new node cluster interface keeps going down

[mbj047]

Hi,
Just added a new node to my cluster in which it was a 5.4 upgraded to 6.2 then joined. Joining failed the first attempt so I removed and re-added the second join seems to work better. Cluster network is on a separate interface connected to an unmanaged switch no router with 5 other proxmox server.
To my problem the network interface and bridge that is for the cluster will go down randomly causing it to drop out of the cluster. After ifdown, ifup 6 million times it eventually comes back up, resyncs, and is good for 5 or 6 hours. I just enabled stp on the bridge in hopes that is an issue but I doubt it.

Any help would be appreciated.

 

[Dominic]

I just enabled stp on the bridge in hopes

Did that help anything? Do syslog or corosync log say anything useful?

 

[mbj047]

No to all of those questions. I did a fresh install and tried to add it to the cluster using the same ip and hostname ended up bringing down the cluster. I finally got it to work correctly by giving the node a new ip and hostname then joining it to the cluster. All good now.

 

[ermanishchawla]

Please refrain from adding system with same IP back to existing cluster directly, first remove the configuration of old server from cluster and then add the same