Proxmox Worker Deployment Failure: 5-minute Timeout

[YzeeGeo]

Hello everyone,

I am encountering a persistent issue when deploying Proxmox Workers on remote sites with limited bandwidth (approx. 20 Mbps).

The Setup:

• VBR Server: v13

• Proxmox VE: 9.x

• Remote Site: Connected via a 20 Mbps link.

• Worker Image Size: ~950 MB.

The Problem:
Every time I try to "Add Worker" on these specific sites, the upload starts correctly, but it fails systematically at exactly 5 minutes (300 seconds).

Logs Analysis:
In the Veeam Task logs, I see the following error:
ERROR | [ProxmoxNode][HOSTNAME]: ==> Response "Post" "https://<IP>:8006/api2/json/nodes/<NODE>/storage/local/upload", "status: 995 - OperationAborted", duration: "5 min 1 msec"

On the Proxmox side, journalctl -u pveproxy reports:
problem with client ::ffff:<VEEAM_IP>; Connection reset by peer

What I've tested so far:

• Increased Proxmox API Timeout: I created /etc/default/pveproxy with TIMEOUT=1800 and restarted the service. No change; the connection is still reset at 5:00.

• Bandwidth Check: At 20 Mbps, the 950 MB file needs about 8-10 minutes to upload. Since it cuts at 5:00, only about 500-600 MB is transferred (confirmed by watching the pveupload-* file in /var/tmp on the PVE node).

• Manual SCP: Transferring the same .img file via SCP to the PVE node works perfectly, which rules out a basic network stability issue.

Question:
It seems there is a hardcoded 300s timeout within the Veeam Proxmox Transport/Appliance service for the HTTP POST upload to the Proxmox API.

• Is there a configuration file within the Veeam Appliance to increase this specific UploadTimeout?

• If not, is there a supported way to "pre-stage" the Worker image on the Proxmox storage so the wizard detects it instead of trying to re-upload it?

This is a major blocker for distributed environments where high-speed fiber isn't available on every branch.

Thanks for your help!

[YzeeGeo]

Hello everyone,

I wanted to share a solution for those struggling to deploy Proxmox Workers on remote sites with limited bandwidth (e.g., 20 Mbps).

The Issue:
When adding a Worker to a remote Proxmox node, the ~950 MB image upload would systematically fail at exactly 5 minutes (300 seconds) with the following error:
Response "Post" ".../upload", "status: 995 - OperationAborted", duration: "5 min 1 msec"

On the Proxmox side, logs showed: Connection reset by peer.

At 20 Mbps, the upload requires about 8-10 minutes, but the Veeam Appliance has a hardcoded default timeout of 5 minutes for REST API connections, causing it to kill the transfer before completion.

The Solution:
The timeout is managed within the Appliance's internal configuration. Here is how to increase it:

Log into the Veeam Appliance Web UI.
Go to Logs and services -> Host Configuration.
Locate the file: /etc/veeam/platform-service-pve/appsettings.json.
Export the file to your computer.
Open the JSON file and locate the following two blocks: "HypervisorRestClient" and "UpdaterRestClient".
Change the "ConnectionTimeout": "5 min" value to "15 min" (or more depending on your speed) in both blocks.
Import the modified file back via the Web UI.
Reboot the Veeam Backup server/Appliance to apply the changes.

After applying this, the upload finished successfully in about 9 minutes on my 20 Mbps link. I hope this helps anyone dealing with distributed Proxmox infrastructures!