Couchbase Server

restarting couchbase-server when node is out of disk space results in having two beam.smp processes because old one refuses to shut down ( happens in key-value + xdcr use case on source cluster)