Testing backup jobs over Remote Access VPN

[ksimon]

Case #06048748

Additional Details: We are attempting to build an infrastructure for testing backup jobs for multiple clients.

In our old scenario, we had a site-to-site VPN to each of these clients. It came in over our main Internet service. This worked fine, provided no one wanted to access the Internet during a full restore job.

We have built a new network that is on its own Internet connection. We are able to create a remote access VPN connection from the client B&R machines and then add a Hyper-V server to that B&R machine.

However, when you attempt a restore job from the client B&R machine, it fails when the 'receiving' Hyper-V machine can not find the repository.

I suspect that Veeam B&R is orchestrating the restore job by telling the Hyper-V machine to pull the data from the repo directly.
This will always fail in this scenario.

Is there a solution for this issue? Preferably one that does not involve multiple new servers.

[Mildur]

Hi Ken

That is correct.
When you do an entire VM restore or instant recovery, data flow is from repository directly to the backup repository.
We have data movers on the HyperV Server and repository server. They need to be able to talk to each other for the restore.

Here is a graphic for the instant recovery (manual testing of vm recovery):
https://helpcenter.veeam.com/docs/backu ... ml?ver=120

We are able to create a remote access VPN connection from the client B&R machines and then add a Hyper-V server to that B&R machine.

Please note, that for such a scenario, all tenant backup server must have the same build level of Veeam Backup & Replication. We deploy our data mover services (and others) to the HyperV machines. If you add the HyperV server to a backup server with a different build number, those services will be replaced or adding may fail if newer ones were detected.

Best,
Fabian