Hello Timmy Even when you write and read using quorum, you still don't have isolation. Example:
Client A write "John Doe" to 3 replicas. Since CL = Quorum, the coordinator waits for 2 acks from the replicas before telling client A that the write is successful. Now suppose that between the first ack and second ack, another client B writes "Helen Sue" at quorum. If client A reads the data immediately after the 2nd ack, it would see "Helen Sue" and not "John Doe". Since you don't have isolation, linearization is not possible. The only way to achieve it is to rely on Paxos using lightweight transaction. But even then, you only have linearization on 1 partition only. On Thu, Oct 16, 2014 at 1:45 AM, Timmy Turner <[email protected]> wrote: > Cassandra in general can't provide guarantee any ordering of the executed > queries, since nodes may fail or rejoin the in arbitrary points in time. > > But why can't it provide ordering for queries run at at least the quorum > level? Given that none of the updates get lost, why would order still an > issue? > > Can you maybe illustrate a scenario which shows how/where the order would > get lost if writes and reads always occurred with quorum consistency? >
