Question regarding application of re-IP rules

[MSMSMSMS]

Support case ID is #01817516

We have Veeam 9 Enterprise Plus installed and configured on primary and on disaster recovery location. We are replicating several of our VM's using Veeam with WAN acceleration from primary to disaster recovery location. In case when we don't use re-IP rules, failover is finished in ~20 seconds, but when we use re-IP rules, failover takes ~15 minutes. Failover step that takes longest time to finish is “Applying re-IP rules” with something like ~13 minutes. When looking at logs, I can see that Veeam takes 13 minutes to do following:

[SVEEAM1] [FC] Checking whether filesystem in mountpoint [C:\Users\veeamsa\AppData\Local\Temp\obrmi0fu.vsn] is accessible.

Judging from that line and network traffic to SVEEAM1 VM, it seems like Veeam is doing that check remotely. Instead of using Veeam installation on disaster recovery location to do that check, it is using Veeam installation on primary location and transferring needed data over slow WAN link from disaster recovery to primary location.
Is it possible to configure this differently, so that re-IP uses Veeam installation on disaster recovery side?
Thanks,

MSMSMSMS

[foggy]

What do you mean by Veeam installation? Do you have two Veeam B&R backup servers, one per each location, or do you mean two proxy servers?

Basically, to apply re-IP rules during failover, the replica VM disk is mounted to the backup server. On slow WAN link this operation can take considerable time. It is generally recommended to have backup server responsible for replication jobs in target location to be able to perform automated failover/failback operations via Veeam B&R console in case of DR (when you lose your source site).

[MSMSMSMS]

Hello foggy,

Thanks for your reply. We have two Veeam B&R backup servers, one per each location. We don't consider our disaster recovery location equally secure in terms of physical access to our primary location, so we wanted to keep primary location vCenter credentials (that Veeam uses) out of secondary location, but having backup server on DR location handling replication jobs does have good sides (e.g. availability of Veeam Faileover plans and replica management through Veeam in case of complete primary location loss and much faster re-ip). We now need to decide between security (or our perception of it) and usability.

We do have one question regarding your proposed solution. Should we create new repository on primary location through "Veeam backup server on disaster recovery location" so we can keep replica digests close to source, or can we use existing repository on primary location currently used by backups made by Veeam server on primary location (our Veeam servers are VM's with all components on single server) for storing replica digests for replicas managed from disaster recovery Veeam backup server?

Thanks, MSMSMSMS

[skrause]

The digests etc. for the replication jobs should be on a repository in the same datacenter as the Veeam server managing the replication (aka the one in the DR location).

They shouldn't take up much disk space at all.

[foggy]

You can use any existing location in the source site, though it should be added as a repository on the DR Veeam B&R instance.

[foggy]

The digests etc. for the replication jobs should be on a repository in the same datacenter as the Veeam server managing the replication (aka the one in the DR location).

Not actually, it is recommended to store them closer to the replication job source, so in the source site, regardless of the backup server location. They are required to compare source data with what is already on target to avoid unnecessary traffic.

[skrause]

Ahh I had that backwards. I never noticed any issues, but I have a low-latency high-bandwidth link between my two locations and only replicate a few critical application servers.

[foggy]

Yes, this is mostly recommended for remote replicas via slower links.

[MSMSMSMS]

Hello Foggy,

Thanks for claryfiying it. In the end I've decided to create new repository (through Veaam B&R on DR side) on primary location (replication source site), instead of using existing repository used for Veeam backups on primary location. It seemed like cleaner option

Best regards, Marinko