Skip to content

Proxmox VE: single offline node causes all entities to become unavailable #165968

@jkmesches

Description

@jkmesches

The problem

When any node in a multi-node Proxmox cluster goes offline, all entities across all nodes become unavailable — including entities belonging to nodes that are still online and reachable.

I have a 5-node cluster. The integration connects through a node that stays online. When a different node shuts down, the coordinator's update cycle fails entirely and every entity goes unavailable until the offline node returns or I manually reload the integration.

The Proxmox API handles this fine on its own. From a surviving node:

  • GET /nodes succeeds and correctly reports the downed node as "status": "offline"
  • GET /nodes/<offline-node>/qemu fails because the API tries to SSH-proxy to the offline node: proxy handler failed: ssh: connect to host <IP> port 22: No route to host

The coordinator iterates over all nodes returned by /nodes and queries each one for VMs/containers with no per-node error handling. When the query for the offline node raises a ResourceException, it isn't caught individually — it aborts the entire update and raises UpdateFailed with the message "No active nodes were found on the Proxmox VE server," even though 4 out of 5 nodes are online and responding.

The node list already includes a status field that could be used to skip offline nodes. Ideally the coordinator would also catch per-node exceptions so a single unreachable node doesn't take down monitoring for the whole cluster.

What version of Home Assistant Core has the issue?

2026.3.2

What was the last working version of Home Assistant Core?

N/A — this appears to be a longstanding issue.

What type of installation are you running?

Home Assistant OS

Integration causing the issue

Proxmox VE

Link to integration documentation on our website

https://www.home-assistant.io/integrations/proxmoxve/

Diagnostics information

See attached file (redacted).

Example YAML snippet

__No response__

Anything in the logs that might be useful for us?

Logger: homeassistant.components.proxmoxve.coordinator
Source: helpers/update_coordinator.py:497
Error fetching proxmoxve data: No active nodes were found on the Proxmox VE server

Logger: homeassistant.components.proxmoxve.coordinator
Source: helpers/update_coordinator.py:432
Timeout fetching proxmoxve data

Additional information

Proxmox API output from a surviving node while another node is shut down:

root@node:~# pvesh get /nodes --output-format json
[...{"id":"node/offlinenode","node":"offlinenode","status":"offline","type":"node"}...]

root@node:~# pvesh get /nodes/offlinenode/qemu --output-format json
proxy handler failed: ssh: connect to host <IP> port 22: No route to host

The API already provides the information needed to handle this gracefully — the integration just doesn't use it.

Metadata

Metadata

Assignees

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions