. Use chmod command to fix permissions.
chmod 600 /etc/ssl/certs/server.key
8.
Restart stunnel:
killall stunnel
stunnel
9.3.3. Task 3: Importing the Certificate Chain into the Pool
After obtaining certificates, you must import (install) the certificates onto the XenServer pool master and
synchronize the hosts in the pool to use those certificates. Then, you can configure XenServer to check the
certificate's identity and validity each time Workload Balancing connects to a XenServer host.
1.
Copy the signed certificate, root certificate and, if the certificate authority has one, the intermediate
certificate from the certificate authority onto the XenServer pool master.
2.
Install the root certificate on the pool master:
xe pool-certificate-install filename=<root_ca>.pem
3.
If applicable, install the intermediate certificate on the pool master:
68
xe pool-certificate-install filename=<intermediate_ca>.pem
4.
Verify both the certificates installed correctly by running this command on the pool master:
xe pool-certificate-list
Running this command lists all installed SSL certificates. If the certificates installed successfully, they appear
in this list.
5.
Synchronize the certificate on the pool master to all hosts in the pool:
xe pool-certificate-sync
Running the pool-certificate-sync command on the master synchronizes the certificates and certificate
revocation lists on all the pool servers with the pool master. This ensures all hosts in the pool use the same
certificates.
6.
Instruct XenServer to verify a certificate before connecting to the Workload Balancing virtual appliance. Run
the following command on the pool master:
xe pool-param-set wlb-verify-cert=<true> uuid=<uuid_of_pool>
Tip:
Pressing the Tab key automatically populates the UUID of the pool.
7.
If, before you enabled certificate verification, you specified an IP address in the Connect to WLB dialog, you
may be prompted to reconnect the pool to Workload Balancing.
In this case, specify the FQDN for the Workload Balancing appliance in the Address box in the Connect to
WLB dialog—exactly as it appears in the certificate's Common Name (CN). (You must enter the FQDN since
the Common Name and the name XenServer uses to connect must match.)
Troubleshooting Tips
• If, after configuring certificate verification, the pool cannot connect to Workload Balancing, check to see if the
pool can connect if you turn certificate verification off (by running xe pool-param-set wlb-verify-cert=<true>
uuid=<uuid_of_pool>). If it can connect with verification off, the issue is in your certificate configuration. If it
cannot connect, the issue is in either your Workload Balancing credentials or your network connection.
• Some commercial certificate authorities, such as Verisign, provide tools to verify the certificate installed
correctly. Consider running these tools if these procedures fail to help isolate the issue. If these tools require
specifying an SSL port, specify port 8012 or whatever port you set during Workload Balancing Configuration.
• If, after following these procedures, an error message appears on the WLB tab stating, “There was an error
connecting to the WLB server,” there may be a conflict between the Common Name in the certificate and the
name of the Workload Balancing virtual appliance. The Workload Balancing virtual-appliance name and the
Common Name of the certificate must match exactly.
69
Index
disk size
increasing, 43
displaying
settings dialog, 18
A
E
adjusting
optimization mode, 19
aggressiveness
configuring, 32
automated recommendations
advanced settings, 30
controlling, 30
overview, 21
automatic
optimization acceptance, 21
enabling
power management, 21
excluding
hosts from recommendations, 29
F
fixed optimization mode
configuring, 19
H
C
High Availability
interaction with WLB, 7
certificates
certificate authority certificate, verifying, 66
certificate chain, importing, 68
Certificate Signing Request, generating, 66
configuring, 64
new, specifying, 68
self-signed, verifying, 65
changing
PostgreSQL account, 57
WLB server, 39
Workload Balancing account, 57
commands
Workload Balancing, 56
configuring
critical thresholds, 27
power management, 20
connecting
pool to WLB, 36
consolidation process
Maximum Density Mode, 26
critical thresholds
configuring, 27
default settings, 27
customizing
Workload Balancing, 47
I
IP address
changing Workload Balancing, 41
obtaining for WLB, 39
L
logging in
Workload Balancing appliance, 56
logs
enabling traces, 57
increasing detail in logs, 58
M
Maximum Density Mode
consolidation process, 26
Maximum Performance Mode
recommendation process, 25
metric weightings
configuring, 28
O
optimal server
resuming VM on, 6
starting, 5
optimization mode
fixed mode, 19
scheduled mode, 19
scheduling, 19
settings, 19
optimization recommendations
accepting, 7
how WLB makes, 24
options
configuring, 18
D
database
archiving, 45
changing account, 57
changing maintenance time, 46
client authentication, 45
configuration files, 45
grooming, 44, 57
grooming parameters, 44
managing, 43
password, 45
pg_hba.conf file, 45
postgresql.conf file, 46
disconnecting
Workload Balancing, 42
P
pausing
Workload Balancing, 41
70
PostgreSQL
changing password, 57
power management
configuring, 20
enabling, 21
increasing detail in logs, 58
troubleshooting
connection to WLB, 54
error messages, 53
firewalls, 53
issues changing servers, 55
overview, 53
recommendations, stop, 54
WLB credentials, 53
Trusted Authority certificates
configuring, 64
R
RBAC
Workload Balancing, 39
recommendation count
setting, 31
recommendation process
Maximum Performance Mode, 25
recommendation severity
setting, 32
recommendations
accepting, 7
accepting, optimization, 7
excluding hosts from, 29
reconfiguring
pool to use different WLB server, 39
renaming
WLB user account, 57
reports
Chargeback, 11
generating, 9
glossary, 11
Host Health History, 13
introduction, 3
overview, 9
Pool Health, 15
Pool Health History, 16
Pool Optimization History, 16
Pool Optimization Performance History, 13
Virtual Machine Motion History, 17
Virtual Machine Performance History, 17
requirements
Workload Balancing, 3
resource utilization
process for evaluating, 24
U
uninstalling
Workload Balancing, 43
updating credentials
Workload Balancing, 40
upgrading
CentOS, 48
Internet access, with, 49
Internet access, without, 49
user accounts
setting permissions, 39
V
virtual disk size
increasing, 43
VM migration interval
setting, 31
VM placement
how to start VM, 5
overview, 4
W
wlb config
running (proc.), 57
wlb.conf
changing, 57
editing (proc.), 57
wlbconf
editing (proc.), 57
wlbconfig
running command, 57
Workload Balancing
administering, 36
administering, overview, 36
commands, 56
configuring, 18, 18
customizing, 47
disconnecting, 42
IP address, 39
Maintenance Mode, 42
overview, 2
pausing, 41
reports, 9
requirements, 3
settings, 18, 18
S
scheduled optimization mode
configuring, 19
scheduling
optimization mode, 19
securing
Workload Balancing, 64
settings
configuring, 18, 18
displaying dialog, 18
star ratings, 4
stopping
Workload Balancing, 41
T
trace flags
71
stopping, 41
uninstalling, 43
using, 4
Workload Balancing account
changing, 57
Workload Balancing commands
wlb restart, 56
wlb start, 56
wlb status, 56
wlb stop, 56
workload reports
Chargeback, 11
generating, 9
glossary, 11, 11
Host Health History, 13
introduction, 3
overview, 9
Pool Health, 15
Pool Health History, 16
Pool Optimization History, 16
Pool Optimization Performance History, 13
Virtual Machine Motion History, 17
Virtual Machine Performance History, 17
72
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement