Created cluster, all VMs gone

[hashime]

Hello,

I have a proxmox server running and wanted to replace it with a better one, so i created a cluster on the old one, added the new one as a node. The Problem is, all VMs on the old one are gone. They still sit on the filesystem, but they are absent from the UI.
Also creating new VMs wont work anymore.
Any advice?

 

[udo]

Hello,

I have a proxmox server running and wanted to replace it with a better one, so i created a cluster on the old one, added the new one as a node. The Problem is, all VMs on the old one are gone. They still sit on the filesystem, but they are absent from the UI.
Also creating new VMs wont work anymore.
Any advice?

Hi,
you must do something wrong. Do you have create the cluster on the old node (with the VMs) first?

It's sounds, that your cluster was created on the new one and you joined the old one - in this case the content of /etc/pve is gone on the old node.

Anyway! Are your VMs are still running? In this case, you can see many info about the config in the command-line - use "ps -aux" to see the kvm processes.

Creating VMs don't work? Sound like no quorum...

What is the output of following from both nodes?

pvecm nodes
pvecm status

Udo

 

[hashime]

Created the cluster on the old node, storage disappeared from the web interface, vms are not running, since they are not startable. "connecting" to the local storage times out. It has been behaving weird since the jessie upgrade. Had to manually fix apparmor permissions for libvirtd and ownership of /dev/kvm. Seems upgrading is not really supported. Status and node output is as expected.

 

[udo]

Status and node output is as expected.

Hi,
difficult to help with no real output...

Created the cluster on the old node, storage disappeared from the web interface,..
It has been behaving weird since the jessie upgrade...

wait, do you talk about "creating cluster" or upgrade from version? to 4.x??


Was the old-node an standard installation? Any backups of /etc/pve?

Udo