Encountered issues when attempting to bootstrap CloudStack with Proxmox

[ShadowJonathan]

Here are a few things I encountered while attempting to install/add Proxmox to cloudstack;

I started using Debian 13, the latest, considering that some/most things would work well on it.

However, there's no .debs for trixie, which I already addressed here: #12251

Then, when using debian 12's .debs, I encountered #12252 in that there's no documentation stating this repository needed to be added. Furthermore, debian 13 is also not supported in that repository, it seems.

Installing the connector manually (which worked), I finally got cloudstack properly setup.

Going through the setup (cloudstack-setup-databases, cloudstack-setup-management) was also no issue.

I then attempted to install a new zone, with the following zone-wide settings:

I could only select "external", not proxmox or the likes.

Then, when arriving at the step to add a cluster with a host, I was met with mostly blank pages:

Then, when launching the wizard, it fails on this page:

Which, after some debugging, is a client-side error here:

cloudstack/ui/src/views/infra/zone/ZoneWizardLaunchZone.vue

Line 1366 in 53a39d3

url = ['http://', hostname].join('')

(indexOf seems to be a JS client error)

---

Continuing, I go install secondary storage.

I had to remove the added cluster because it wasn't configured with the right proxmox settings, and I couldn't add a proxmox host.

So I install the host following this documentation

When I try to add primary storage, I get the following error message:

Which is because of this part in the proxmox.sh file:

cloudstack/extensions/Proxmox/proxmox.sh

Line 63 in 71f47d6

check_required_fields vm_internal_name url user token secret node

I skip primary storage, as it seems I can't add this, or that it's configured in some other way.

I add a template, to test launch a VM:

That works, then I try to create an instance from it, which fails with this:

I try to run it again, specifying the host, but this time it fails again:

I realise the zone was disabled, enable it, and recreate the VM, which gets me this error:

Looking at the logs, it gives me this:

2025-12-16 18:18:29,159 ERROR [c.c.v.ClusteredVirtualMachineManagerImpl] (Work-Job-Executor-10:[ctx-ca2a9424, job-49/job-51, ctx-8cb0346d]) (logid:ce949871) Unable to orchestrate start VM instance {"id":3,"instanceName":"i-2-3VM","state":"Stopped","type":"User","uuid":"53f4a94c-3124-4368-952e-7da10c9f472f"} due to [Cannot find an available cluster in Pod 2 to start domain router for Ovm.
Ovm won't support any system vm including domain router, please make sure you have a cluster with hypervisor type of any of xenserver/KVM/Vmware in the same pod with Ovm cluster. And there is at least one host in UP status in that cluster.].

Which is where I'll leave it for now, as this seems to be a proper blocker.

[weizhouapache]

@ShadowJonathan
as far as I know, proxmox-only zone (or other external provider) is not supported.
please correct me if I am wrong. cc @harikrishna-patnala @shwstppr @abh1sar

I have once created a zone with only proxmox, but it required some db hacks

[ShadowJonathan]

IMO, if this is the case, then "External" shouldn't be selectable in the zone creation wizard.

There's a few other points like these where the zone wizard attempts to allow you to select configuration that it then later on (when launching) tells you is impossible and unsupported, and - like this case - where it allows a certain action, but is unsupported in the end.

[abh1sar]

'External' as a hypervisor option should not be shown in the Zone creation wizard, as a primary hypervisor (kvm, vmware and xenserver) is needwd to run the systemvms.
Also update the documentation if it is not clear enough.

Created issue #12363 to track

[ShadowJonathan]

thank you! :D