[SOLVED] After failed cluster join, config in /etc/pve messed up

[alh]

Hi, after an error during cluster join operation the config files and dirs in /etc/pve were deleted/are missing.

root@srv1:/etc/pve# ls -lh
total 0
lrwxr-xr-x 1 root www-data 0 Jan  1  1970 local -> nodes/srv1
lrwxr-xr-x 1 root www-data 0 Jan  1  1970 lxc -> nodes/srv1/lxc
lrwxr-xr-x 1 root www-data 0 Jan  1  1970 openvz -> nodes/srv1/openvz
lrwxr-xr-x 1 root www-data 0 Jan  1  1970 qemu-server -> nodes/srv1/qemu-server
drwxr-xr-x 2 root www-data 0 Jul 19 02:33 virtual-guest

Is there a way to fix this without reinstalling?

 

[Dominic]

Have you restarted the server after the failed join? What is

sysstemctl status pve-cluster
journalctl -u pve-cluster

?

 

[alh]

Yes, I had restarted the server after the failed join and I also tried restarting the different services. In the end I just did the following:

# separate node
systemctl stop pve-cluster
systemctl stop corosync
pmxcfs -l
rm /etc/pve/corosync.conf
rm -r /etc/corosync/*
killall pmxcfs
systemctl start pve-cluster
rm /var/lib/corosync/*

# re-create nodes dir
mkdir -p /etc/pve/nodes/srv1

# re-create certs
pvecm updatecerts --force

# copy from another/similar node
# - datacenter.cfg
# - storage.cfg
# - user.cfg

Not sure this is the correct way but it gave me back the WebGUI etc.

On a sidenote: Is the cluster mode suitable to centrally manage all PVEs across several sites (over the internet) or should it only be used inside one datacenter (latency etc.)?

 

[Dominic]

Clusters require low latency and it is therefore recommended to have the servers as close together as possible. You can find more information the Cluster Network Requirements section of the Admin Guide.