What identifiers does the Cluster host use to identify a VM

[TimRyan]

I have an 8 node cluster with 21 VM's at the moment. The Cluster Host will allocate a noVNC session for all but the three on three new nodes. this is a refusal;

https://192.168.0.112:8006/?console...vmname=ubuserve-1&node=prox-6&resize=off&cmd=

This is an acceptance;

https://192.168.0.112:8006/?console...vmname=w-serve-10&node=prox-5&resize=off&cmd=

This is the syslog traffic from the Host node when this took place

Nov 20 17:25:35 prox-1 pvedaemon[37647]: starting vnc proxy UPIDrox-1:0000930F:00171779:655C070F:vncproxy:116:root@pam:
Nov 20 17:25:35 prox-1 pvedaemon[1172]: <root@pam> starting task UPIDrox-1:0000930F:00171779:655C070F:vncproxy:116:root@pam:
Nov 20 17:25:36 prox-1 pvedaemon[37647]: Failed to run vncproxy.
Nov 20 17:25:36 prox-1 pvedaemon[1172]: <root@pam> end task UPIDrox-1:0000930F:00171779:655C070F:vncproxy:116:root@pam: Failed to run vncproxy.

Nov 20 17:28:50 prox-1 pvedaemon[38097]: starting vnc proxy UPIDrox-1:000094D1:0017634B:655C07D2:vncproxy:115:root@pam:
Nov 20 17:28:50 prox-1 pvedaemon[1173]: <root@pam> starting task UPIDrox-1:000094D1:0017634B:655C07D2:vncproxy:115:root@pam:
Nov 20 17:29:06 prox-1 pvedaemon[1172]: <root@pam> successful auth for user 'root@pam'
Nov 20 17:29:16 prox-1 pvedaemon[1173]: <root@pam> end task UPIDrox-1:000094D1:0017634B:655C07D2:vncproxy:115:root@pam: OK

What is the Host prox-1 finding that leads it to refuse the noVNC session for one but not the other, and how might I fix this?

This is the qm config dump from the failed VM

root@prox-6:~# qm config 116
boot: order=scsi0;ide2;net0
cores: 4
cpu: x86-64-v2-AES
description: VM116 Available at 192.168.0.158 ubuserver-1
ide2: local:iso/ubuntu-22.04.3-live-server-amd64.iso,media=cdrom,size=2083390K
memory: 12288
meta: creation-qemu=8.1.2,ctime=1700432813
name: ubuserve-1
net0: virtio=EE:7D:00:1E:3A:7F,bridge=vmbr0,firewall=1
numa: 0
onboot: 1
ostype: l26
scsi0: local-lvm:vm-116-disk-0,iothread=1,size=1T
scsihw: virtio-scsi-single
smbios1: uuid=b4c7fa76-75e5-43f3-8e26-44713b424e00
sockets: 1
vmgenid: d9892f1c-2ad7-4edd-a204-e3615682a457

Is the vmgenid the term that the Host ID's with?

 

[LnxBil]

In general, the only primary key (if you're familiar with databases) is the VMID. It identifies and references eveything.

What is the Host prox-1 finding that leads it to refuse the noVNC session for one but not the other, and how might I fix this?

Please check the SSH connection from and to all nodes. This may be the problem.

 

[TimRyan]

I am able to from my windows workstation 192.168.0.144, on the same subnet 192.168.0.1/24

SSH to VM118, then on that shell SSH to the prox-1 host at 192.168.0.112 and from that shell SSH back to the VM118 at 192.168.0.158 without issue