Anyone had success on Proxmox Worker on shared storage

[sys92]

On kb docs its mentioned that : "The default local storage must be enabled on all hosts where worker VMs will be deployed. If you cannot use the default storage in your environment, contact Veeam Customer Support." when I raised ticket, engineer was suggesting to use local storage only.

Kindly someone share the method to configure shared storage for proxmox workers (Since local storage disk is just 480GB I cannot use it for worker).

I have a shared volume (FC-SAN LVM) available where all the vm are hosted .

I am reaching forum since support request case was not useful (Case #07905795 ). we procured license exclusively for vmware to proxmox migration now basic support sucks!

[rovshan.pashayev]

Hi,

Veeam Plug-in for Proxmox VE requires at least one file-level storage (such as NFS) configured in the Proxmox Virtual Environment. This is because the plug-in does not support environments that rely solely on LVM-based storage, please see the User Guide.

This storage is used not only to deploy worker VMs but also to store data generated during backup of protected VMs.

As a workaround, you can create a VM on your LVM storage and configure it as an NFS server. Then, add this NFS share to Proxmox Storage. This will allow you to store workers and the data generated during backup of protected VMs.

[rovshan.pashayev]

Hi @s@sys92

Were you able to try the NFS‑based workaround and see if it suited your environment?

[pascalb97]

Hi,

I have similar questions, since we are facing the same challenge of integrating Veeam with Proxmox using FC-SAN based shared LVM storage (like many others).

1. When enabling snapshot-as-volume-chains (I am able to snapshot VMs in Proxmox using that), does Veeam still store snapshot data on that file-level snapshot storage?
2. If so, how big should that Snapshot storage be? We plan to migrate VMs with up to ca. 30TB of used space from our VMware infra to Proxmox and we are running many VMs in general. These are VMs we are already backing up with Veeam configured for VMware. Do I theoretically need a 30TB snapshot storage for that (when running a full backup for example)?

I am looking forward to you answers, thanks for your support!

[pablo.gutierrez]

Same problem, to add Proxmox Worker only show local-lvm y not shared LVM (iSCSI Storage), any workaround?

[rovshan.pashayev]

Hi Pablo,
You’ll need to submit a Support Case so the Support team can guide you toward a possible solution, depending on your environment.
Please keep in mind that, in any case, you will still need to use file‑level storage for the worker(s)

[bgatestmg]

we just went through this also and support was not very helpful. What we did for the interim was create the workers on NFS storage as we had a qnap kicking around we could easily configure for this and afterwards i just migrated the disk to the iSCSI storage we use with our nimbles and it seems to work alright. Definitely not ideal by any means.

[pfrancoeur]

Hi,

I'm trying to find a way to deploy the Veeam worker iso somwhere else than the "local" (/var/lib/vz/templates/iso/) storage?

Our proxmox inastallation is running from a 32Gb local storage that is somewhat limited and adding 1.1Gb brings us below 30%.

I tried disabling the local storage completely but it fails the worker deployment.

Any other way to move it somewhere else?

[rovshan.pashayev]

Hi @pfrancoeur,

I moved your post to the existing topic.
You’ll need to submit a Support Case so the Support team can guide you toward a possible solution, depending on your environment.
Please keep in mind that, in any case, you will still need to use file‑level storage for the worker(s)

[sys92]

This issue has not been addressed so far.

There should be flexibility in selecting the datastore/storage location for worker VM deployment, similar to the flexibility Veeam already provides for network and snapshot datastore selection.

Currently, worker VM deployment is restricted to local storage only. In many production environments, this is not practical, as some nodes may not have sufficient local disk capacity available. For example, deploying a single worker VM may require approximately 100 GB of free space, which cannot always be guaranteed on every node.

Allowing administrators to select the target datastore or shared storage for worker VM deployment would provide greater operational flexibility and help avoid deployment failures caused by limited local storage availability.

I am not sure whether this limitation has been addressed in version 13 or later releases. However, as of Veeam Backup & Replication 12.3.x, this feature appears to be unavailable. Unfortunately in version 13, VMware 6.x is removed having extended support for 6.x could have helped many legacy infra