Clovertech

process terminates – hcicmd -p xxx -c ". output_cycle&q

› Clovertech Forums › Read Only Archives › Cloverleaf › Cloverleaf › process terminates – hcicmd -p xxx -c ". output_cycle&q

  • Creator
    Topic
  • May 18, 2010 at 4:18 pm #51764
    Lawrence Williams
    Participant

      AIX 5.3

      Cloverleaf 5.7

      The automatic log cycling feature has not worked in a consistent manner (it doesn’t cycle the logfile when the size reaches the set amount)…..so, I have written a script that checks the logfile sizes and if the size is over 10MB runs the command to cycle the logfile manually………..”hcicmd -p $process -c “. output_cycle”

      I have now seen a couple instances in which the process terminates after the logfile is cycled, anyone have any suggestions ?

      Here is an excerpt from the logfile for the process that crashed (omis2):

      [prod:prod:INFO/0:

    • Creator
      Topic
    Viewing 10 reply threads
    • Author
      Replies
      • May 20, 2010 at 7:53 pm #71677
        Lawrence Williams
        Participant

          bump

        • May 24, 2010 at 9:06 pm #71678
          Lawrence Williams
          Participant

            no one else has seen this sort of problem ?

          • May 25, 2010 at 3:18 pm #71679
            Michael Hertel
            Participant

              See this:

              https://usspvlclovertch2.infor.com/viewtopic.php?t=3213

            • May 25, 2010 at 3:26 pm #71680
              Lawrence Williams
              Participant

                Thanks Michael, I had already read over and even posted to that thread.

                I have disabled the automatic log cycling to try and rule it out as the problem, however I still get the same problem (the process terminates).

                I will be installing the REV2 patch soon (hopefully within the next week or so) and I am hoping it may help with this problem.

              • May 25, 2010 at 3:33 pm #71681
                Michael Hertel
                Participant

                  Sorry, didn’t see your previous post.

                  The snippet you’ve provided is interesting. It looks like a normal shutdown is occuring. You said “crash” earlier. Does it appear to go thru a normal shutdown in the process log or are you getting a panic?

                • May 25, 2010 at 3:43 pm #71682
                  Lawrence Williams
                  Participant

                    No panic…. the process seems to cycle the logfile and restart and then I’ll see a few lines of INFO in the new logfile (i.e.  “Checking for leaked handles in General interpreter…..”) and then I get the following:

                    [cmd :cmd :INFO/0:    omis2_cmd:05/25/2010 09:30:26] Removing engine command port file.

                    [prod:prod:INFO/0:  STARTUP_TID:05/25/2010 09:30:26] Removing pid file.

                    [prod:prod:INFO/0:  STARTUP_TID:05/25/2010 09:30:26] Engine process 1400964 is terminating

                    [prod:prod:INFO/0:  STARTUP_TID:05/25/2010 09:30:26] Ended at Tue May 25 09:30:26 2010

                  • May 25, 2010 at 3:58 pm #71683
                    Michael Hertel
                    Participant

                      Is there something like the following also in your log?

                      [cmd :cmd :INFO/0:   others_cmd:05/25/2010 08:53:16] Receiving a command

                      [cmd :cmd :INFO/0:   others_cmd:05/25/2010 08:53:16] Receiving a command

                      [cmd :cmd :INFO/0:   others_cmd:05/25/2010 08:53:16] Received command: ‘. die’

                      [cmd :cmd :INFO/0:   others_cmd:05/25/2010 08:53:16] Doing ‘die’ command with args ”

                      [prod:prod:INFO/0: others_xlate:05/25/2010 08:53:16] Checking for leaked handles in the Xlate interpreter…

                    • May 25, 2010 at 4:26 pm #71684
                      Lawrence Williams
                      Participant

                        I don’t see anything that matches that specifically…. I am attaching the logfile with the termination line.

                      • May 25, 2010 at 4:35 pm #71685
                        Michael Hertel
                        Participant

                          Could you also post the .old version of the log?

                          I’d also like to see the log with the cycle command executing.

                          Thanks

                        • May 25, 2010 at 4:50 pm #71686
                          Lawrence Williams
                          Participant

                            Here’s the originating logfile……I’ve taken out all the message related logging (patient information).

                          • May 25, 2010 at 6:18 pm #71687
                            Michael Hertel
                            Participant

                              Hmmmm, no clues there.

                              I’d try turning up the EO by going into Netconfig>process>configure, select the process and change EO default to enable_all.

                              Then rerun the test and look at both logs again.

                              One other thing, the previous topic suggested compression, could you post both new logs and your siteInfo file? Even if you’ve disabled autocycling, I’m thinking that if you do have something defined for compression, that maybe that is getting in the way. But I’m fishing.

                              Again, I find it interesting that it is going through a normal shutdown rather than a panic.

                          • Author
                            Replies
                          Viewing 10 reply threads
                          • The forum ‘Cloverleaf’ is closed to new topics and replies.