5 Replies Latest reply on Jun 17, 2009 10:17 AM by bmg4357

    SQL Server Performance Orion NPM 9.1 SP4 APM 2.0 SP2

    bmg4357

      I have migrated my Orion database twice in an effort to get better performance and I am still fighting with issues.  I now have the database on a quad proc 3.2 GHz W2K3 X64physical server, 4GB RAM, RAID 5.  I am monitoring 1504 elements with NPM and 1321 components with APM.

      The problem I am having is two fold.  Fist of all, the Disk Queue length on the SQL 2005 X64 SP2 back-end server is running consistently between 5 and 9. 

      The second problem that I am having is the biggie.  The virtual memory utilization will hit about 4.5GB after a couple of days.  The CPU then spikes to 100% consistently.  APM and NPM stop logging and the web interface goes down.  If I cycle the Orion services on the application server, the problem comes right back again and the virtual memory utilization on the SQL box never decreases.  If I try to restart the SQL processes, the MSSQL process hangs in the stopping state.  I then have to either use Process Explorer to kill the process and then restart it or I have to reboot the box.  I'm then ok for another two days.  I could schedlue a nightly reboot of the SQL box, but that would only cover up the problem and not resolve the underlying issue.

      This is driving me crazy.  Support is of no help on this, and now I can't even get them to reply to my emails.  I'm very frustrated at this situation.  Any help would be appreciated.