none
What is the Performance Metric Provider timer job? RRS feed

  • Question

  • Hi, Every day at certain times of day, the Performance Metric Provider failed with:

    An update conflict has occurred, and you must re-try this action. The object SPWebService was updated by OurDomain\OurSPAdmin, in the OWSTIMER (3304) process, on machine OURSERVER. View the tracing log for more information about the conflict.

    I can't find any information about this specific timer job. What does it do? How does it work? How we can make benefit out of it?

    Does anyone understand this timer job?

    Thank you.

    Monday, August 18, 2014 11:50 PM

All replies

  • Thank you Mike for your reply.

    However, the article links you provided talks about Database Performance Metric Provider, not Performance Metric Provider. These are two different timer jobs.

    I'm getting error from Performance Metric Provider, and no errors from Database Performance Metric Provider.

    Regards

    Tuesday, August 19, 2014 1:52 AM
  • Hi Conway,

    As I understand, you encountered event id 6398 several times a day. Please refer to the link below and clear SharePoint configuration cache:

    http://support.microsoft.com/kb/939308/en-us

    Similar issue:

    http://sharepointcherie.blogspot.com/2013/06/sharepoint-2013-event-id-6398-update.html 

    Regards,

    Please remember to mark the replies as answers if they help and unmark them if they provide no help. If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com .


    Rebecca Tu
    TechNet Community Support

    Wednesday, August 20, 2014 2:14 AM
  • Hi Rebecca,

    Thank you for your reply. Unfortunately you are not answering my question as well.

    I'm not looking for a fix to the error message, as it only happens twice a day. And I do know about the cache-clearing procedure already, which made no difference to the error.

    The purpose of this thread is to find out information about the Performance Metric Provider timer job. What does it do? How does it work? How we can make benefit out of it?

    Thanks.

    Wednesday, August 20, 2014 2:39 AM
  • Hi Conway,

    There is hardly find article specific for Performance Metric Provider timer job. However, we know it as a SharePoint timer job which collects performance metrics data per: http://technet.microsoft.com/en-us/library/cc678870(v=office.15).aspx

    For performance metrics, you could find related information in: http://technet.microsoft.com/en-us/library/gg593613(v=office.15).aspx

    Thanks for the understanding.

    Regards,


    Rebecca Tu
    TechNet Community Support


    • Edited by Rebecca Tu Thursday, August 21, 2014 9:10 AM
    Thursday, August 21, 2014 9:09 AM
  • Thank you again Rebecca.

    Yes there isn't any informative articles about this timer job, which is why I've come here to open this thread.

    Now it looks like users are not going to benefit directly from this timer job.

    It'd be good to at least understand things like...if this timer job can be turned off, or change its schedule without having bad impact on the farm, or how to troubleshoot it when it fails...

    But thank you for spending the time. I'll wait a bit longer to see if there are any more people contributing to this thread.

    Regards

    Thursday, August 21, 2014 9:42 PM
  • I had the EXACT same question go forward to MS Support.

    Here is the reply:

    The information about the "performance metric provider" timer job is very slim. What I’ve been able to determine is that it’s a part of Health Data Collection reporting. SharePoint uses this timer job to collect snapshots of the performance every minute and compile that into the overall health. The occasional conflict does not impact the health data collection. We all get these events, even in a 2-3 server farm. You’ll have them more frequently but they can still be ignored.

    Tuesday, November 17, 2015 2:49 PM
  • That sounds like it's part of the Request Management system used to assess the health of the SharePoint servers so that the load balancing process can work. That matches with the description MS provide and with the frequency of the timer job.

    That means that a couple of failures don't mean much in themselves, it'll correct after a minute. On the down side a failure may mean that the health report doesn't get updated for a minute which means you may get poor routing of requests until the job re-runs. Finally the question of when the job fails is of interest, logically it'll probably fail more when the farm is heavily loaded, which is the exact time it needs to be working

    It also means that it shouldn't be disabled as by doing so you'd cripple the load balancing, at best it will just have to keep using the latest, out of date, health report for each server, at worst it'll curl up and die.

    Thanks for the info Cory.

    Tuesday, November 17, 2015 3:08 PM