Replication fail over on SQL (2005 sp3) cluster

Hi there,
We have transactional replication running on a SQL cluster (2005 sp3) which failed over to the secondary node the other day. The fail over seemed to work fine and the db was still available to the application; however, replication could no longer get from the distributor to the subscriber. Once we failed the node back over to the primary node replication started working again immediately.
Just wondering if anyone may know why replication would lose connectivity when it failed over. I would have thought that it would have just kept 'trucking along'.
Thanks,
Jeff