@dheath

my feeling at the moment is we have a misunderstanding of clustering rather than a cluster issue, seems a bit early to start advocating rebuilding it. 🙂

@peacesells

SQLAgent jobs are held in msdb and there is only one msdb per instance. You seem fixated on the nodes, is this a failover cluster with just one instance, or do you have two instances, one active on each node? (An active\active cluster)