Migrating VMs to new host with shared storage

[dtawse]

Hello,

I'm looking for some advice on how best to tackle the following.

We currently have a single Hyper-V host server that is running several Windows VMs. The host is connected to a NAS device via iSCSI (Ethernet). Each of the VMs has one or more VHDs on this NAS. Veeam B&R 9.5 is running on the host and backing up to external HDDs.

I have a new host that I would like to move the VMs to, while keeping the VHDs on the NAS. I was thinking of using replication/failover to achieve this but I'm unsure of a couple of things. My rough plan is:

1. In Veeam, configure replication job of all production VMs to new host. Exclude VHDs hosted on NAS.
2. Run replication.
3. Shut down all production VMs on current host.
4. Run replication job to capture current state.
5. Failover VMs.
6. Shut down replicas (they auto-start on failover).
7. On new host, configure iSCSI and manually attach VHDs hosted on NAS to replicas.
8. Boot replicas and confirm functionality.
9. Perform permanent failover of all VMs.

Is this a reasonable approach? The main sticking point is with the VHDs on the NAS. These are very large so I can't replicate them to the new host and wouldn't want to anyway as they should stay on the NAS. Unfortunately, Veeam removes any excluded disks from the replicated VM settings so, as far as I can tell, they need to be added manually. I'm just hping that the automatic boot of the replicas on failover (missing the NAS VHDs) won't cause any problems.

Additionally, after the migration of the VMs, I'd like to move the Veeam installation to the new host, keeping all settings and history. Particularly, I want to avoid having to re-seed our offsite (CloudConnect) backup. How best to go about this?

Thanks

[foggy]

The less impact would be if you replicate the entire VMs (without any exclusions) and explicitly specify location for the disks currently stored on NAS to keep them on NAS (if there's enough space to temporarily store them twice). For seamless switch I recommend using planned failover. As for the backup server migration, then just use configuration backup and restore approach.

[dtawse]

Thanks very much for the reply, foggy.

Unfortunately, there is not sufficient space on the NAS to replicate (duplicate) the existing VHDs. Given this, is my approach the best reasonable option?

For the backup server migration, will the steps you've linked to avoid the need to re-seed? I ask because I seem to recall going through a similar process in the past and, having recovered the configuration on a new backup server, Veeam didn't seem to recognise the existing backup files as belonging to the restored jobs and initiated full backups of everything.

Thanks

[foggy]

Unfortunately, there is not sufficient space on the NAS to replicate (duplicate) the existing VHDs. Given this, is my approach the best reasonable option?

Yes, in this case manual disk re-attachment would be required. I would try with a test VM first, though.

For the backup server migration, will the steps you've linked to avoid the need to re-seed?

Provided repository servers, paths, etc. do not change, re-seed should not be required in this case.

[dtawse]

Great. Thanks very much for the help!