Thanks for your feedback @tracker , if you can please join on GitHub I would be happy to accept new features .
Re-balance partitions when a node is back online should already work :)
You can build a distributed cache on top of it , or any other distributed system :)