1 Reply Latest reply: Nov 15, 2012 2:21 PM by robvarga RSS

    GetAll for a partitioned cache

    user644056
      Hi,

      Can somebody point me to the docs on how this is handled?
      In particular invoking getAll for 100 keys on 3 storage nodes what should happen? Expect 3 network calls only; where does the aggregation take place?

      Thank you,
      Horea
        • 1. Re: GetAll for a partitioned cache
          robvarga
          user644056 wrote:
          Hi,

          Can somebody point me to the docs on how this is handled?
          In particular invoking getAll for 100 keys on 3 storage nodes what should happen? Expect 3 network calls only; where does the aggregation take place?

          Thank you,
          Horea
          Hi Horea,

          in the simplest case you should see 2 or 3 Coherence requests sent depending on whether you call getAll on one of these 3 storage nodes or a different node (which obviously does not need a request). They would be sent in parallel, and Coherence would return information from the 3 nodes to the node on which getAll was called which waits for responses from all partitions involved.

          If the cluster is rebalancing because of nodes joining or leaving, the calling node may send additional request(s) to a new node if a partition(s) containing some of the keys have moved after sending the original request.

          Best regards,

          Robert