I have exactly the same question.
I need load balancing for 1 queue on different servers
Yes. Clustering is in the plans for 4.0.
Since 4.0 is behind the original schedule, a singleton form of clustering was implemented in 3.2.3.
Question - ins't this solvable by simply having two consumers (MDBs) for the same queue deployed on two different nodes?
(Transactions can be used for HA).
What exactly does JMS clustering mean?