[PVE-User] Cluster won't reform so I can't restart VMs

[Uwe Sauter]

If it is a multicast problem and your cluster is not that big (~10 nodes) you could switch to using "udpu" in corosync.conf

totem {
[…]
  config_version: +=1 ######### increment with every change you do
  transport: udpu
}




Am 11.08.2017 um 13:48 schrieb Alexandre DERUMIER:
> seem to be a multicast problem.
> 
> does it work with omping ?
> 
> 
> ----- Mail original -----
> De: "Chris Tomkins" <christ at brandwatch.com>
> À: "proxmoxve" <pve-user at pve.proxmox.com>
> Envoyé: Vendredi 11 Août 2017 12:02:00
> Objet: [PVE-User] Cluster won't reform so I can't restart VMs
> 
> Hi Proxmox users, 
> 
> I have a 4 node cluster. It has been in production for a few months with 
> few/no issues. 
> 
> This morning one of my admins reported that each node appeared isolated 
> ("Total votes: 1"). All VMs were up and unaffected. Unfortunately I 
> made the mistake of stopping VMs on 3 of the nodes to apply updates and 
> reboot as I assumed this would clear the issue. Now the scenario remains 
> the same but the VMs are down on 3 of the nodes and it won't allow me to 
> start them as I have no quorum. 
> 
> No config changes were made and this cluster was fine and had quorum last 
> time I looked (last week). 
> 
> I don't want to take the wrong action and make this worse - any advice 
> would be greatly appreciated! 
> 
> hypervisors ar1406/ar1600/ar1601 are up to date and have been rebooted this 
> morning. ar1407 has not been rebooted or updated (yet) as the VMs on it are 
> critical. 
> 
> Thanks, 
> 
> Chris 
> 
> [LIVE]root at ar1406:~# for i in ar1406 ar1407 ar1600 ar1601; do ssh $i 'cat 
> /etc/pve/.members'; done 
> { 
> "nodename": "ar1406", 
> "version": 3, 
> "cluster": { "name": "netteamcluster", "version": 4, "nodes": 4, "quorate": 
> 0 }, 
> "nodelist": { 
> "ar1407": { "id": 2, "online": 0}, 
> "ar1601": { "id": 3, "online": 0}, 
> "ar1600": { "id": 4, "online": 0}, 
> "ar1406": { "id": 1, "online": 1, "ip": "10.0.6.201"} 
> } 
> } 
> { 
> "nodename": "ar1407", 
> "version": 3, 
> "cluster": { "name": "netteamcluster", "version": 4, "nodes": 4, "quorate": 
> 0 }, 
> "nodelist": { 
> "ar1407": { "id": 2, "online": 1, "ip": "10.0.6.202"}, 
> "ar1601": { "id": 3, "online": 0}, 
> "ar1600": { "id": 4, "online": 0}, 
> "ar1406": { "id": 1, "online": 0} 
> } 
> } 
> { 
> "nodename": "ar1600", 
> "version": 3, 
> "cluster": { "name": "netteamcluster", "version": 4, "nodes": 4, "quorate": 
> 0 }, 
> "nodelist": { 
> "ar1407": { "id": 2, "online": 0}, 
> "ar1601": { "id": 3, "online": 0}, 
> "ar1600": { "id": 4, "online": 1, "ip": "10.0.6.203"}, 
> "ar1406": { "id": 1, "online": 0} 
> } 
> } 
> { 
> "nodename": "ar1601", 
> "version": 3, 
> "cluster": { "name": "netteamcluster", "version": 4, "nodes": 4, "quorate": 
> 0 }, 
> "nodelist": { 
> "ar1407": { "id": 2, "online": 0}, 
> "ar1601": { "id": 3, "online": 1, "ip": "10.0.6.204"}, 
> "ar1600": { "id": 4, "online": 0}, 
> "ar1406": { "id": 1, "online": 0} 
> } 
> } 
>