quorum failed - CLM CONFIGURATION CHANGE [help!]

[trv]

Hi. I need help. I'm fighting with quorum last week and can't do anything.
I have 3 nodes. 2x nodes is ok, but 3rd one can't sync with others

[FONT=Droid Sans Mono]root@blackcap:~# pvecm nodes[/FONT]Node  Sts   Inc   Joined               Name
   1   M  2878936   2015-06-16 18:14:43  blackcap
   4   X      0                        phoenix
   5   X      0                        sparrow

root@blackcap:~# tail -f /var/log/syslog
Jun 16 18:16:03 blackcap corosync[212860]:   [CLM   ] Members Left:
Jun 16 18:16:03 blackcap corosync[212860]:   [CLM   ] Members Joined:
Jun 16 18:16:03 blackcap corosync[212860]:   [CLM   ] CLM CONFIGURATION CHANGE
Jun 16 18:16:03 blackcap corosync[212860]:   [CLM   ] New Configuration:
Jun 16 18:16:03 blackcap corosync[212860]:   [CLM   ] #011r(0) ip(192.168.30.58)
Jun 16 18:16:03 blackcap corosync[212860]:   [CLM   ] Members Left:
Jun 16 18:16:03 blackcap corosync[212860]:   [CLM   ] Members Joined:
Jun 16 18:16:03 blackcap corosync[212860]:   [TOTEM ] A processor joined or left the membership and a new membership was formed.
Jun 16 18:16:03 blackcap corosync[212860]:   [CPG   ] chosen downlist: sender r(0) ip(192.168.30.58) ; members(old:1 left:0) [FONT=Droid Sans Mono]Jun 16 18:16:03 blackcap corosync[212860]:   [MAIN  ] Completed service synchronization, ready to provide service.[/FONT]

[FONT=Droid Sans Mono]root@phoenix:~# pvecm nodes[/FONT]Node  Sts   Inc   Joined               Name
   1   X      0                        blackcap
   4   M  2621688   2015-06-16 17:21:11  phoenix [FONT=Droid Sans Mono]   5   X  2621700                        sparrow[/FONT]

[FONT=Droid Sans Mono]root@phoenix:~# cat /etc/pve/cluster.conf[/FONT]<?xml version="1.0"?>
<cluster name="birds1" config_version="9">

  <cman keyfile="/var/lib/pve-cluster/corosync.authkey">
  </cman>

  <clusternodes>


  <clusternode name="phoenix" votes="1" nodeid="4"/>
  <clusternode name="sparrow" votes="1" nodeid="5"/>
<clusternode name="blackcap" votes="1" nodeid="1"/></clusternodes>

</cluster>

I do many things, but no luck to restore normal working of cluster.
Please help.
proxmox 3.4

 

[spirit]

Hi,

seem to be a multicast problem.

Do you have already read this wiki ?

https://pve.proxmox.com/wiki/Multicast_notes

(about multicast, igmp, igmp snooping).

Also try to test the multicast with omping (howto in the this wiki)