Is anybody in Chronicle Map users community needs a strongly consistent cluster of 100% redundant Chronicle Maps (no partitioning) with synchronous operation-based replication, e. g. map.put() is successful only after the operation is applied to the quorum of cluster nodes? (While replication is logically synchronous, in Java API it might be asynchronous, returning CompletableFuture from the map.put() method).
If somebody is interested, please identify youself -- I would like to try to implement this using the Atomix project:
http://atomix.io/
Write throughput would obviously be not even close to the existing Chronicle Map replication (open source in Chronicle Map 2 and closed in 3), but cluster guarantees are stronger, cluster is more reliable under failure conditions and much easier to bootrstrap and operate.