Veeam 9.5 u4a SQL DB didn't failover in Always on SQL Cluster

[aman4God]

We have a new Veeam install that is pointed to an SQL Listener on our SQL 2016 Enterprise cluster that is an Always On Availability group. I am not a DBA so forgive any bad terminology. The cluster failed over last night and now the Veeam database won't connect. Does Veeam 9.5 support connection to an Always on group? Is there a way to configure the BR server for this setup or does it need to be moved to a stand-alone server? This is the first time I have installed in this configuration and I cannot find any documentation related to always on SQL.

[Dima P.]

Hello Stanton,

Unfortunately there is no native integration with SQL AON. From the top of my head you need to manually point your Veeam B&R server to a new SQL server (another SQL node) by updating the registry hive

HKEY_LOCAL_MACHINE\SOFTWARE\Veeam\Veeam Backup and Replication with the new SQL server name / SQL database name (new credentials might be required as well)

In order to do that correctly please raise a support case. Hope that helps!

[skrause]

Are you connecting to the cluster's AD hostname or are you connecting via IP?

While Veeam might not seamlessly reconnect automatically if the primary server changes in the cluster, if you are connecting to the Cluster's AD object then any applications that connect to it should reconnect to the cluster when their services are restarted.

We use SQL Always On extensively and our connection strings are always to the Cluster's AD entry.

(ex: Cluster=PRODDB, Server1=PRODDB01, Server2=PRODDB02 where we point all of our SQL connections to PRODDB)

I will be doing some extensive testing of Veeam products (VBR, Enterprise Manager, and VeeamONE) in a couple of months as we rebuild/migrate our datacenters so I can't say for certain how Veeam would behave. We currently have one application that freaks out when the cluster fails over and has to have its services restarted but it doesn't require a connection string change.

[Andreas Neufert]

AD object way + Veeam Backup Service and Veeam Enterprise Manager Service restart worked im the cases that I remember from some versions ago.