Restoring Configuration Database to new Server

[Lyriclumpclass]

Hello All,

I have a quick question on the aformentioned subject. We have one B&R 9.5 server that currently does everything. This B&R server backups the configuration to an seperate file share daily. We use the B&R server to replicate a few VM's throughout the day (some hourly, some every 1/2 hour). What would happen if I restored the (daily) configuation database to a new VM with a different hostname/ip (i.e following the Migration path on https://helpcenter.veeam.com/docs/backu ... tml?ver=95) ?

More specifically, what would happen to the replicas given that the configuation database is restored to an earlier version ?
Is B&R smart enough to get all the info it requires to reconcile the database with the replicas directly from the ESXi infrastructure (presuming that there is some metadata here) ? If so, will it then need to do a 'full' replication cycle or will it happily carry on as before ?

[veremin]

More specifically, what would happen to the replicas given that the configuation database is restored to an earlier version ?

Could you elaborate on that portion a bit? What will happen, if there were some replication cycles not captured in configuration backup? Those will not be restored during restore of configuration, since the latter simply does not contain those.

Modifications in target infrastructure happened after configuration backup will also go unnoticed.

Thanks.

[foggy]

When replication job starts, it synchronizes actual replica configuration with the database. So everything will be smooth and the job will perform incremental run, as usually. The only thing you should keep in mind is replica metadata - in case it is not available for the latest existing restore point, digests recalculation process will be triggered.

[Lyriclumpclass]

Thanks for the quick assistance from the Veeam people. I'm quite impressed by the level of service as always (both now and in the past) !

So, from the responses, I am given to believe the following :
a) Whatever metadata is stored on the ESXi server (if any) it is not enough to reconstruct whatever B&R needs to resume replication since the last backup (in the case of moving B&R to a new server either deliberatly or if the current server fails). If this restore operation is done it will 'trigger the digests recalculation process' to bring it back up to scratch.
b) The 'digests recalculation process' involves scanning the entire source VM and the target (i.e. no CBT) and is likely to be quite a slow process ?
c) If there is a discrepency between the configuration and the replica then it will perform (b) after which it will 'clean up' and delete any ESXi snapshots since the last configuration backup.
d) The fact that the new B&R server has a new host name is irrelevent.
e) If I were to do things properly, then none of (a-c) are necessary. By this I mean
1) Disable all jobs (backup, replication etc).
2) Perform one last backup of the configuration
3) Disable all the various Veeam Services on the existing B&R server
4) Install Veeam on the new B&R server and move the licence to it
5) Restore from the backup done in (2)