I am Using SCOM 2019 UR2 to monitor a number of virtual SQL Instances, configured to use the health service SID for authentication
Just recently it has been noticed that when an SQL instance fails over from one host to another the scom agent on the old host still attempts to query the SQL instance (using its own computer account), resulting in Multiple SQL login failures in the SQL logs (only the localhost is authorised in SQL )
this is not something i've seen before nor can I confirm it is happening to all sql instances that migrate(i have no test environment to start this type of testing) so i was hoping someone may have seen this previously or possibly offer a another insight ?
the performance metrics and health monitors for the SQL instance still seem to collect and perform as usual so it appears both computers are querying the instance following a migration and scom is non the wiser offering no access denied errors on either hosts
if I clear the cache on the agent the problem is resolved , so i have to setup some monitoring to catch the SQL login errors and automate a clearing of the agent cache but i dont really want to just patch up the issue and walk away
Thanks in advance