none
SQL 2008R2 instance taking long time to fail-over after SQL 2012 Upgrade.

    Question

  • Hi,

    I've a 2-Node cluster of Windows 2008 R2 Ent where we are having multiple SQL instances running. I've upgraded an instance to SQL Server 2012 from SQL Server 2008 R2. Ever since failover of other instances running SQL Server 2008R2 is taking long time to bring SQL Services online in the cluster. And it fails after 3 minutes (Pending Timeout default setting).

    I've verified the SQL instance during the state change (from offline to online), I found out that SQL Server is available that time just cluster is not able to change the state and it time out. There is no information in cluster logs at that point of time.

    Have someone faced the similar issue and any pointer to check it further? Can it be related to cluster resource DLLs which were updated while upgrading to SQL 2012?

    Looking forward to your cooperation.

    Regards,

    Divesh Mathur

    Cluster check during SQL upgrade


    Tuesday, November 26, 2013 12:20 PM

All replies

  • Hello,

    Please review the SQL Server error log from the node where it fails to come online. The Sql Server Error log may give you the cause if the sql server service itself is not able to start.
    If the sql server was able to start without any errors, please also check eventviewer and cluster logs for further analysis.

    Reference:View Events and Logs for a Failover Cluster

    Regards,
    Fanny Liu


    Fanny Liu
    TechNet Community Support

    Wednesday, November 27, 2013 7:37 AM
    Moderator
  • Thanks Fanny for the reply!

    SQL Server is able to start, see below logs.

    2013-11-26 10:31:58.19 spid8s      Recovery is complete. This is an informational message only. No user action is required.
    2013-11-26 10:31:58.19 Server      SQL Server is now ready for client connections. This is an informational message; no user action is required.
    2013-11-26 10:34:27.36 spid53      Attempting to load library 'xpstar.dll' into memory. This is an informational message only. No user action is required.
    2013-11-26 10:34:27.36 spid53      Using 'xpstar.dll' version '2009.100.1600' to execute extended stored procedure 'xp_instance_regread'. This is an informational message only; no user action is required.

    Also as mentioned earlier, there is no logs in the cluster logs during that period. Also in the event viewer the only log I found is below...

    Log Name:      System
    Source:        Microsoft-Windows-FailoverClustering
    Date:          11/26/2013 10:35:55 AM
    Event ID:      1069
    Task Category: Resource Control Manager
    Level:         Error
    Keywords:     
    User:          SYSTEM
    Computer:      <>
    Description:
    Cluster resource 'SQL Server ()' in clustered service or application <>failed.

    Regards,

    Divesh Mathur

    Wednesday, November 27, 2013 8:13 AM
  • Go through the below link it may helps you

    http://support.microsoft.com/kb/883732/en-us

    Wednesday, November 27, 2013 8:52 AM
  • I've verified the registry settings and they are similar on both nodes :-(
    Wednesday, November 27, 2013 9:34 AM
  • Check DNS server that cluster name should be there if  It is not there for any reasons(what I am suspected) then create a new with the same IP and cluster name.

    take off line both nodes one by one and failover.

    Wednesday, November 27, 2013 10:51 AM
  • Hi Pradeep,

    DNS record exists for the SQL Network Name and IP is assigned to it. Also during the failover SQL Network name, IP and All disks come online as usual...only the SQL Service resource is not able to change state in given time and timeout.

    Regards,

    Divesh Mathur

    Wednesday, November 27, 2013 11:16 AM