Start a new topic
Answered

Can EventSentry monitor Disk I/O utilization?

I can't find it in the feature information.


Best Answer

Yes, EventSentry monitors the disk utilization by default, this is available in the "Performance System" system health monitoring package. EventSentry monitors Windows performance through the performance counters made available by Windows.


Windows provides a number of performance counters to monitor disk performance, including:

  • % Disk Time
  • Avg. Disk Bytes / Transfer
  • Avg. Disk Queue Length (enabled by default)
  • Avg. Disk sec/Transfer
  • Disk Transfers/sec
  • Split IO/Sec
  • Disk Bytes/sec


You can monitor all those counters with EventSentry in any interval, graph the results and setup alerts as well. Our demo shows some examples of performance metrics and how you can access reports: http://demo.eventsentry.com/performancetrends.


Hope this helps.


Answer

Yes, EventSentry monitors the disk utilization by default, this is available in the "Performance System" system health monitoring package. EventSentry monitors Windows performance through the performance counters made available by Windows.


Windows provides a number of performance counters to monitor disk performance, including:

  • % Disk Time
  • Avg. Disk Bytes / Transfer
  • Avg. Disk Queue Length (enabled by default)
  • Avg. Disk sec/Transfer
  • Disk Transfers/sec
  • Split IO/Sec
  • Disk Bytes/sec


You can monitor all those counters with EventSentry in any interval, graph the results and setup alerts as well. Our demo shows some examples of performance metrics and how you can access reports: http://demo.eventsentry.com/performancetrends.


Hope this helps.

Hi Ingmar,


I am facing this trigger alert often

The performance counter "Performance System\Disk Queue" ("PhysicalDisk(*)\Avg. Disk Queue Length" instance "0 C:") on host < > exceeded the threshold of 20 the current values are: Average: 21Minimum: 0Maximum: 453 Top 3 instances: _Total (21)0 C: (21)2 F: (0) Counter Description:Avg. Disk Queue Length is the average number of both read and write requests that were queued for the selected disk during the sample interval. To adjust this threshold or exclude this alert: 1. Open the management console 2. Navigate to Packages -> System Health 3. Locate the package "Performance System" expand it and click Performance / SNMP 4. Locate the object "Disk Queue" and double-click it 5. Click the "Alert" tab to adjust the threshold or clear the event log alert check box 6. Save or Save and Deploy the configuration




Disk Queue Interval: 5 Alert >20 Time Period 10 mins Database 5 mins

Is there any recommendable values to be set on trigger ? or Shall we disable and neglect this ?

Hi,


There are a few things you can do to suppress this alert/error. I would recommend to first tweak the threshold of this counter so you get fewer alerts. This performance counter is still valuable since it informs you of high disk activity on a system that could indicate a problem. However, the default values may be a little bit too sensitive.


I would suggest you first increase the threshold from 20 to 50 and also increase the time period from 10 minutes to 30 minutes. This should definitely reduce the numbers of alerts generated. Of course you can tweak it even further, e.g. increase the time period to one hour etc.


If that doesn't help, then you can always delete the counter as well, or simply exclude the actual alert - we have a screen cast on excluding alerts here: https://youtu.be/kk3AHhP5zKo.



I hope this helps,

Ingmar.


1 person likes this

Hi,


Thanks for the quick response. It is helpful.


Regards,

Sriganth

Login or Signup to post a comment