ConfigServer Community Forum

Just a note that I ran into a problem with certain cluster members were not responding. I double checked their configurations and everything looked good. Restarting LFD wouldn't fix anything. When I ran csf --cping from the master node two of them reported "no reply" even though when I looked at their logs you could see "lfd[16189]: Cluster member <ip> (CA/Canada/<hostname>) said PING!"

It turns out that they were a few CSF versions behind (Upgrading csf from v9.27 to 9.29...) Upgrading all the cluster members to 9.29 fixed the issue and all could talk to each other again.

I'm not sure its a bug or what but I guess its pretty good to make sure that all members in the cluster are upgraded the same version at the same time.

That is completely to be expected since v9.29 is the first version to have responding cluster members, so all members have to be running v9.29+ to respond.

That makes sense. Something else I'm noticing is that cluster member denied reasons are getting cut off. Is that by design?

No, it should not be cutting it off. I'm hoping v9.30 which we're about to release will resolve this.

Just a note that v9.30 fixed this now. Nothing is getting cut off any more