with 3node cluster setup deployed on k8s, periodically we are facing issue with server admin credentials

[cpallavi]

Hi Team,

We have a 3 node cluster with couch version 3.2.1 deployed on k8s cluster.

The db was working working totally fine for about 4-6 weeks, post which, we started getting issue, wherein few databases were not getting load with error "This database failed to load" . On further investigation , found out that on clicking on different icons we were getting issue, "you are not a server admin"

Post which, when we checked the /opt/couchdb/etc/local.d/password.ini file, we found out that chttpd_auth and coucdb uuid was missing from password.ini file

The above issue we are facing only with multi node couch cluster.

Can someone please help with this? or how do we make sure, the 3 node cluster is running without any issue!

Thanks pallavi C

[cpallavi]

Any updates on this?

[nickva]

@cpallavi make sure all the nodes in the cluster share the same config file (ini) passwords, server uuids values and cookie secret, and of course Erlang cluster cookies so that nodes can communicate with each other. It sounds like it may be a kube issue, maybe it's not mounting the volume correctly?