How to remove node from "Remotes"
- Thread starter bigAboo
- Start date
Correct. Just to be sure, did you read https://pve.proxmox.com/pve-docs/pve-admin-guide.html#_remove_a_cluster_node ?
"At this point, you must power off hp4 and ensure that it will not power on again (in the network) with its current configuration."
There seem to be some pitfalls...
Solve it - delete token from node and edit remotes file in /etc/proxmox-datacenter-manager- work fine
Am I wrong or is this a link to the PVE documentation? Here we are talking about the PDM, not a PVE cluster so I hope you can remove and add nodes as needed?
Correct. Maybe.
Unfortunately I am not a PDM user. Sorry if I linked to unfitting documentation...
You can already use a single PVE node/cluster with multiple PDM instances. Just use a different API token name.
The report from the original reporter here is a different error where we should provide a workflow to force remove such nodes and/or remotes - would be nice to get a issue opened at https://bugzilla.proxmox.com/ to keep track of this.
The report from the original reporter here is a different error where we should provide a workflow to force remove such nodes and/or remotes - would be nice to get a issue opened at https://bugzilla.proxmox.com/ to keep track of this.
Thanks for the clarification, Thomas.
I ran into a similar issue recently while re-adding a node after certificate changes in PDM. In my case, the stale remote entry and API token mismatch were causing conflicts.
What worked for me was:
It might be helpful if PDM provided a forced cleanup option or better error messaging when a node is stuck in this state. Opening a Bugzilla issue sounds like a good step to track this
Curious if others have seen similar behavior or found a cleaner approach?
I ran into a similar issue recently while re-adding a node after certificate changes in PDM. In my case, the stale remote entry and API token mismatch were causing conflicts.
What worked for me was:
- Removing the outdated API token from the node
- Cleaning up the corresponding remote entry in /etc/proxmox-datacenter-manager/remotes.cfg
- Then re-adding the node using a new token name (as you suggested)
It might be helpful if PDM provided a forced cleanup option or better error messaging when a node is stuck in this state. Opening a Bugzilla issue sounds like a good step to track this
Curious if others have seen similar behavior or found a cleaner approach?