If the health checker script is getting errors, then I'd say there are health issues with the cluster.
If you are running unsupported 2.0 code, then it could be any number of things.
I would open a TAC case to see if the cluster can be brought into a healthy state to facilitate upgrading...
Kirk...