SQL Server Performance

Problem with clusters

Discussion in 'SQL Server 2005 Clustering' started by satya.sqldba, Dec 7, 2006.

  1. satya.sqldba New Member

    Hello everyone

    We have 4 production servers that are clustered, all Active/Passive. We ahave been having a strange problem from few days. There is one particular table thats troubling the whole server,when any operations(even select) are done on it, failover is occuring. We have tried to rebuils the database, but the same thing is happening once again. What could be the trouble? Any solutions?

  2. satya.sqldba New Member

    The trouble was particularly happening because of one database to be precise. Any backup, restore, running triggers, anything of that sort the cluster is failing over back and forth.

  3. MohammedU New Member

    First check the SQL error logs and Event logs as well as cluster log for any errors...
    Run the check table against problematic table after restoring the db on to any other server for corruption...

    What happens if you run..."Select top 1 * from problemtable"...
    Does it cause failover?

    Try diabling the parallalism and see what happens...
    In sql 7 parallalism caused one of cluster server to failover many times...

    Mohammed U.
  4. satya Moderator

    You need to be more clear by looking at Clustering logs and event viewer logs.

    Satya SKJ
    Microsoft SQL Server MVP
    Contributing Editor & Forums Moderator
    This posting is provided AS IS with no rights for the sake of knowledge sharing.
  5. sachinsamuel New Member

    Also check if there is any hardware problem to the server. It may be a case that the primary server is failing back because them
  6. sumit_rec New Member

    Hi Satya

    Check the following ::

    1> ErrorLogs for any Access Violations, Assertions and/or any stack dumps that are coming up
    2> Application and System Event Logs on Both Node 1 & 2 for any hardware errors.
    3> The Cluster.log for any errors on the Resources
    4> Check the LOG Directory for any SQL Server Dumps (*.dmp and *.mdmp)

    If you do see the above like any Access Violations and/or any dumps contact MS Support. MS Support can find if there is any fix to those dumps.

    Check for Hardware errors as well.

    Also does the Table in Question reference any Trigers which call Extended Stored Procs or any COM objects.

    Another important things check for Dependencies on the Cluster as well.

    Check for any
    1> 3rd Party DLLs loaded on SQL Server Process Space (You can find this by using Process Explorer on Windows systems)
    2> Any Filter Drivers which needs to be updated. Use Fltmc command to find out.

    In short Download MPS Reports from Microsoft support Site and check the above files on the cabinet files.

    Sumit Sarkar

    Sumit sarkar
  7. satya.sqldba New Member

    Thanks Sumit

    There was an OS config problem. Issue resolved. The OS team took care of it

  8. MohammedU New Member

    Satya,<br /><br />If possible...Can you please post some more details about your solution to your problem<img src='/community/emoticons/emotion-1.gif' alt=':)' /><br /><br />MohammedU.<br />Moderator<br />SQL-Server-Performance.com
  9. satya.sqldba New Member


    I am a begginer in clustering and haven't setup the clusters by myself. I had very little info in a very brief email from the OS team that the problem got solved and mentioned that it was due to config problem.

  10. MohammedU New Member

    Not a problem... take it easy...


Share This Page