My problem is that on a single node of the cluster I can mount a GFS2 volume, however as soon as I try to write to the volume, access to GFS2 freezes on all nodes (Simple ls hangs even). The hang finally clears up with the original two nodes regaining access, but the third node vomits errors all over (See below from syslog). The other two nodes can read and write to GFS2 just fine until this node joins the cluster and tries to write. In my digging I found this thread:
But that sounds different enough that I don't think it is my issue.
Syslog bits from one of the "good" nodes: