Using SQL Server 2019 with 200+ databases in a 2 node FCI, what is considered best practice regarding SQL Server Agent when performing manual failover for SQL Server CU patching. Should we manually stop the SQL Server Agent before failover?
I ask because we see that sometimes failover takes over 10 minutes, and sometimes failover takes less than 3 seconds.
Should we manually stop the SQL Server Agent before failover?
I don't see why. When a failover occurs, SQL Agent is going to be asked to shut down due to the engine shutting down and also from a request from the cluster to shutdown to move the services.
I ask because we see that sometimes failover takes over 10 minutes, and sometimes failover takes less than 3 seconds.
Do you have any data that shows it's due to SQL Agent?