[SOLVED] Qdevice not voting

[sg29]

Hi,

After a powerloss this weekend with a power supply fail, i noticied this morning that my HA has not worked as expected.

i think this is due to my Qdevice :

Quorum information
------------------
Date:             Mon May  2 11:42:21 2022
Quorum provider:  corosync_votequorum
Nodes:            2
Node ID:          0x00000002
Ring ID:          2.9b9
Quorate:          Yes


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


Membership information
----------------------
Nodeid      Votes    Qdevice Name
0x00000002          1   A,NV,NMW 192.168.51.8 (local)
0x00000003          1   A,NV,NMW 192.168.51.10
0x00000000          0            Qdevice (votes 1)

I have tried to remove qdevice and adding it again, no change.

How could i fix that ?

Thank you.

 

[oguz]

hi,

are the corosync services on your qdevice working as expected?

do you see any errors in journals related to this issue?

 

[sg29]

on my qdevice node :

corosync.service is active (running)
corosync-qnetd.service isactive (running)
corosync-qdevice.service failed with that error :

  L'unité (unit) corosync-qdevice.service a commencé à démarrer.
mai 02 14:11:44 PXQ corosync-qdevice[21143]: Can't read quorum.device.model cmap key.
mai 02 14:11:44 PXQ systemd[1]: corosync-qdevice.service: Main process exited, code=exited, status=1/FAILURE
 Subject: Unit process exited
 Defined-By: systemd
  Support: https://www.debian.org/support

and corosync-quorumtool give this (not sur i understand well, only one of the 2 node detected ?):

Quorum information
------------------
Date:             Mon May  2 14:18:38 2022
Quorum provider:  corosync_votequorum
Nodes:            1
Node ID:          1
Ring ID:          1.f
Quorate:          Yes

Votequorum information
----------------------
Expected votes:   1
Highest expected: 1
Total votes:      1
Quorum:           1
Flags:            Quorate Qdevice

Membership information
----------------------
    Nodeid      Votes    Qdevice Name
         1          1  NA,NV,NMW node1 (local)
         0          0            Qdevice (votes 0)

Other question is : is that normal that pvecm status give NV for all node in my first post ?

 

[oguz]

is SSH working correctly between the nodes and the qdevice?

which distribution is your qdevice running?

 

[sg29]

Finaly managet to get it work.

Make a fresh install of my Qdevice (debian 11 for rpi3B)

apt install corosync-qnetd
apt install corosync-qdevice

nano /etc/ssh/sshd_config
PermitRootLogin yes
systemctl restart ssh

And on my two nodes :
ssh-keygen -f "/root/.ssh/known_hosts" -R "x.x.x.x" (qdevice ip)
rm -R /etc/corosync/qdevice/net/nssdb/
rm -R /etc/corosync/qnetd/nssdb/

pvecm qdevice setup x.x.x.x --force
(got some corosync error at this point but systemctl and journalctl -xe where all green)
pvecm qdevice remove
pvecm qdevice setup x.x.x.x --force


and then :

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

Membership information
----------------------
    Nodeid      Votes    Qdevice Name
0x00000002          1    A,V,NMW 192.168.51.8 (local)
0x00000003          1    A,V,NMW 192.168.51.10
0x00000000          1            Qdevice