sailfin
  1. sailfin
  2. SAILFIN-1603

For an umbrella 1599. After 2-4 hours memory grew and reached full heap.

    Details

    • Type: Bug Bug
    • Status: Resolved
    • Priority: Major Major
    • Resolution: Fixed
    • Affects Version/s: 1.0
    • Fix Version/s: milestone 1
    • Component/s: session_replication
    • Labels:
      None
    • Environment:

      Operating System: Solaris X86
      Platform: All

    • Issuezilla Id:
      1,603

      Description

      ******************************************************************************************************

      • Template v0.1 ( 05/01/08 )
      • Sailfin Stress test issue
        ******************************************************************************************************
        Sailfin Build : 61 nightly from 01/30/2009
        Cluster size : 10
        Happens in a single instance (y/n) ? : NA
        Test id : st2_4_presence_subscribe-refresh
        Location of the test : as-telco-sqe/stress-ws/presence
        JDK version : 1.6.0_07
        CLB used : Yes
        HW LB used : No.
        SSR: Enabled

      Started 9 SIPp clients with the following command to effectively get 85cps per
      instance in the 10 instance cluster.
      sipp command :
      sipp -t t1 -sf st2_4_presence_subscribe-refresh-failure.xml -r 95 -m 55555 -d
      585000 -nd -trace_err -trace_screen -buff_size 33554432 -reconnect_close false
      -max_reconnect 10 -reconnect_sleep 3000 <sailfin_host:sip_port>

      *****************************************************************************
      The scenario is as follows :

      1. Start the traffic and let it run till the first round of reSUBSCRIBE completes.
      2. Kill the nodeagent and instance processes (kill -9). Pick one which does not
      have a sipp client connection
      3. Let the traffic continues for about 30 minutes.
      4. Restart the stopped instance. Let the test run for 24x1.

      After about 2 - 4 hours (different time for different instances) since the
      instance was restarted (on step 4), the memory started to grow quickly for all
      instances, except one. And reached full heap. I've attached one gc.log.

        Issue Links

          Activity

          Hide
          easarina added a comment -

          Created an attachment (id=957)
          grep GC gc.log|grep -v ParNew|grep -v Rescan

          Show
          easarina added a comment - Created an attachment (id=957) grep GC gc.log|grep -v ParNew|grep -v Rescan
          Hide
          Scott Oaks added a comment -

          Not seeing in current tests.

          Show
          Scott Oaks added a comment - Not seeing in current tests.

            People

            • Assignee:
              jluehe
              Reporter:
              easarina
            • Votes:
              0 Vote for this issue
              Watchers:
              0 Start watching this issue

              Dates

              • Created:
                Updated:
                Resolved: