Results 1 to 2 of 2
  1. #1
    Join Date
    Nov 2007

    Unanswered: Clustered sql Server 2005 restarted suddenly

    Hi EveryOne,

    WIndows 2003 sp2 and sqlserver 2005 standard edition -sp2.

    Found following erors in The cluster and event logs.

    error: [sqsrvres] CheckQueryProcessorAlive: sqlexecdirect

    error: [sqsrvres] OnlineThread: QP is not online.

    error: [sqsrvres] printODBCError: sqlstate = HYT00;
    native error = 0; message = [Microsoft][ODBC SQL Server
    Driver]Timeout expired

    [sqsrvres] printODBCError: sqlstate = 08S01; native error = 0; message = [Microsoft][SQL Native Client]Communication link failure

    [sqagtres] SvcStop: service did not stop; giving up.
    [sqsrvres] OnlineThread: asked to terminate while waiting for QP.

    SQL Server is terminating in response to a 'stop' request from Service Control Manager

    Earlier we had sudden Sql agent restart but this time sql server services were restarted .

    What could be he reasons for this.

    This machine has 64 GB ram and sql server has been alloted 30 Gb.
    I found 30 Gb Usage in the dump file of Sqlserver

    Can a memory Starved sql server could be a rason why sql server would have restarted


  2. #2
    Join Date
    Jun 2009
    I know this is an old thread but just wondering if anyone has any ideas on this subject? I'm having almost the same issue (only once so far - it just started). I get all of the above errors but haven't had the service stop or had a cluster failover yet. I'm concerned that if the problem gets worse we will have failovers.

    I've done some reading and I've got a few theories:
    1. Cluster networking looks like it could be tweaked. Both the private and public connections are setup for internal and all communications. I'm wondering if this could be causing issues with the heartbeat in certain circumstances??? (Although it's worked fine for over 2 years)
    2. We upgraded our ram from 16GB to 32GB a few weeks ago. I read someone who claims that it's possible to cache too much data in memory. In some cases if major updates are made to tables in memory then when it comes time to commit those changes to disk the IO subsystem can be saturated and become unresponsive???

    I'm still digging but I see no other issues on this server. There doesn't appear to be any memory pressure and there are no other problems in any of the logs etc.

    The only other difference to the above post is that my server is a SQL2000 Enterprise cluster - not SQL2005.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts