sailfin
  1. sailfin
  2. SAILFIN-2017

High session loss (SipApplicationSession unexpectedly null) observed in the converged app system test

    Details

    • Type: Bug Bug
    • Status: Resolved
    • Priority: Major Major
    • Resolution: Incomplete
    • Affects Version/s: 2.0
    • Fix Version/s: milestone 1
    • Component/s: session_replication
    • Labels:
      None
    • Environment:

      Operating System: All
      Platform: All

      Description

      Build 31d
      Parent Issue - 1983

      360 "SipApplicationSession unexpectedly null" messages in the server logs
      indicate session loss. These can be attributed to loss of replication messages
      due to the instance failure. Considering a 1 second window of vulnerability this
      value should not be higher than 74 sessions (37 SIP + 37 HTTP cps).

      See issue 2016 for logs.

        Issue Links

          Activity

          Hide
          varunrupela added a comment -

          updated dependency and added keyword

          Show
          varunrupela added a comment - updated dependency and added keyword
          Hide
          Mahesh Kannan added a comment -

          As mentioned in 2016, it is possible that the use of single timer thread to
          manage three artifacts (SS, SAS and HTTP) could lead to more 75 such messages.

          I think this is not a show stopper.

          Show
          Mahesh Kannan added a comment - As mentioned in 2016, it is possible that the use of single timer thread to manage three artifacts (SS, SAS and HTTP) could lead to more 75 such messages. I think this is not a show stopper.
          Hide
          varunrupela added a comment -

          From the last run with default FLUSH_INTERVAL_FOR_REMOVAL_MILLIS setting, the
          number of "unexpectedly null" messages in the cluster is 62. There are 62
          matching 481 responses in the cluster.

          The number is small enough to be within the exit criteria and hence closing the
          issue for this soft failure.

          The total active sessions used for comparison is call-rate * call-duration =
          37*~90= ~3330 sessions.

          Show
          varunrupela added a comment - From the last run with default FLUSH_INTERVAL_FOR_REMOVAL_MILLIS setting, the number of "unexpectedly null" messages in the cluster is 62. There are 62 matching 481 responses in the cluster. The number is small enough to be within the exit criteria and hence closing the issue for this soft failure. The total active sessions used for comparison is call-rate * call-duration = 37*~90= ~3330 sessions.

            People

            • Assignee:
              Mahesh Kannan
              Reporter:
              varunrupela
            • Votes:
              0 Vote for this issue
              Watchers:
              0 Start watching this issue

              Dates

              • Created:
                Updated:
                Resolved: