• Is this a two node cluster? Just to understand, you're running 4 instances, two per node. If that's the case then you probably would have each server instance at 25% memory. You would really need some resources in this case, as you said.

    How are you planning to rebalance on they fly? You're going to run "sp_configure" after the failover to set "max_server_memory"?

    "select * from sys.dm_os_cluster_nodes" will get you the node names but not if the instance is running on it.