[SOLVED] Unable to remove old cluster nodes

[cdukes]

The GUI shows two old servers that are no longer in the cluster, but when I run pvecm nodes, they are not listed:

pvecm nodes

Membership information
----------------------
    Nodeid      Votes Name
         1          2 bfs (local)

How can I get rid of them from the UI?

 

[dietmar]

remove directory /etc/pve/nodes/<oldnodename>

 

[cdukes]

Thanks! Strangely, it only removed one of the two old nodes (even though I deleted both dirs)

 

[cdukes]

I can't seem to get rid of the other one. Any ideas?

 

[t.lamprecht]

Try restarting the cluster file system:

systemctl restart pve-cluster

If that doesn't work can you please post the output of:

cat /etc/pve/.members
cat /etc/pve/corosync.conf # on PVE >= 4.0
cat /etc/pve/cluster.conf # on PVE < 4.0

 

[cdukes]

/etc/pve/.members

{
"nodename": "bfs",
"version": 3,
"cluster": { "name": "clusterA", "version": 6, "nodes": 2, "quorate": 1 },
"nodelist": {
  "serverA": { "id": 2, "online": 0},
  "serverB": { "id": 1, "online": 1, "ip": "192.168.28.200"}
  }
}

/etc/pve/corosync.conf

logging {
  debug: off
  to_syslog: yes
}

nodelist {
  node {
    name: serverA
    nodeid: 2
    quorum_votes: 1
    ring0_addr: serverA
  }

  node {
    name: serverB
    nodeid: 1
    quorum_votes: 2
    ring0_addr: serverB
  }

}

quorum {
  provider: corosync_votequorum
}

totem {
  cluster_name: clusterA
  config_version: 6
  ip_version: ipv4
  secauth: on
  version: 2
  interface {
    bindnetaddr: 192.168.28.200
    mcastaddr: 224.0.28.1
    ringnumber: 0
  }

}

Is it because of the quorum votes?

pvecm s

Quorum information
------------------
Date:             Tue May  3 06:46:20 2016
Quorum provider:  corosync_votequorum
Nodes:            1
Node ID:          0x00000001
Ring ID:          4816
Quorate:          Yes

Votequorum information
----------------------
Expected votes:   3
Highest expected: 3
Total votes:      2
Quorum:           2
Flags:            Quorate

Membership information
----------------------
    Nodeid      Votes Name
0x00000001          2 192.168.28.200 (local)

I don't have any other servers in this cluster - can I just remove the cluster entirely?

 

[t.lamprecht]

Have you executed

pvecm delnode serverA

after removing the node (physically) from the cluster? It is still in the corosync config which handles the cluster stuff, that's the reason you see it still in the GUI. After the delnode command it will be gone.

I don't have any other servers in this cluster - can I just remove the cluster entirely?

Yes you could (and you then can also recreate a new one later if you want), please look that you execute the commands in the order I post them:

rm /etc/pve/corosync.conf

systemctl stop pve-cluster

systemctl stop corosync
rm /etc/corosync/corosync.conf

systemctl start pve-cluster

 

[cdukes]

The delnode command didn't work, but the others did, thanks!

 

[cdukes]

How do I mark the thread as solved?

 

[t.lamprecht]

The delnode command didn't work, but the others did, thanks!

Strange, no error? But good that you could "de-cluster" your node

How do I mark the thread as solved?

At the top right of the thread (over your first Post) is a "Threadtool" Link/Menu, there you can edit the thread and mark it as solved.

 

[cdukes]

Strange, no error? But good that you could "de-cluster" your node

No error, just returned to the command prompt. Strange indeed
But all is well now. Thanks again!