Post reply

Service Pack Strategy

  • February 27, 2012 at 7:07 am

    #252197

    Hi all

    We need to service pack one of our SQL two node clusters, our boss is asking if this can be done in ours. At present I am thinking about service packing the passive node, doing a failover at the night, then service packing the other node after a days running.

    Obviously we will have a day then where one node is Service Packed and the other isn't. What do people think about that strategy? I can see an advantage in easy roll back should anything be incorrect, however I am worried about having the cluster at two different versions.

    Thanks

  • February 27, 2012 at 7:44 am

    #1452071

    Here is a KB that will provide some guidance for your rolling patch process.

    ______________________________________________________________________________________________
    Forum posting etiquette.[/url] Get your answers faster.

  • February 28, 2012 at 2:37 am

    #1452442

    Thanks for that, seems it not best practice although would probably work.

  • February 29, 2012 at 11:43 am

    #1453319

    In SQL2008 In an Active/Passive configuration I simply run the service pack or CU on the inactive node then reboot. When the server is back up I fail over SQL Server to the updated server. Once it is up and running I connect to it and make sure everything looks OK. Then I patch the other node, reboot and then will fail SQL Server over to it to make sure SQL Server runs fine on that node as well. DONE!

    If you patch the inactive node and stop you run the fear that if something happens to the active node that SQL Server will fail over (without your control) to the patched node and get upgraded blindly.

Viewing 4 posts - 1 through 4 (of 4 total)

You must be logged in to reply to this topic. Login to reply