Agent Backup Failover Cluster - Replacing SQL Servers

Userlevel 5



What we have:

  • A Protection Group (protection-sql) containing SQL Server Cluster: sql_c01 with two sql Servers. sql1 and sql2
  • I have one agent Backup Job bj-sql-01 , Type “Failover Cluster”. Useing a sobr with aws capacity and archive tiers. 
  • Anything is working absolutely fine. Backup ist fine. Offloading is fine. 

Whats going on:

  • Both sql servers will be renewed. There will be two new sql server. sql1-new and sql2-new (agent already installed)
  • After installation, both new sql servers will useing the same SQL Server Cluster sql-c01
  • sql1 and sql2 will be turned off.

My Accepetion:

Should I build a new backup job (clone) or is it possible to keep the existing backup job. bj-sql-01 ?

  • I would manually initiate a Protection Group rescan. The new sql-server should appear in the Protection Group protection-sql the OLD sql server should disappear, because they do not belong to the sql cluster anymore.
  • The existing backup job should recocognize the missing sql Server but will continue backup the new servers. I expect an active full backup, because auf new servers. 



Best answer by JMeixner 25 April 2022, 15:17

View original


Userlevel 7
Badge +22



Do you need to keep any particular backups prior to decommission? If so I’d go for a new job so you can tell from backup job name which was old and new. I’m going to assume your recovery requirements will be quite limited once you’ve finished your migration as, assuming your AD survives, you wouldn’t do a full VM restore of your old SQL servers anyway due to the cluster not referencing the old sql servers anymore.


So in most scenarios, reusing existing will be fine. But for cleanliness, may as well make a new job.

Userlevel 7
Badge +17

If I had to do this, I would create a new Job for the new cluster and let the data of the old cluster age out and delete the remainders after some time…

As @MicoolPaul said, it is more clean for me this way...