How to disable Worker Failover to remote sites over MPLS?

[YzeeGeo]

Hello,

I am currently deploying Veeam V13 to manage 8 remote sites connected via MPLS. Each site has its own Proxmox VE node, a local Dell DD3300 storage, and a local Veeam Worker.

The Issue:
This weekend, a local Worker failed to boot. The Proxmox "Start" task looped for 36 hours without success. Instead of failing the job, Veeam automatically selected a Worker from a different remote site over the MPLS to perform the backup of a 600GB VM.

This caused a massive saturation of our WAN links for two days.

Current Setup:

• Veeam Version: 13 (latest build).

• Architecture: 1 Central VBR server, 8 remote Proxmox nodes.

• Repositories: 8 Dell DD3300, each configured with a local Linux Gateway (Debian) to keep traffic local.

The Problem:
I cannot find any option in the V13 Proxmox Worker or Job settings to:

• Disable Network Mode (NBD) failover: I want the job to fail if HotAdd is unavailable, rather than falling back to the network over the WAN.

• Restrict Worker Affinity: I need to strictly bind a Proxmox node to its local Worker and forbid any "cross-site" help.

In VMware, we have the "Proxy Selection" and "Transport Mode" (HotAdd only) options. In the Proxmox V13 integration, these settings seem to be missing or fully automated.

Questions:

• Is there a registry key or a hidden setting to disable "Automatic Worker Selection" across the entire infrastructure?

• How can I ensure that Site A never uses a Worker from Site B, even if the local Worker is offline?

I want the backup to FAIL rather than saturate my MPLS.