Management Processors (Windows Server 2012, Windows Server 2008, and Windows Server 2003)

Applies To: Operations Manager 2007, System Center 2012 - Operations Manager, System Center 2012 R2 Operations Manager, System Center 2012 SP1 - Operations Manager, Windows Server 2012, Windows Server 2012 R2

Windows Server 2012, Windows Server 2008, and Windows Server 2003 management packs can monitor individual instances of processors or all instances together. By default, the health of the processors is monitored as a total of all instances. If you are interested in monitoring individual processor instances, you can do so by enabling the Object Discoveries for Windows Server 2008 Processor or Object Discoveries for Windows Server 2003 Processor objects. After Object Discoveries has been enabled, the processors will be discovered within 24 hours after which they will become monitored and performance data will be collected. The following is a procedure for enabling Object Discoveries for processors running Windows Server 2008 or Windows Server 2003 for use with the Windows Server 2008 and Windows Server 2003 management packs.

To enable Object Discoveries for Windows Server 2008 Processor

1. Log on to the computer with an account that is a member of the Operations Manager Administrators role for the Operations Manager 2007 management group.

2. In the Operations console, click the Authoring button.

3. In the Administration pane, expand Authoring, expand Management Pack Objects, and then click Object Discoveries.

4. In the Operations Manager 2007 toolbar, click Scope.

5. On the Scope Management Pack Objects by target(s) dialog box, click Clear All.

6. In the Look for box, type Windows Server 2008 Processor, select the Windows Server 2008 Processor check box, and then click OK.

7. In the Object Discoveries pane, expand Discovered Type: Windows Server 2008 Processor.

8. Right-click Discover Windows CPUs, point to Overrides, point to Override the Object Discovery, and click For a specific object of type: Windows Server 2008 Operating System.

   Note

   By selecting For a specific object of type: Windows Server 2008 Operating System, you decide which objects are affected by this change. If you select For all objects of type: Windows Server 2008 Operating System, every object will be affected by this change.

9. In the Select Object dialog box, click the computer on which you want to monitor the individual processors, and then click OK.

10. In the Override Properties dialog box, locate the row with the ParameterName of Enabled, click Override, click the OverrideSetting menu, click True, and then click OK.

After you have enabled an override for the processor, rules, and monitors for the targets, Windows Server 2008 Processor or Windows Server 2003 Processor will start to work on the CPUs.

Monitoring Total Processor Performance

Many rules, tasks, and monitors in the management pack are used for monitoring processor performance. We recommend that you at least monitor the items listed in the following table.

Evaluate the default settings for the following parameters and compare them against your business needs. If your management strategy could benefit from a change in these values, use overrides to make the necessary changes. For more information about overrides, see the "Overrides in Operations Manager 2007" topic in Operations Manager 2007 Help

Total CPU Utilization Percentage (Monitor)

CPU utilization is the percentage of elapsed time that the processor spends to run a non-idle thread. It is calculated by measuring the duration of the idle thread that is active in the sample interval and subtracting that time from interval duration. CPU utilization is the primary indicator of processor activity, and this monitor displays the average percentage of busy time observed during the sample interval.

CPU queue length is the current length of the system work queue for this CPU.

By default, the threshold for this monitor is a CPU utilization of 95 percent along with a CPU queue length greater than 15 measured once every 2 minutes using five samples to compute the threshold.

Total Processor % Interrupt Time (Collection Rule)

This rule collects the Total Instance of the % Interrupt Time performance counter. By default, a sample is taken every 5 minutes. % Interrupt Time monitors the overall average processor utilization that occurred in Interrupt mode. Only interrupt service routines (ISRs), which are device driver functions run in Interrupt mode. Excessive % Interrupt Time can identify that a device is malfunctioning and serves as a secondary indicator that a device might be contributing to a processor bottleneck.

Processor % Processor Time Total (Collection Rule)

This rule collects the Total Instance of the % Processor Time performance counter. By default, a sample is taken every 5 minutes. % Processor Time is the percentage of time when the processor is not running the idle thread and it is assumed that the processor is busy on behalf of real work. % Processor Time is the primary indicator of a processor bottleneck. You should be concerned of sustained periods of % Processor Time over 80 to 90 percent.

Total Processor % DPC Time (Collection Rule)

This rule collects the Total Instance of the % DPC Time performance counter. By default, a sample is taken every 5 minutes. % DPC Time monitors the percentage of time that the processor spent in routines known as deferred procedures calls, which are device driver scheduled routes which are called from ISRs. Excessive %DPC Time might be an indication of a hardware or device driver problem.

Monitoring Individual Processor Performance

The following monitors and rules are enabled when you enable Object Discoveries for processors.

Processor % Interrupt Time (Collection Rule)

This rule collects the Processor of the % Interrupt Time performance counter. By default, a sample is taken every 5 minutes. % Interrupt Time monitors the overall average processor utilization that occurred in Interrupt mode. Only interrupt service routines (ISRs), which are device driver functions run in Interrupt mode. Excessive % Interrupt Time can identify that a device is malfunctioning and serves as a secondary indicator that a device might be contributing to a processor bottleneck.

Processor % Processor Time (Collection Rule)

This rule collects the Processor of the % Processor Time performance counter. By default, a sample is taken every 5 minutes. % Processor Time is the percentage of time when the processor is not running the idle thread and it is assumed that the processor is busy on behalf of real work. % Processor Time is the primary indicator of a processor bottleneck. You should be concerned of sustained periods of % Processor Time over 80 to 90 percent.

Processor % DPC Time (Collection Rule)

This rule collects the Processor of the % DPC Time performance counter. By default, a sample is taken every 5 minutes. % DPC Time monitors the percentage of time that the processor spent in routines known as deferred procedures calls, which are device driver scheduled routes which are called from ISRs. Excessive %DPC Time might be an indication of a hardware or device driver problem.

CPU Utilization Percentage (Monitor)

CPU utilization is the percentage of elapsed time that the processor spends to run a non-idle thread. It is calculated by measuring the duration of the idle thread that is active in the sample interval and subtracting that time from interval duration. CPU utilization is the primary indicator of processor activity, and this monitor displays the average percentage of busy time observed during the sample interval.

CPU queue length is the current length of the server work queue for this CPU.

By default, the threshold for this monitor is CPU utilization of 95 percent measured once every 2 minutes using 5 samples to compute the threshold.