Dedicated Veeam Backup network in VCF8+

[Lopese]

Hello
I am designing a dedicated back-up network to include two different vsphere clusters existing on the same management LAN.

My goal is to route all backup traffic through a dedicated network.
How to i achieve this goal?

Esxi hosts have default host managent vmkernal1 (172.30.80.0/24)

Do I create a separate vmkernal2, with service as management for each esxi host in (10.50.10.0/25).
Create Veeam proxy server with nic in same subnet as vmk2.
Create a static route pointing traffic in (10.50.10.0/25) through veeam proxy VM, join veeam management server with backup lan network ip for hosts (10.50.10.0/25).?

Thank you.

[vnikiforov]

Hello, Eddie,

Deploy one proxy VM per cluster that leverages the HotAdd backup method; this is the recommended option. Give each proxy two vNICs: one on the management LAN (so VBR can task it and reach vCenter), one on 10.50.10.0/25 (the backup VLAN). Each host in the cluster already has a second VMKernel in that same subnet with the Management service enabled. And VMDK reads should then happen internally via hot-add.

On the repository side, give the repo (or you may use a single repository gateway for the repositories for example) server a NIC in 10.50.10.0/25 as well.
Add 10.50.10.0/25 as a Preferred Network in VBR.
Specifying Preferred Networks

That's what forces the proxy>repo data stream onto the backup VLAN.
Without a repo-side NIC on that subnet, Veeam Backup & Replication has nowhere to route it and it will fall back to management network.
If that implementation doesn't work for some reason, I'd recommend you to open a case with Veeam Technical support on that matter, who can analyze the setup with the logs.