Cluster daemons don't stop on one node when using 'ccs --stopall', or when stopping a cluster via the Conga interface in a RHEL 6 High Availability cluster

Issue

• clvmd hangs on one random node when stopping cluster using ccs -h $HOSTNAME --stopall
• The ccs --stopall command never finishes
• LVM commands block after I stop the entire cluster with ccs
• When I use ccs --stop, other nodes lose quorum
• After running the ccs --stopall command, one of the members got the message "Quorum Dissolved", and only the other members stopped successfully.
• After attempting to stop a cluster using the Conga web interface, one of the members got the message "Quorum Dissolved", and only the other members stopped successfully.

Message from syslogd@node1 at Jun  8 08:55:04 ...
 rgmanager[19747]: #1: Quorum Dissolved