SQL Server Failover Cluster – Steps to Apply Windows Updates to SQL Failover Cluster

sql serversql server 2014

What are the steps to apply windows updates to a active/passive two node sql failover cluster? My initial thought was to apply windows updates to passive node/reboot and then failover and repeat process but not sure. Do I need to stop the db's and sql agent jobs before failover?

sql2014
windows 2012r2

Best Answer

You do not need to stop any jobs or services to manually failover. Please see the Microsoft Docs - Perform a Forced Manual Failover on manually failing over.

These are the general steps you should follow for applying any updates to your failover cluster instances:

Backup all databases.
Failover all SQL cluster roles to Node A.
Install the patches on Node B (passive).
Restart Node B.
Failover all SQL cluster roles to Node B.
Install the patches on Node A (passive).
Restart Node A

More information on the installing updates process can be found in these Microsoft Docs - Upgrade a failover cluster instance.

Related Solutions

Sql-server – SQL Server cluster failover

Failover clustering works to provide constant service in the event of a failure (defining a failure as an "abnormal termination of the previously active application, server, system, or network" - http://en.wikipedia.org/wiki/Failover). Manual manipulation (start, restart, pause, and stop) upon the service being clustered (SQL Server, File Services, etc.) does not qualify as an abnormal termination of the service.

If I recall, at least in Windows Server 2008, you can simulate a failover test case in the Failover Management Console under Administration Tools. Check this out for a list of testing methods: http://blogs.technet.com/b/vipulshah/archive/2009/06/17/failover-cluster-testing-methods.aspx

Sql-server – SQL Server Mirroring kicks in before cluster has time to fail over

Database mirroring questions from TechEd: failovers and partner timeouts

... when the cluster is failing over and the principal instance is unavailable for 60-90 seconds, the other instances in the mirroring session will time-out after ten seconds and initiate a mirroring failover... The solution to the problem is to change the partner time-out value - a facility that isn't well known. To change the partner time-out value for a mirroring session, use the following code:
ALTER DATABASE mydatabase SET PARTNER TIMEOUT 90;

Best Answer

Related Solutions

Sql-server – SQL Server cluster failover

Sql-server – SQL Server Mirroring kicks in before cluster has time to fail over

Related Question