July 17, 2009 at 6:52 am
Hi,
Attempted to install SP3 on our cluster last night, and had a few problems. Any help/suggestions greatly appreciated...
Our cluster has two-nodes (widows 2003 SP2 64 bit), running one instance per node (SQL 2005 Enterprise SP2), in an active-active configuration.
I moved both instances (PROD01 and PROD02) onto a single node (called SRCSILVER) and installed SP3. The installer correctly picked up both instances, updated them, and updated the other components (Integration Services etc).
I then installed SP3 on the other node (called SRCGOLD), which successfully updated non-database engine components (Integration services etc). I then rebooted SRCGOLD and when it came back online, I moved the resources over to SRCGOLD (both instances and cluster group). I then rebooted SRCSILVER. When SRCSILVER came back online, I moved the cluster group and PROD02 back onto it.
When I came to check the versions PROD01 (on SRCGOLD) was showing 9.0.3042 (SP2) and PROD02 (on SRCSILVER) was showing 9.0.4035 (SP3). I then moved PROD01 back onto SRCSILVER, to see if I could, and it showed version 9.0.4035.
As a next step I thought i’d try to run SP3 on SRCGOLD, with both instances running on that node. When I did, the setup failed on the database engine components. Here is the relevant snippet from the hotfix log.
Attempting to pause the 32 bit ngen queue
07/16/2009 20:05:06.034 Attempting to pause the 64 bit ngen queue
07/16/2009 20:05:06.345 Installing product: SQL9
07/16/2009 20:05:06.345 Installing instance: MSSQLSERVER
07/16/2009 20:05:06.345 Enumerating passive cluster nodes
07/16/2009 20:05:06.363 Patching available passive node: SRCSILVER
07/16/2009 20:05:06.363 Waiting for first completed passive node
07/16/2009 20:05:06.363 Patching passive node: SRCSILVER
07/16/2009 20:05:06.655 Task Scheduler: Created remote folder for product instance target \\SRCSILVER
07/16/2009 20:05:14.632 Task Scheduler: Transferred payload to remote product instance target \\SRCSILVER
07/16/2009 20:05:14.632 Task Scheduler: Transferred payload to remote product instance target \\SRCSILVER
07/16/2009 20:05:14.668 Task Scheduler: Created new scheduled task for product instance target \\SRCSILVER
07/16/2009 20:05:14.668 Task Scheduler: SaveTaskToDisk(): Got task interface
07/16/2009 20:05:14.705 Task Scheduler: Saved new scheduled task for product instance target \\SRCSILVER
07/16/2009 20:05:14.705 Task Scheduler: Created scheduled task for product instance target \\SRCSILVER
07/16/2009 20:05:14.705 Task Scheduler: Started scheduled task for product instance target \\SRCSILVER
07/16/2009 20:05:14.705 Task Scheduler: Started scheduled task for product instance target \\SRCSILVER
07/16/2009 20:05:18.209 Task Scheduler: Scheduled task for product instance target has completed
07/16/2009 20:05:18.209 Task Scheduler: Waiting for exit code from scheduled task for product instance target \\SRCSILVER
07/16/2009 20:05:24.050 Task Scheduler: Received exit code 11006 from scheduled task for product instance target \\SRCSILVER
07/16/2009 20:05:24.050 Task Scheduler: Result code for scheduled task for product instance target has been received
07/16/2009 20:05:24.050 Task Scheduler: Removed scheduled task for product instance target \\SRCSILVER
07/16/2009 20:05:24.178 Task Scheduler: Removed remote folder for product instance target \\SRCSILVER
07/16/2009 20:05:24.178 Remote process completed for product instance target
07/16/2009 20:05:24.178 Exit code for passive node: SRCSILVER = 11006
07/16/2009 20:05:29.727 The following exception occurred: No passive nodes were successfully patched Date: 07/16/2009 20:05:29.727 File: \depot\sqlvault\stable\setupmainl1\setup\sqlse\sqlsedll\instance.cpp Line: 3510
07/16/2009 20:05:30.110 Watson: Param1 = Unknown
07/16/2009 20:05:30.110 Watson: Param2 = 0x2b01
07/16/2009 20:05:30.110 Watson: Param3 = Unknown
07/16/2009 20:05:30.110 Watson: Param4 = 0x0
07/16/2009 20:05:30.110 Watson: Param5 = instance.cpp@3510
07/16/2009 20:05:30.110 Watson: Param6 = Unknown
07/16/2009 20:05:30.110 Watson: Param7 = SQL9
07/16/2009 20:05:30.110 Watson: Param8 = Hotfix@
07/16/2009 20:05:30.110 Watson: Param9 = x64
07/16/2009 20:05:30.110 Watson: Param10 = 4035
07/16/2009 20:05:30.110 Installed product: SQL9
07/16/2009 20:05:30.110 Hotfix package completed
07/16/2009 20:05:30.110 Attempting to continue the 32 bit ngen queue
I’ve ensured there are no remote connections to either node (apart from the console connection on SRCGOLD), as I have read this can cause issues. Scheduled tasks are enabled on both nodes and the service account (that is used for both nodes) and myself have permissions to create them.
Was wondering if anyone had experienced similar issues when installing SP3 (or service packs/cumulative updates in general) and been able to fix it? SP2 did go on ok – but that was installed one node/instance at a time.
Thanks,
Matt
July 17, 2009 at 8:07 am
Just a shot-in-the-darh here but I've had problems on clusters if I'm logged on to both nodes at the same time. I think you may need to install on your SRCGOLD side only after you've logged off the SRCSILVER side.
Also; The obvious - make sure your at least local admin on both sides, Make sure the Task Scheduler service is running.
Lastly, I've found it helps to be logged on as "console".
Good luck, let us know what happens.
Tim White
July 17, 2009 at 8:29 am
Thanks for the swift reply Tim 🙂
Checked the above. Ran as admin on console. Found this which might help...
http://www.mssqltips.com/tip.asp?tip=1668
... will post results when attempt it.
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply