Applies to: Exchange Server 2007 SP1, Exchange Server 2007
Topic Last Modified: 2007-09-26

This topic provides information about performance counters and suggestions about how to improve processor performance. Generally, identifying that a server is processor-bound is straightforward. Use the performance monitor counters listed in the following table to determine whether there are any contentions on the processors.

Performance counters for processors

Counter Expected values

Processor\% Processor Time (_Total)

Indicates the percentage of time the processor is running non-idle threads.

You can use this counter to monitor the overall utilization of the processors or the per-processor utilization.

  • The average CPU utilization should be below 90 percent at all times.

System\Processor Queue Length

Indicates the number of threads in the processor queue.

There is a single queue for processor time, even on computers with multiple processors. This counter shows ready threads only, not threads that are currently running.

  • This counter should be less than 2.

  • Infrequent spikes that do not correlate to the % Processor Time is above 75 percent are not an issue.

Example of a Processor-Bound Problem

The following figure shows a problem due to excessive processor utilization that a server running Microsoft Exchange is experiencing. The CPU utilization is approximately 90 percent (as indicated by the Processor\% Processor Time counter), and the System\Processor Queue Length counter is consistently high.

Monitor processor using Performance snap-in

To make sure that Exchange is responsible for this CPU utilization, use the Performance snap-in to view the Processor\% Processor Time counter for all processes. Evaluate these processes to determine if an Exchange process is the major cause for the CPU utilization.

Monitor % Processor Time using Performance snap-in

Improving Processor Performance

You can improve processor performance by taking the following actions:

  • Make sure that database maintenance and backups occur during off-peak hours and are staged   To reduce the overall impact on the server, make sure that input/output (I/O)-intensive, CPU-intensive, or memory-consuming tasks (such as backup and maintenance) occur outside normal operating hours. You can further lessen the effect of these resource-intensive tasks by staging (setting different start and preferably end times for tasks) the maintenance and backup of databases or storage groups.

  • Offload tasks to other servers   Many tasks that an Exchange server performs can be set to occur only on dedicated servers. For example, you can deploy dedicated public folder servers if your organization relies heavily on them.

For More Information

For more information about resolving performance issues related to Exchange 2007, see Performance Issues.

Monitoring the % Processor Time counter using the Performance snap-in
Monitoring a processor using the Performance snap-in