Hi all,
We have a 4 node cluster and everything was fine until two of these nodes got disconnected. Now we have a situation where we seems to have lost the quorum with no way to get it back. We especially have some errors we can not get rid of, so any help would be appreciated:
The information I can give (please ask for more if necessary):
Logs from pve-cluster services:
This CS_ERR_BAD_HANDLE come back in any command about corosync quorum (like corosync-quorumtool -s )
Do you have any idea?
We have a 4 node cluster and everything was fine until two of these nodes got disconnected. Now we have a situation where we seems to have lost the quorum with no way to get it back. We especially have some errors we can not get rid of, so any help would be appreciated:
The information I can give (please ask for more if necessary):
Code:
# pveversion
pve-manager/8.0.4/d258a813cfa6b390 (running kernel: 6.2.16-14-pve)
Logs from pve-cluster services:
Code:
pmxcfs[3670817]: [quorum] crit: quorum_initialize returned wrong quorum_type: 0
Code:
# pvecm expected 1
You cannot change expected votes, corosync is not using votequorum
Code:
# pvecm status
Cluster information
-------------------
Name: INFRA
Config Version: 6
Transport: knet
Secure auth: on
Unable to start votequorum status tracking: CS_ERR_BAD_HANDLE
This CS_ERR_BAD_HANDLE come back in any command about corosync quorum (like corosync-quorumtool -s )
Do you have any idea?