This discussion has been locked. The information referenced herein may be inaccurate due to age, software updates, or external references.
You can no longer post new replies to this discussion. If you have a similar question you can start a new discussion in this forum.

Some metrics for 1 SQL Server monitored instance are not working

FormerMember
FormerMember

 

Some metrics for only 1 monitored SQL Server instance are not working while the same metrics for other SQL instances are working just fine.

1. The metrics are enabled by default, but sometimes it will become disabled because of that error msg: "SQL Server performance data not being populated in master..sysperfinfo."

2. I already cleared the chart cach (few times).

3. Already restarted the Ignite PI service on the repo instance.

4. WMI counters are enabled (from Regedit.exe I checked).

5. WMI service on the monitored instance has been restarted already.

6. SQL Server service account on that server is the Local System built-in account.

Please find attached, it might help you to help me solve this issue.

Thanx

Hany

attachments.zip
  • FormerMember
    0 FormerMember

    Hi hany helmy,

    1. The issues with the sysperinfo table are concerning but I'm curious if these problems stem from outside the tool, can you do a SELECT * FROM SYS.SYSPERINFO? What about as the DPA / Ignite monitoring user?

    I shot you an email from our ticketing system with other troubleshooting as well for the rest of the points listed. Once we have finished working the ticket, I'll post a solution on this forum for future googlers who stumble across the same errors in an attempt to self resolve.

    - Jon Shaulis

  • FormerMember
    0 FormerMember in reply to FormerMember

    Thanks for your feedback, the results of the master..sysperfinfo is NULL (0 rows) using the monitoring user account (sysadmin) for that particular server only while 4 other servers, it`s retrieving the results as it should be. So I guess it`s a permission issue but I cannot resolve it myself as I am not fully aware of the software requirements, so plz need ur assistance through web session to close this tkt.


    Thanx

    Hany 

  • FormerMember
    0 FormerMember in reply to FormerMember

    Hi hany helmy,

    We've got a meeting scheduled tomorrow to resolve the issue, I'll circle back here to log the solution. As a note on the permissions, the monitoring user needs SYSADMIN to correctly monitor and cannot be "sa". We'll likely do an update connection as that typically re-grants or fixes any permissions issues, although there was more than just permission problems occurring in your instance. (Heavy WMI load issues, we will work on those too.)

  • FormerMember
    0 FormerMember in reply to FormerMember

    For that particular instance, it`s working on a different domain than the repo and all other monitored instances, and the monitoring user is a domain account which is already a sysadmin.

  • FormerMember
    0 FormerMember in reply to FormerMember

    Hi hany helmy,

    I just wanted to post the solution here for anyone who googles and finds the similar search result. What we found is that the DMV's we pull the metrics from are blank and contain no information. We are an agentless tool and rely on queries to pull the data. So our solution here is to refer back to the original configuration of SQL Server or engage Microsoft services to find why the system tables contain no information / data as without it, we won't be able to pull that information to create our metrics.