[SOLVED] Corosync wrong bindnetaddr

[vRod]

Hi all,

I have a proxmox cluster consisting of 2 hosts (pmx01, pmx02). Cluster name is cluster01 and consisted of 2 other hosts which are now gone (pve01, pve02).

Now, so far most things seems to be working right (except for some qlge driver port flapping) but I have noticed some recent retransmit totem errors. when looking in the corosync.conf, i see the ip from the old pve02 as bindnetaddr:

root@pmx02:~# cat /etc/corosync/corosync.conf
logging {
  debug: off
  to_syslog: yes
}

nodelist {
  node {
    name: pmx01
    nodeid: 1
    quorum_votes: 1
    ring0_addr: 10.72.0.101
  }
  node {
    name: pmx02
    nodeid: 2
    quorum_votes: 1
    ring0_addr: 10.72.0.102
  }
}

quorum {
  provider: corosync_votequorum
}

totem {
  cluster_name: cluster01
  config_version: 12
  interface {
    bindnetaddr: 10.72.0.201
    ringnumber: 0
  }
  ip_version: ipv4
  secauth: on
  version: 2
}

I have tried to change the bindnetaddr to the ip of pmx01 and altered the config_version, but it changes back to .201.

Does anyone know how to fix this, or if it even needs to be fixed?

 

[Chris]

Hi,
have you tried to follow the procedure mentioned in the docs about how to change the corosync config: https://pve.proxmox.com/pve-docs/pve-admin-guide.html#_corosync_configuration

 

[vRod]

Hi,
have you tried to follow the procedure mentioned in the docs about how to change the corosync config: https://pve.proxmox.com/pve-docs/pve-admin-guide.html#_corosync_configuration

Thank you, I had edited the /etc/corosync/corosync.conf file! Now that i changed the one in pve, it looks like it worked!

just out of curiosity, was this really important to change back? I mean, the cluster looks to have worked nevertheless, I just noticed that it had the wrong address randomly.

 

[Chris]

No worries, as long as the bindnetaddr is an address form the subnet configured on the network interface you should be fine.