locked
Need alerting on Page File % usage RRS feed

  • Question

  • Hello-

    We are using SCOM 2012R2 for our environment- this is exactly what I need to accomplish

    1. Monitor File Page Percentage used

    2. Establish a warning and critical state, lets say warning at 90% usage and Critical at 95% on only the page file

    3. We have email alerting and is running perfectly- numerous tests

    Does SCOM have the ability to do this out of the box??? If so could you walk through it for me

    Thank you for your time

    GunnySFA

    Tuesday, October 7, 2014 4:09 PM

Answers

  • We got a similar monitor to run using almost the exact same as the above but with a couple of differences.

    1) Use a "Consecutive Samples over Threshold" monitor instead of the suggested "Simple Threshold" monitor.

    This will allow you to set a number of consecutive samples it has to be spiked at before generating an alert. Otherwise you run the risk of getting alerts simply because a server happened to have the Page File at over 80% for the 5 minute interval your monitor is set for once during the day, then it calmed down again. A SQL DB can cause that during a brief 10 minute period or so. The Consecutive Samples monitor, however, will force the server monitored to have to have been at 80% Page File usage, every 10 minutes (interval) for over 4 consecutive samples (40 minutes) straight before sending out alert. We find that more useful. We dont care that server AA was at 90% page file usage today for 15 minutes during a mid-day backup. However, we do care that server AA was at 90% page file usage today for well over an hour straight (10 minute interval, 6 samples) and it hasnt cleared back to a healthy state yet. 

    2) We only wanted a pure total % utilization alert for the Page File so our monitor is not set for all instances. We left the all instances check box cleared and set the instance as [ _Total ] (without the brackets). This is a direct correlation to PerfMon where the Paging File is being monitored for % Usage counter for the _Total instance. 

    • Proposed as answer by Blake Mengotto Wednesday, October 8, 2014 3:58 PM
    • Marked as answer by Yan Li_ Friday, October 17, 2014 6:54 AM
    Wednesday, October 8, 2014 2:22 PM
  • Hi Gunny,

    SCOM Can do more than that. Yes it is possible.

    You need to create 2 unit monitors for that. 1 for Warning 90% usage and the second for Critical at 95%.

    Create a unit monitor with the below performance counter for Page file usage in %

    Monitor target : Windows Server Operating System (Will alert if this appears on all the agents running windows server o/s)

    Select the items as below:

    Object: Paging File
    Counter: % Usage
    Instance: Check the box below as all instances are needed.

    Click next and mention the threshold 90%

    And select the alerting type as per the screen shots.

    Last Check enable the alert and the severity (Warning or critical) 

    Create the same way for another Critical for 95 %

    Apologize for a mixed screenshot. Microsoft allows only 2 photos per answer. 

    If screen shot is not visible open it in a new window to enlarge.

    Let me know if any doubts.


    Gautam.75801

    Tuesday, October 7, 2014 7:21 PM

All replies

  • Hi Gunny,

    SCOM Can do more than that. Yes it is possible.

    You need to create 2 unit monitors for that. 1 for Warning 90% usage and the second for Critical at 95%.

    Create a unit monitor with the below performance counter for Page file usage in %

    Monitor target : Windows Server Operating System (Will alert if this appears on all the agents running windows server o/s)

    Select the items as below:

    Object: Paging File
    Counter: % Usage
    Instance: Check the box below as all instances are needed.

    Click next and mention the threshold 90%

    And select the alerting type as per the screen shots.

    Last Check enable the alert and the severity (Warning or critical) 

    Create the same way for another Critical for 95 %

    Apologize for a mixed screenshot. Microsoft allows only 2 photos per answer. 

    If screen shot is not visible open it in a new window to enlarge.

    Let me know if any doubts.


    Gautam.75801

    Tuesday, October 7, 2014 7:21 PM
  • We got a similar monitor to run using almost the exact same as the above but with a couple of differences.

    1) Use a "Consecutive Samples over Threshold" monitor instead of the suggested "Simple Threshold" monitor.

    This will allow you to set a number of consecutive samples it has to be spiked at before generating an alert. Otherwise you run the risk of getting alerts simply because a server happened to have the Page File at over 80% for the 5 minute interval your monitor is set for once during the day, then it calmed down again. A SQL DB can cause that during a brief 10 minute period or so. The Consecutive Samples monitor, however, will force the server monitored to have to have been at 80% Page File usage, every 10 minutes (interval) for over 4 consecutive samples (40 minutes) straight before sending out alert. We find that more useful. We dont care that server AA was at 90% page file usage today for 15 minutes during a mid-day backup. However, we do care that server AA was at 90% page file usage today for well over an hour straight (10 minute interval, 6 samples) and it hasnt cleared back to a healthy state yet. 

    2) We only wanted a pure total % utilization alert for the Page File so our monitor is not set for all instances. We left the all instances check box cleared and set the instance as [ _Total ] (without the brackets). This is a direct correlation to PerfMon where the Paging File is being monitored for % Usage counter for the _Total instance. 

    • Proposed as answer by Blake Mengotto Wednesday, October 8, 2014 3:58 PM
    • Marked as answer by Yan Li_ Friday, October 17, 2014 6:54 AM
    Wednesday, October 8, 2014 2:22 PM