Click here to monitor SSC
SQLServerCentral is supported by Redgate
 
Log in  ::  Register  ::  Not logged in
 
 
 
        
Home       Members    Calendar    Who's On


Add to briefcase

Problem with one side of a cluster Expand / Collapse
Author
Message
Posted Monday, August 5, 2013 8:24 AM
Hall of Fame

Hall of FameHall of FameHall of FameHall of FameHall of FameHall of FameHall of FameHall of FameHall of Fame

Group: General Forum Members
Last Login: Today @ 11:49 AM
Points: 3,173, Visits: 2,341
Good morning,

W have a SQL 2008 R2 cluster set up as an active/active. We had the default instance running on the A side and three other instances running on the B. Before I got here, they needed to patch to SP2 CU5 to fix an issue with one of the applications using the server. The A side was successfully patched, but there was a problem on the B side.The admin had to restore the operating system from backup. This caused some of the original installation files to disappear. Consequently the update could not find necessary files and failed. The two sides are now at different version levels. The cluster was switched to an active/passive with the B side being the passive.

I am not sure how to fix this issue. I am thinking we may need to do a complete reinstall of SQL on B, but that is convoluted with 4 different instances and analysis services and reporting services all installed.

I have not worked with a sql cluster since 2007. (SQL2005).

Any ideas?
Thanks


Mike

“I know that you believe you understand what you think I said, but I'm not sure you realize that what you heard is not what I meant.”...Robert McCloskey

___________________________________________________________________
Post #1480940
Posted Tuesday, August 6, 2013 8:22 AM
SSC-Addicted

SSC-AddictedSSC-AddictedSSC-AddictedSSC-AddictedSSC-AddictedSSC-AddictedSSC-AddictedSSC-Addicted

Group: General Forum Members
Last Login: Today @ 4:26 AM
Points: 452, Visits: 962
I personally would remove the passive node from the cluster and reinstall. Remember when adding a node to the cluster all setup options are taken from the active node so it should be a case of Next Next Next on the install wizard.

Once reinstalled, patch the passive node to the correct level then perfom a failover enusing all comes back online.

SSRS is not cluster aware so you should not really have any problems there, if the servers only host the SSRS databases and not the front end aspect then you should have no issues. But if you are using both as SSRS Web servers in a scaleout deployment you will need to recreate all the settings on your current passive node.


MCITP SQL 2005, MCSA SQL 2012
Post #1481359
Posted Tuesday, August 6, 2013 11:00 AM
Hall of Fame

Hall of FameHall of FameHall of FameHall of FameHall of FameHall of FameHall of FameHall of FameHall of Fame

Group: General Forum Members
Last Login: Today @ 11:49 AM
Points: 3,173, Visits: 2,341
That is what we are currently planning on doing.
Thanks.


Mike

“I know that you believe you understand what you think I said, but I'm not sure you realize that what you heard is not what I meant.”...Robert McCloskey

___________________________________________________________________
Post #1481457
Posted Tuesday, August 6, 2013 10:32 PM
Valued Member

Valued MemberValued MemberValued MemberValued MemberValued MemberValued MemberValued MemberValued Member

Group: General Forum Members
Last Login: Thursday, April 21, 2016 12:50 AM
Points: 67, Visits: 471
I actually had a similar issue once where the first node had an error during cluster setup but the added node didn't so to install service packs i had to fail everything over to the second node then opted for a complete rebuild of the first node and rejoined it to the cluster.
Post #1481642
Posted Monday, August 19, 2013 8:33 AM
SSC Journeyman

SSC JourneymanSSC JourneymanSSC JourneymanSSC JourneymanSSC JourneymanSSC JourneymanSSC JourneymanSSC Journeyman

Group: General Forum Members
Last Login: Tuesday, November 12, 2013 9:40 AM
Points: 81, Visits: 29
This was exactly what we had to do. We evicted the node from the cluster,
completely rebuilt the server, joined it to the cluster, then reinstalled sql and patched to correct level.
Everything is now at the same level and failover works correctly.
Post #1485848
« Prev Topic | Next Topic »

Add to briefcase

Permissions Expand / Collapse