Proxmox crash after renaming a host as is described in the official documents!

[mjdtjm]

Hi there,
I've just renamed a host in a single node proxmox installation as is described in the official documents here:
https://pve.proxmox.com/wiki/Renaming_a_PVE_node

After doing all steps in the document and rebooting the node, the /etc/pve directory is empty (not mounted), and the GUI in port 8006 is not accessible!
What's happened?!

 

[mjdtjm]

Unfortunately it seems that the above mentioned official document is not complete!
I suggest the following howto for hassle-free renaming of a node:
----------------------------------------------------------------------------
Step by step howto rename a proxmox node/host:

1. Shutdown all of the VMs and containers that reside on the node.
2. Run these commands in a root shell:
# service pve-cluster stop
# pmxcfs --local
3. Perform the steps in http://wiki.debian.org/HowTo/ChangeHostname
4. Perform the steps in https://pve.proxmox.com/wiki/Renaming_a_PVE_node
Be careful: you should use a utility like "mc (Midnight Commander)" to manually build new hierarchy under "/etc/pve/nodes/<New Name>" and move files from "/etc/pve/nodes/<Old Name>" to it.
5. Run this to find configuration files that need manual corrections :
# find /etc -type f -print0|xargs -0 -I ff grep -l <Old Name> ff
6. Investigate one by one of found files and correct the important ones (such as /etc/pve/storage.cfg).
7. Now, reboot the node and hope to have the running node with new name!

 

[Michael.Uray]

I had the same issue today when I tried to rename a node.
It was very strange, the file /etc/pve/lxc/101.conf did not exist anymore and I was also not able to create one (error: file already exists).
In the end I changed the name back how it was before and did a reboot to get the server back into operation.

 

[Dunuin]

"/etc/pve" is not a normal folder. It's the content of the config database mounted there by the pve-cluster service. If the service can't start, the folder will be empty. See: https://pve.proxmox.com/wiki/Proxmox_Cluster_File_System_(pmxcfs)

So the question is probably why your pve-cluster service wasn't running anymore and a systemctl status pve-cluster.service probably could have been giving you some hints on if/why its not working.