Performance Counters and Threshold for Exchange Server

We'll continue Exchange Monitoring and now delve into one of the most important topics of Performance Monitoring. In this post we'll discuss about the various Performance Counters and their Thresholds that should be kept in mind when performing a Cumulative HealthCheck.


Please Note: Some changes might be noticed with different versions of Exchange Server

Type	Object\Counter	Threshold	Action
Memory	Memory\Available Bytes	Less than 4MB	Check for memory leaks and add RAM if necessary
Memory	Memory\Pages/ sec	20	Investigate paging settings
Paging File	Paging File\% Usage	Above 70%	The Paging File value should match up with the previous two values. A value exceeding 70% is not healthy for the system.
Disk	Physical Disk\ Free Space Logical Disk\ Free Space	15%	Clear more disk space. Increase logical or physical disk space
Disk	Physical Disk\ Disk Time Logical Disk\ Disk Time	90%	The disk is not being read quickly enough. This could be a hardware issue. It could also be that the amount of data on disk is too large.
Disk	Physical Disk\Disk Reads/sec, Physical Disk\Disk Writes/sec	Depends on Manufacturer	The writing the reading from the disk is slow. We may need to upgrade the disk or disk drivers
Processor	Processor\ % Processor Time	85%	Find the process intensive processes and move them to separate processors (if you have a multi-processor machine) or add more processing power (by adding another processor or by upgrading to a faster processor)
Processor	Processor\ Interrupts/ sec	Processor specific; 1000 is a standard	If the counter value increases without additional processes, the cause could be hardware related
Server	Server\ Bytes Total/ sec	Depends on the network	If all your servers’ Bytes Total/ sec is the same and similar to the maximum network speed, you might need to increase the network bandwidth
Server	Server\ Pool Paged Peak	Physical RAM available	Should not be greater than the physical RAM value.

Performance Object	Performance Counter	Details
MSExchangeIS	VM Largest Block Size	A healthy server has more than 200,000,000 bytes (200 MB) as the largest free block. If the value is lower, carefully monitor the server.
Memory	Pool Page Bytes	When the /3GB switch is used, amounts larger than 200 MB indicate a problem except when backups are running. During backups, each page in the cache manager is copied into the pool page, which causes an increase in pool page size.
Memory	Pool Nonpaged Bytes	When the /3GB switch is used, amounts larger than 100 MB indicate a problem.
Memory	Free System Page Table Entries	Amounts less than 3000 indicate a problem.
Process	Private Bytes	An upward trend indicates a potential memory leak.

Every Checklist thats circulated among team members should have these Thresholds mentioned clearly for reference.