I have a cluster with 2 nodes, and Veeam is backing up databases every day and logs every 5 minutes. This works great and I can easily recover a single database to any point in time.
Then I was wondering what will happen if I completely lose a node. Let's say, worst case scenario, OS is gone. The AG will handle the failover and databases will still be accessible.
How can I restore the lost node?
I went into testing. To avoid a full reinstall of the node I took a backup of the VM. But obviously databases contained in the backup are older than those on the primary node of the cluster. So when I tried this I end up being stuck with databases out of sync.
First thing I did after the backup was restored was to restart the machine to kill. Still out of sync. As far as I understand, the backup I used was not a SQL database backup, and databases were in use (not that good) and the second error the databases on the primary and the broken secondary took different paths.
I found a really inefficient way to do it: remove databases from the AG, delete them on the secondary, then re-add them to the AG and let the secondary node sync.
Is it possible to resync databases more efficiently? Particularly, without losing the primary (meaning no restoration on the primary)?