Hi Atle, > The paper looks interesting, but could you perhaps give me an outline of the > differences between distributed key-value stores (like www.hazelcast.com), > and distributed shared memory?
I don't know how the distributed key-value store is implemented (though I could hazard a guess), but I do know a little about DSM since I spent 4 years in an office with Savas Parastatidis who was really clued up on it. Briefly, in the model that's swimming round my head, an object can live in any local memory space. If it's referenced by another process in another memory space, it is copied to that space. Should any state changes happen to that object, it becomes the primary and any other copies are invalidated. Rinse and repeat. Jim _______________________________________________ Neo4j mailing list User@lists.neo4j.org https://lists.neo4j.org/mailman/listinfo/user
