Each site is running local backups using Veeam and Site A is also currently running a replica job to site B (using VBR installed at Site A). This was initially 'seeded' using the initial replica to external HD which was shipped to Site-B.
To ensure I have VBR DB details available at Site-B site in case of an DR outage at Site-A, I'm thinking of running the replice jobs from the VBR server at site B (instead of site A). Few questions:
Will this impact performance - i.e. with the VBR server at the remote end doing the 'processing'
Can I export the existing job data from the site-A VBR to site-B VBR server?
If i need to recreate the jobs, how would the initially re-seeding work? I imagine this wouldn't be possible over the WAN as the external HD would tbe 'local' to Site-B and all the data is in Site-A?
Should I not worry about this, and just replicate the DB from Site-A to Site-B instead?
3. That wouldn't work, because you need to have an external HD locally at site A to be able to "seed" VMs from site A. Because seeding copies VM files to a external device, and if you try to do "seeding" from site B, all the traffic would go over WAN link. And I believe this is not what you are willing to see, right?
4. Yes, you may replicate the database using SQL server built-in options for that.
For 2 - I don't actually want to export all the jobs (such as the local backup jobs) - just the Replica job.
There seems to be a lot of dependency on the VBR server in Veeam compared to other products that are able to 'read' existing backup targets to repopulate all the job metadata. Can be quite important in emergency situation where all you might have access to (and should really need)is the backup media.
I'm assuming if I moved a VM to a different Virtual Center cluster then that would invalidate all the Veeam DB data as well and I'd have to start fresh?
Yes, I fully agree with you. If you have a look at the second link, you will see this functionality and other job management enhancements are in our roadmap already. But thank you for your feedback!
You're not quite right here. All the VMs are processed by their moref number. This number is assigned by a vCenter, so if you keep your VMs within one vCenter, they will have the same moref number and you wouldn't need to run your backup jobs from scratch after moving VMs to a different cluster.
Sorry - yeah I meant a different Virtual Center Server (we're about to decomission our exising VCenter server).
I guess if VC is the only way that VM's are uniquely identified then there's no way around it. Would be nice if Veeam could do some type of intelligent mapping when pointing a job at an existing backup media for example.. (eg Name + last host + some other identifier or even a manual mapping).
-