The /etc/pve directory disappeared

Jonatanas12

New Member
May 29, 2023
3
0
1
The directory /etc/pve when making changes to the server disappeared, I reverted the changes but still can not see the content, what do I do? I have no backups.
 
As you can see there is nothing in the directory and when I want to restart the services it does not let me
 

Attachments

  • Screenshot_4.jpg
    Screenshot_4.jpg
    36.4 KB · Views: 23
  • Screenshot_3.jpg
    Screenshot_3.jpg
    129.2 KB · Views: 20
  • Screenshot_2.jpg
    Screenshot_2.jpg
    212.7 KB · Views: 20
As you can see there is nothing in the directory and when I want to restart the services it does not let me
That's normal when the pve-cluster service does not start. Can you find the actual reason for the failure in journalctl -b (you might have to scroll and search a bit)?
The directory /etc/pve when making changes to the server disappeared, I reverted the changes but still can not see the content, what do I do?
Maybe you can tell a little about the changes and how you reverted them?
I have no backups.
It's fine if you don't care about the VMs/containers.
 
Last edited:
Changes were made to the hosts file to rename the clusters and then returned to the same name.

As you can see I have no more errors that the keys do not exist. and another one I found about ip
 

Attachments

  • Screenshot_5.jpg
    Screenshot_5.jpg
    519.4 KB · Views: 12
  • Screenshot_6.jpg
    Screenshot_6.jpg
    518.9 KB · Views: 10
  • Screenshot_8.jpg
    Screenshot_8.jpg
    59.9 KB · Views: 10
There might be something wrong in one of your /etc/hosts file. Please post it and make sure each IP + name is correct in the /etc/hosts file.
Renaming a node while still in cluster, might make some issues. The correct way is to remove the node and the readd it again to the cluster. Please also post /etc/pve/corosync.conf and /etc/network/interfaces of each host.
 
Last edited:
  • Like
Reactions: leesteken
I just had to add a comment here. A fairly new Proxmox host that has been running for a couple weeks with numerous reboots, was updated today. Once I rebooted, I started getting error messages and I noticed that the /etc/pve folder was empty. I don't believe I changed anything since the last reboot, but I wouldn't bet money on it.

However, I did find that the /etc/hosts file contained an entry:
127.0.0.1 localhost.localdomain localhost
127.0.0.1 ps1.example.com ps1

I changed it to the public IP for the host, similar to the following:
72.12.34.56 ps1.example.com ps1

low and behold, the pve-cluster service was able to start up, and a reboot put everything back to normal. Thank god I was able to ssh into the host, as I had already arrived home for the weekend.
 
127.0.0.1 ps1.example.com ps1
How did that line get in there? Did you enter 127.0,0.1 in the Proxmox installer (instead of a local network IP address)? It should (always) contain the IP address of the Proxmox host that is reachable from other systems in the (local) network
 
Sorry for the late reply. I have installed ProxMox many times, and I would never enter 127.0.0.1 as IP address. It's a rule that none of my servers even use DHCP and instead use a reserved bank of internal or public static IPs. Of course, it could have been a brain fart, but that's what I didn't understand: The device was working fine for weeks and even rebooted, but an update on that day and a reboot displayed the problem.