post vmware reinstall veeam reconfiguration question

[cmdrriker]

We're preparing to do a clean install of esxi 5.1 on two servers that are currently running esx 4.1. A new virtual center
will also be created. The esx servers are both currently veeam replication hosts. After the esxi hosts are newly installed, and
added to the new virtual center, what needs to be done to Veeam to resume the replication jobs? Some of the steps I am thinking
need to be done are:

remove the old virtual center and hosts from veeam (will there be a problem with this? should this be done prior to the hosts being taken down?)

add the new virtual center and hosts to veeam

rescan the datastores to find the existing replicas

use the existing replicas as seeds for the new replication jobs that will be created.

Is there anything else that would need to be done to resume replication after a clean vmware install?

The Veeam version is 6.5

[Vitaliy S.]

Hello Will,

If you remove vCenter Server, you will need to re-create all replication jobs and re-map source VMs to VM replicas. No issues should be observed, but keep in mind that existing restore points of your VM replicas will be removed.

Hope this helps!

[cmdrriker]

Vitaly,

Thank you for the heads up on the restore point removal. I have a question though, because I'm not sure I understand why. The replicas will not be moving, they will remain in the same place on the same luns that will be re-mounted on the servers after esx4.1 has been wiped and esxi 5.1 has been installed. Then the replicas will be registered on the newly installed esxi hosts.(which will be in the new virtual center) Then I planned to create new jobs, using the existing replicas as seeds, and storing the new replica in the same location as the parent seed. I guess I hoped the replication might then continue where it left off. Maybe it won't because the vm id's will have changed? If the replication won't continue where it left off, thereby preserving the restore points, then is there any benefit to using the seed method versus just creating a new job from scratch?

[foggy]

I guess I hoped the replication might then continue where it left off. Maybe it won't because the vm id's will have changed?

Exactly.

If the replication won't continue where it left off, thereby preserving the restore points, then is there any benefit to using the seed method versus just creating a new job from scratch?

Seeding allows to avoid full VM data transfer, only changes will be sent to synchronize the replica with the original VM.

[cmdrriker]

Ok, one question though, after a successful replication using the seed, can I then delete the previous replica used for the seed, or do those vmdk's become the parent disks for the new replica?

[veremin]

The latter, the bottom line of mapping functionality is that already existing VM becomes a target for replication job.

If you put replica mapping into use, during the first run, only differences between VMs will be calculated and transferred. These differences later will be used to synchronize the state of VM in DR site and the original one in Production site.After successful synchronization, there will be only two restore points. One of which is the latest state of the mapped VM (target VM), another is the latest state of the original VM on the source host.

Thanks.

[cmdrriker]

Thank you, but does this mean that I can never delete the seed replica, because the new replica based on the the seed will point to the seed vmdk as a base for the new replica?

[foggy]

Yes, the VM used as a seed for the replica, will become your new replica VM. So you should not delete it.