First stop the corosync and the pve-cluster services on the node:
systemctl stop pve-cluster
systemctl stop corosync
Start the cluster filesystem again in local mode:
pmxcfs -l
Delete the corosync configuration files:
rm /etc/pve/corosync.conf
rm /etc/corosync/*
You can now start the filesystem again as normal service:
killall pmxcfs
systemctl start pve-cluster
