I'm currently trying to run Couchbase Server under CoreOS + Docker (blog post here <http://tleyden.github.io/blog/2014/12/15/running-a-sync-gateway-cluster-under-coreos-on-aws/>), and running into a problem: if I run "fleetctl stop couchbase-server" followed by "fleetctl start couchbase-server", it's losing it's configuration data which is stored in /opt/couchbase, and so it is unable to rejoin the cluster when it comes back up.
I saw that there was a yaml config <https://github.com/brooklyncentral/clocker/blob/master/examples/src/main/assembly/files/blueprints/couchbase.yaml> for Couchbase under Clocker, and wanted to compare to the way it was being done on Clocker to hopefully find a solution. If I run a Couchbase Server cluster node under Clocker, can nodes be restarted and rejoin the cluster? Also, which Dockerfile is using for building the docker image that's used?
