Hello,
I had a three node cluster (PVE2, PVE3, and PVE4) setup while we prepared to migrate from VMWare. After successfully migrating the VMWare server to PVE2, I reformatted the VMWare server and joined it to the cluster as PVE1 and then went though the KB steps to properly remove PVE4 from the cluster. Everything seemed fine and we haven't had any issues until I noticed PVE4 still listed under HA quorum. My troubleshooting so far has uncovered the following:
pvestatus nodes is correct for all three current nodes. The removed node is not present.
root@pve1:~# pvecm nodes
Membership information
----------------------
Nodeid Votes Name
2 1 pve3
3 1 pve2
4 1 pve1 (local)
pvecm status looks good too...
root@pve1:~# pvecm status
Cluster information
-------------------
Name: KMCTest
Config Version: 5
Transport: knet
Secure auth: on
Quorum information
------------------
Date: Tue Jan 7 09:43:27 2025
Quorum provider: corosync_votequorum
Nodes: 3
Node ID: 0x00000004
Ring ID: 2.1b5
Quorate: Yes
Votequorum information
----------------------
Expected votes: 3
Highest expected: 3
Total votes: 3
Quorum: 2
Flags: Quorate
Membership information
----------------------
Nodeid Votes Name
0x00000002 1 192.168.100.3
0x00000003 1 192.168.100.2
0x00000004 1 192.168.100.1 (local)
But I notice that the previous original nodes (PVE2 and PVE3) show the following content under ~/pve/nodes. It does not show PVE1 and does show the removed PVE4 node.
root@pve3:~/pve/nodes# ls
pve2 pve3 pve4
If I check on PVE1, there is no pve folder under ~/ (root)
root@pve1:~# ls
root@pve1:~# pwd
/root
Not sure how to fix this. Any guidance would be very welcome.
I had a three node cluster (PVE2, PVE3, and PVE4) setup while we prepared to migrate from VMWare. After successfully migrating the VMWare server to PVE2, I reformatted the VMWare server and joined it to the cluster as PVE1 and then went though the KB steps to properly remove PVE4 from the cluster. Everything seemed fine and we haven't had any issues until I noticed PVE4 still listed under HA quorum. My troubleshooting so far has uncovered the following:
pvestatus nodes is correct for all three current nodes. The removed node is not present.
root@pve1:~# pvecm nodes
Membership information
----------------------
Nodeid Votes Name
2 1 pve3
3 1 pve2
4 1 pve1 (local)
pvecm status looks good too...
root@pve1:~# pvecm status
Cluster information
-------------------
Name: KMCTest
Config Version: 5
Transport: knet
Secure auth: on
Quorum information
------------------
Date: Tue Jan 7 09:43:27 2025
Quorum provider: corosync_votequorum
Nodes: 3
Node ID: 0x00000004
Ring ID: 2.1b5
Quorate: Yes
Votequorum information
----------------------
Expected votes: 3
Highest expected: 3
Total votes: 3
Quorum: 2
Flags: Quorate
Membership information
----------------------
Nodeid Votes Name
0x00000002 1 192.168.100.3
0x00000003 1 192.168.100.2
0x00000004 1 192.168.100.1 (local)
But I notice that the previous original nodes (PVE2 and PVE3) show the following content under ~/pve/nodes. It does not show PVE1 and does show the removed PVE4 node.
root@pve3:~/pve/nodes# ls
pve2 pve3 pve4
If I check on PVE1, there is no pve folder under ~/ (root)
root@pve1:~# ls
root@pve1:~# pwd
/root
Not sure how to fix this. Any guidance would be very welcome.
