none
performance issue of OPSMGRDW

    Question

  • Hi

    I am able to see lot of data copying and immediately deleting from the Perf.PerformanceStage table of OPSMGRDW. Please let me know what data copying to this table. Every 1 second it is copying 2000 records and immediately deleting and causing performance issue.

    Regards

    Donald D'souza

     

    Monday, June 14, 2010 9:53 PM

Answers

  • This is part of the normal data flow of data into the Data warehouse.

    When a Management Server writes Performance Data to the Data Warehouse, the data is written into Perf.PerformanceStage.  Once there, there's a periodic job that calls the StandardDatasetMaintenance stored procedure to move the data from the Perf.PerformanceStage into the Perf.PerfRaw_GUID tables.

    This process should not be running every second though.  It will run frequently (every ~30 seconds or so) on every management server that you have.

    What sort of performance problems are you seeing?


    Michael Pearson
    OpsMgr Performance Test Team
    http://blogs.technet.com/michaelpearson/

    This posting is provided "AS IS" with no warranties, and confers no rights. Use of attachments are subject to the terms specified at http://www.microsoft.com/info/cpyright.htm
    Tuesday, June 15, 2010 1:02 AM

All replies

  • This is part of the normal data flow of data into the Data warehouse.

    When a Management Server writes Performance Data to the Data Warehouse, the data is written into Perf.PerformanceStage.  Once there, there's a periodic job that calls the StandardDatasetMaintenance stored procedure to move the data from the Perf.PerformanceStage into the Perf.PerfRaw_GUID tables.

    This process should not be running every second though.  It will run frequently (every ~30 seconds or so) on every management server that you have.

    What sort of performance problems are you seeing?


    Michael Pearson
    OpsMgr Performance Test Team
    http://blogs.technet.com/michaelpearson/

    This posting is provided "AS IS" with no warranties, and confers no rights. Use of attachments are subject to the terms specified at http://www.microsoft.com/info/cpyright.htm
    Tuesday, June 15, 2010 1:02 AM
  • Hi Michael

    We are getting complaints from SAN team SCOM DW is driving 250MB/s out of the storage array.

    Regards

    Donald D'souza

    Tuesday, June 15, 2010 9:14 AM
  • But that's the way it works.  Your SAN should be able to handle it :)
    Microsoft Corporation
    Tuesday, June 15, 2010 3:34 PM
  • Ok, 250MB/sec on the SAN.  How many minutes / seconds is that for?  Is that 250MB/sec read, write, or read+write?  What's the ratio of reads to writes?

    If its a sustained 250MB/sec constantly 24x7, that's probably a problem.  Its it 250MB/sec for a few seconds, that's a good thing.  Tell your SAN team they are doing a good job.

    If the SAN team is concerned that the OpsMgr DW is using too many shared resources, can they throttle the IOPS to the OpsMgr DW LUN's? 

     


    Michael Pearson
    OpsMgr Performance Test Team
    http://blogs.technet.com/michaelpearson/

    This posting is provided "AS IS" with no warranties, and confers no rights. Use of attachments are subject to the terms specified at http://www.microsoft.com/info/cpyright.htm
    Tuesday, June 15, 2010 6:31 PM
  • Hi

    We have applied below thing to get ride of the issue

    Adjust autogrowth settings for database OperationsManagerDW to 250 MB.

    SQL Max Server Memory setting to 12GB

    Applied CU10

    But the read/ write is normal during day time ( 250MB/sec read+write), but it increasing in night from 12:00 AM to 08:00 AM(3250/sec read+write).

    Anything you guys suggest?

    Regards

    Donald D'souza
    Friday, July 23, 2010 4:16 PM
  • At this point, I would contact support since there are lots of things to look at here.  The issue could be one or many of the following:

    1. It might not be the OpsMgrDW causing the IO.  There may be another database that's causing the IO.  Tracking that down via a forum like this isn't easy to do and its going to take a long time.
    2. There are one or many Rules collecting too much performance data to your Data Warehouse.  I think there's a good chance of this.
    3. Performance Data isn't the issue but maybe there are other data items which are contributing to the IO load.  Maybe Events?
    4. Is there another process that is reading data from this DW causing all of the IO?
    5. The DW Maintenance procedure has not been running and keeping all of the indexes and stats up to date.  This is also very likely.

    I suggest contacting CSS and let them help you with this. 
    http://support.microsoft.com
    Click on the "Contact Support Professional" link at the bottom of the page.


    Michael Pearson
    OpsMgr Performance Test Team
    http://blogs.technet.com/michaelpearson/

    This posting is provided "AS IS" with no warranties, and confers no rights. Use of attachments are subject to the terms specified at http://www.microsoft.com/info/cpyright.htm
    Friday, July 23, 2010 7:14 PM