Use PBS as a two-node cluster quorum member ?

Fabio

Renowned Member
Jan 30, 2016
63
1
73
52
I read that you can configure PBS as the third quorum node in a two-node cluster with replicas. I set up a two node cluster and VM replicas. Now I put a proxmox backup server and connected it to my cluster. Backups work and are scheduled. I would like to use the PBX as a quorum element in order to check if one of the two nodes fails, and ensure that the VM is passed to the active node. Can you help me please?
 
forgive me if I try to install qdevice on my PBS it gives me an error: root@pbs:~# apt install corosync-qnetd
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
E: Unable to locate package corosync-qnetd
root@pbs:~# apt install corosync-qdevice
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
E: Unable to locate package corosync-qdevice
root@pbs:~#
I'm not understanding how to proceed, can you help me?
 
I tried again. My source.list file is:

-------------------------------------------------- -------------------------------------------------- -----------
deb http://ftp.it.debian.org/debian bookworm main contrib

deb http://ftp.it.debian.org/debian bookworm-updates main contrib

# security updates
deb http://security.debian.org bookworm-security main contrib
-------------------------------------------------- -------------------------------------------------- -----------
I tried again to put the package with apt install corosync-qnetd and it downloaded and installed it.
 
Ok it seems to work. I'll summarize what I did:

put on PBS: apt install corosync-qnetd
put in the two nodes: apt install corosync-qdevice

From one of the two nodes I configured qdevice in the following way: pvecm qdevice setup 192.168.5.30

then with pve status I see:

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

Membership information
----------------------
Nodeid Votes Qdevice Name
0x00000001 1 A,V,NMW 192.168.5.10
0x00000002 1 A,V,NMW 192.168.5.20 (local)
0x00000000 1 Qdevice
root@node2:~#

I did a test by disconnecting the cables of the two network cards of node 2, the system migrated the two virtual machines to node 1. It worked, but it took 2 minutes and 49 seconds.

Is the time it took you right?
Consider that the nodes have active VM replicas