corosync alert [TOTEM ] Invalid packet

[Tanguy Bourdet]

Hello,
I made cluster with 5 Node, and on all node i have massive message in /var/log/daemon.log

All message are the same every second and

Aug 6 12:41:47 pve-1 corosync[1544]: alert [TOTEM ] Invalid packet data
Aug 6 12:41:47 pve-1 corosync[1544]: [TOTEM ] Digest does not match
Aug 6 12:41:47 pve-1 corosync[1544]: [TOTEM ] Received message has invalid digest... ignoring.



Here you have the corosync generated config :

logging {
debug: off
to_syslog: yes
}

nodelist {
node {
name: pve-1
nodeid: 1
quorum_votes: 1
ring0_addr: 10.0.11.1
}
node {
name: pve-2
nodeid: 2
quorum_votes: 1
ring0_addr: 10.0.11.2
}
node {
name: pve-3
nodeid: 3
quorum_votes: 1
ring0_addr: 10.0.11.3
}
node {
name: pve-4
nodeid: 4
quorum_votes: 1
ring0_addr: 10.0.11.4
}
node {
name: pve-5
nodeid: 5
quorum_votes: 1
ring0_addr: 10.0.11.5
}
}

quorum {
provider: corosync_votequorum
}

totem {
cluster_name: SENSIBLE
config_version: 5
interface {
bindnetaddr: 10.0.11.1
ringnumber: 0
}
ip_version: ipv4
secauth: on
version: 2
}




Can you help me ?

Best regards

 

[schinzelh]

I had the same issue today, the pointer for me was in https://forum.proxmox.com/threads/corosync-alerts-and-errors-unicast.43101/#post-207172

The error messages indicates that someone sends messages to the same IP/port ...

I have moved a node from a old cluster to a new one - and it turned out that until i removed the node entry from the old cluster the cluster kept sending traffic to the node, causing the syslog spam.

Removing the node from the old cluster solved it for me.

 

[Tanguy Bourdet]

I have 2 pve-cluster one for old production and one for new. Both have the same cluster name. But not linked.

I'm using new host with fresh install of proxmox.