addNode.sh failing when trying to extend 11gR1 RAC binaries to new node running 11gR2 cluster

We are practising a cluster upgrade from 11gR1 (11.1.0.7) to 11gR2 (11.2.0.4).  All of our databases will remain at 11.1.0.7 for immediate future. The upgrade went fine. We are exercising delete/add node functionality. There were no issues deleting the node. When adding the node, there were no issues extending the cluster to the new node. Attempting to extend the 11.1.0.7 database home to new node using addNode.sh is failing.

We are running the addNode.sh from the 11.1.0.7 ORACLE_HOME, and the ORACLE_HOME variable is also set to the 11.1.0.7 home.

Using the gui: We are seeing what I think are irregularities. On the "Specify Cluster Nodes to Add to Installation" screen, under the Existing nodes section, the Private Node Name of the existing node is blank. Under the Specify New Nodes section, the new node name  and virtual host names are proper, but the Private node name is not populated. After filling in the Private node name, we get this error: "Error:  Cluster checks failure, the selected remote nodes 'test2' are not accessible". And in the status field: "the virtual hostname test2-vip' for node 'test2' is already in use."