A bottleneck from a disk can significantly impact response time for applications running on your system. So we are going to see what are the most commonly seen disk bottlenecks and how they can be monitored and what are the tolerable values for them.

What are performance counters?

Before selecting the objects and counters, we need to define objects and counters in Windows Performance Monitor (PERFMON). A performance object is an entity for which performance data is available and performance counters define the type of data that are available for a performance object. One performance object will have one or more associated performance counters. When added in the Performance Monitor, they will appear in the form of \Performance Object\Performance Counter. For example:  in \PhysicalDisk\% DiskTime, physical disk is the performance object and % Disk Time is the performance counters.

From disk per se, there are two performance objects which we typically monitor:

1. Physical Disk performance object is one that monitors disk drives on the computer. It identifies the instances representing the physical hardware, and the counters are the sum of the access to all partitions on the physical instance.

2. Logical Disk Performance object monitors logical partitions. Performance monitor identifies logical disks by their drive letter or mount point.

Performance Counters for monitoring Disk

• Physical Disk (instance)\Disk Transfers/sec counter for each physical disk and if it goes above 25 disk I/Os per second then you've got poor response time for your disk.
• Physical Disk(instance)\% Idle Time – It is a measure the percent time that your hard disk is idle during the measurement interval, and if you see this counter fall below 20% then you've likely got read/write requests queuing up for your disk which is unable to service these requests in a timely fashion. In this case it's time to upgrade your hardware to use faster disks or scale out your application to better handle the load.
• PhysicalDisk\% Disk Time: Shows the percentages of time the disks are busy servicing read or write requests. The recommended average should be less than 85%.
• PhysicalDisk\Avg. Disk Queue Length: Shows the number of disk requests that are waiting for a disk during the sample interval. The recommended average should be less than 2 per disk.
• PhysicalDisk\Current Disk Queue Length: Shows the number of disk requests that are currently waiting at the time the performance data is collected. The recommended average should be less than 2 per disk.
• PhysicalDisk\Disk Transfers/sec: Shows the number of read and writes per second regardless of the amount of data. The recommended average should be less than 100 per disk. 
• PhysicalDisk\Disk Bytes/sec: Shows the amount of data transmitted per disk per second.
• PhysicalDisk\Avg. Disk Sec/Read: Shows the average in milliseconds (ms) to read from disk. The recommended average should be less than 10 milliseconds (ms).
• PhysicalDisk\Avg. Disk Sec/Write: Shows the average in milliseconds (ms) to write to disk. The recommended average should be less than 10 milliseconds (ms).

 What can be done?

• Increase higher memory to reduce paging to the disks.
• Get faster hard disks to your server.
• Consider using SAN and RAID array if not already done so.
• Create multiple files and filegroups in separate disk drives.
• Relocate the log files to separate disks from the data files.