[SOLVED] QDevice unable to start, no corosync config generated -- Port 5403/tcp required, not open

[esi_y]

5403 open now, I allowed it through earlier. Guess it didn't take, showing open now.

root@pve:/# pvecm status
Cluster information
-------------------
Name:             TwitchyCluster
Config Version:   17
Transport:        knet
Secure auth:      on

Quorum information
------------------
Date:             Wed Dec 20 13:57:39 2023
Quorum provider:  corosync_votequorum
Nodes:            2
Node ID:          0x00000001
Ring ID:          1.332
Quorate:          Yes

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 192.168.50.6 (local)
0x00000002          1    A,V,NMW 192.168.50.11
0x00000000          1            Qdevice

We have votes!

Yeah I really think this should be in the docs for the QDevice setup specifically, it's different port than for other comms.

 

[Twitchytoes]

Yeah I really think this should be in the docs for the QDevice setup specifically, it's different port than for other comms.

Agreed, I wasn't aware of it. Thank you for your help, I was genuinely stumped on this one.

 

[esi_y]

5403 open now, I allowed it through earlier. Guess it didn't take, showing open now.

PS Do not forget to --permanent that rule.

 

[esi_y]

Agreed, I wasn't aware of it. Thank you for your help, I was genuinely stumped on this one.

And funnily, you do not actually need the other ports on the QDevice at all.

@Maximiliano Could please 5403/tcp be added to the docs (best to the QDevice section itself)?

 

[esi_y]

@Twitchytoes Finally, I would suggest you set thread as solved, maybe add to title "port 5403/tcp was not open" for others in the future.

 

[esi_y]

Took the liberty of filing these two now:
https://bugzilla.proxmox.com/show_bug.cgi?id=5139
https://bugzilla.proxmox.com/show_bug.cgi?id=5140

I suggest affected people +1 themselves in the comments.