Qpid
  1. Qpid
  2. QPID-3449

0-10 code path hardcodes session timeout to 60 seconds.

    Details

    • Type: Improvement Improvement
    • Status: Resolved
    • Priority: Major Major
    • Resolution: Fixed
    • Affects Version/s: 0.10, 0.11, 0.12
    • Fix Version/s: 0.13
    • Component/s: Java Broker, Java Client
    • Labels:
      None

      Description

      The 0-10 code path has a hardcoded timeout value (Session#timeout) used to enforce a maximum time permitted for Method invocation (Session#invoke) and sync (Session#sync). This value is 60,000ms.

      Users should have the ability to override this setting on both broker and client side in order that they have flexibility to deal with issues such as slow WAN, contended storage etc that may require this parameter to be tuned.

        Activity

        Hide
        Alex Rudyy added a comment -

        Patch changes transport Session to use timeout value specified via JVM settings "qpid.sync_op_timeout" or "amqj.default_syncwrite_timeout.
        This patch enables tests SyncWaitTimeoutDelayTest and SyncWaitDelayTest

        Show
        Alex Rudyy added a comment - Patch changes transport Session to use timeout value specified via JVM settings "qpid.sync_op_timeout" or "amqj.default_syncwrite_timeout. This patch enables tests SyncWaitTimeoutDelayTest and SyncWaitDelayTest
        Hide
        Alex Rudyy added a comment -

        Robbie, could you please review this patch?

        Show
        Alex Rudyy added a comment - Robbie, could you please review this patch?
        Hide
        Robbie Gemmell added a comment -

        Putting this back to Open status, as it builds on the QPID-3448 which needs reworked. Also, the attempt to closing at the TCP level upon timeout might not actually accomplish this as the close method used will try use the connection normally to negotiate closure; this in itself may also timeout.

        Show
        Robbie Gemmell added a comment - Putting this back to Open status, as it builds on the QPID-3448 which needs reworked. Also, the attempt to closing at the TCP level upon timeout might not actually accomplish this as the close method used will try use the connection normally to negotiate closure; this in itself may also timeout.
        Hide
        Alex Rudyy added a comment -

        Robbie, could you please review an updated version of patch?

        A newly introduced by previous patch SessionTimeoutException was removed.
        Also, connection closure was removed as well.

        SyncWaitTimeoutDelayTest and SyncWaitDelatTest were not enabled as SyncWaitTimeoutDelayTest requires more work.

        Show
        Alex Rudyy added a comment - Robbie, could you please review an updated version of patch? A newly introduced by previous patch SessionTimeoutException was removed. Also, connection closure was removed as well. SyncWaitTimeoutDelayTest and SyncWaitDelatTest were not enabled as SyncWaitTimeoutDelayTest requires more work.
        Hide
        Alex Rudyy added a comment -

        Attached a new patch from me and Robbie.

        Show
        Alex Rudyy added a comment - Attached a new patch from me and Robbie.
        Hide
        Robbie Gemmell added a comment -

        Patch applied.

        Show
        Robbie Gemmell added a comment - Patch applied.

          People

          • Assignee:
            Robbie Gemmell
            Reporter:
            Keith Wall
          • Votes:
            0 Vote for this issue
            Watchers:
            0 Start watching this issue

            Dates

            • Created:
              Updated:
              Resolved:

              Development