[pve-devel] [PATCH-SERIES pve-http-server/pve-manager] fix#4689: rewrite_uri: autofind nodename for qemu/lxc

Wolfgang Bumiller w.bumiller at proxmox.com
Wed May 31 13:49:06 CEST 2023

On Wed, May 31, 2023 at 01:19:50PM +0200, Alexandre Derumier wrote:
> Hi,
> Currently, to manage qemu && lxc vms, we always need to specify nodename in uri.
> This is a problem with automation tools like terraform, where is node is registered
> in the state of terraform.
> (That mean, than if we move the vm on another node, terraform don't known it, and try to create the vm
> again or can't delete the vm,...)
> https://github.com/Telmate/terraform-provider-proxmox/issues/168
> This can also be a potential problem with race, if we need to query /cluster/ressources to find the node, then another
> query on the vm.
> I have some discussion with fabian about it:
> https://bugzilla.proxmox.com/show_bug.cgi?id=4689
> This patch series, simply find the nodename if not specified and rewrite url.
> /nodes/(qemu|lxc)/<vmid> ---> /nodes/<nodename>/<qemu|lxc>/<vmid>

But then `qemu`/`lxc` look like a node name, which is weird.
If we do this, maybe we should use a different start entry altogether to
not mix too many "aliases" into the "real" api tree, like
`/virtual/qemu/*` or `/guests/qemu/*`?

More information about the pve-devel mailing list