Clone VMs in Proxmox Cluster from templates

[apanescu]

Hello,

I have a cluster of 9 nodes in Proxmox. And I want to clone VMs on one of the nodes with a template from another node, but I get an error which says that the Proxmox Node is not allowed for this action. I also attached a screenshot with the error.

Does someone have any clue about this problem? Do I need to upgrade/install some packages to have this option available?

Thanks!
Andre

 

[Stefan_R]

For a linked clone to work across nodes, the template's disks have to be stored on a shared storage (e.g. Ceph).

 

[SamuelL]

For a linked clone to work across nodes, the template's disks have to be stored on a shared storage (e.g. Ceph).

I'm having this same issue except I'm doing a full clone. Both nodes have the same drive config and are both ZFS so there shouldn't be any issues. See attached photo.

 

[Stefan_R]

The drives need to be on shared storage for a full clone too. Cloning the disk over the network is currently not implemented, but you are free to open a feature request on our bugtracker. Note that you can always just migrate the template beforehand and then clone locally.

 

[timdonovan]

"For a linked clone to work across nodes, the template's disks have to be stored on a shared storage "

No worries, I'll do a full clone!

"Oh that won't work either"

Why is everything simple so painful in Proxmox? Cloning is effectively the same as migrating except you don't delete the running VM. It's effectively nothing more than a cp command. Ugggh.

 

[tom]

Why is everything simple so painful in Proxmox?

Templates and cloning works perfectly for me. Do basic checks in your Proxmox VE node/cluster to find the cause of your problems.

If something does not work, you will see clear error messages and log files.

 

[timdonovan]

I'm groaning about the fact that "Cloning the disk over the network is currently not implemented".

Instead I just tried to migrate a server, but it seems like Proxmox migrate totally ignores thin volumes, and doesn't use compression. So the migrate feature is currently shifting 28GB worth of 000000's. slow clap.

 

[tom]

if you have a cluster, just use a shared storage or Ceph. Migration of local disk over a network is always a pain by definition.

 

[timdonovan]

That is fair enough I suppose! Thanks.

 

[paul_b]

I had the same problem and it helped me:

in the storage settings, select nodes instead of all

note that storage is not called local-lvm, there will be an error:
can't clone VM to node 'node01' (VM uses local storage) (500)

 

[lauhub]

As none of the above worked for me, I finally found that it was necessary to remove the local storage's ISO disk that was inserted to make clone working.

So a message like "remove any media from CD/DVD drive before cloning" would be a good clue instead of the cryptic "not allowed for this action".