Clovertech

Thread keep up and OB msg queued

› Clovertech Forums › Read Only Archives › Cloverleaf › Cloverleaf › Thread keep up and OB msg queued

  • Creator
    Topic
  • May 18, 2007 at 11:48 am #49289
    Hongle Sun
    Participant

      Hello, we have found one out bound thread showm status of UP, but the all the outbound message queued up, not sending any message, the thread set up with outbound only , await replay , replyy ‘generation’ with resend_ob_msg in 30 senconds, what should i look into? Thanks

    • Creator
      Topic
    Viewing 8 reply threads
    • Author
      Replies
      • May 18, 2007 at 11:55 am #61380
        James Cobane
        Participant

          Hongle Sun,

          Take a look at the “Last Read” time on the thread compared to the “Last Write”; are you actually receiving any ACK’s back from the receiving system?  I’m assuming (based on your description) that you have the ‘Timeout’ set to 30 for the ‘Await Replies’ and that it is not set to wait forever (-1).  It seems like you’re likely not getting any ACK’s back from the receiving system even though you are connected (status = “UP”).

          Hope this helps.

          Jim Cobane

          Henry Ford Health

        • May 18, 2007 at 12:01 pm #61381
          Hongle Sun
          Participant

            Thanks Jim for quickly responsing. I just got notified that the OUTbound system was recycling at that time, but my question is, with my above setting in place, the thread should be able to reconnect to foreign system as it restores, start sending outbound message, am i right?

            Thanks

          • May 18, 2007 at 12:08 pm #61382
            James Cobane
            Participant

              By default, the engine trys to reconnect every 5 seconds for a TCP/IP connection (this is in the “Auto-Reconnect Options”, “Re-open time” in the protocol settings), so it will keep trying to re-connect.  As long as the receiving system re-starts its communications when it comes back up, Cloverleaf should re-connect without a problem.

              Jim Cobane

              Henry Ford Health

            • May 18, 2007 at 12:25 pm #61383
              Hongle Sun
              Participant

                I check the thread status, it tells me the las receivd and las wt are the same, my interperution for this would be we dint get the ACK from foreign system and the thread keep sending outbound msg in 30 seconds while the OB msg queued. My question is when engine didnt get ACK from foreign system, should the protocol status be UP or openning? correct me if i am wrong, thanks again

              • May 18, 2007 at 1:20 pm #61384
                Hongle Sun
                Participant

                  James, Can i assume the resend_ob_msg not work as expected?

                  Thanks

                • May 18, 2007 at 6:16 pm #61385
                  James Cobane
                  Participant

                    Check to see if the responses you are getting back are ACK’s (i.e. AA) vs. NAK’s (AE or AR).  It’s possible that you are getting the AE/AR type responses, so the engine keeps resending the message (depending on the what the proc that is handling the reply is telling the engine to do).  I’m suspecting that you are getting AE or AR repsonses; not AA’s.

                    Jim Cobane

                    Henry Ford Health

                  • May 28, 2007 at 5:34 pm #61386
                    Hongle Sun
                    Participant

                      Thanks James, we found it is interface in down system newly updated was not sending us ack. Sine the case in good hand, one more question is, with my current resend_ob_msg applied. for example if the down system was recyling and never get a chance to send us an ack. I assumed cloverleaf will keep sending ob msg within time expried interval. My question is the outbound message log file will grow biger and have a lot of redudance msges in it, at this case, whatesle or tcl proc i can use to not gernerate those repeating msges in log file? Thanks

                    • May 29, 2007 at 5:20 pm #61387
                      Russ Ross
                      Participant

                        Our catch all solution to keep this type of situation from getting out of hand is to use an alert to check opque.

                        We configure cloverleaf to issue a level one alert notifying foreign system owners that they have a problem when opque > 200 messages.

                        We configure cloverleaf to issue a level two alert notifying our integration team when opque > 2,000 messages.

                        After we receive a level two alert, we know the foreign system owners are not doing their job in a timely fashion for some reason.

                        Then the on-call person knows to follow up and possibly put the outbound thread in down time mode.

                        Refer to this post for a description of how we handle putting a thread in down time mode:

                        https://usspvlclovertch2.infor.com/viewtopic.php?t=1621” class=”bbcode_url”>https://usspvlclovertch2.infor.com/viewtopic.php?t=1621

                        Russ Ross
                        RussRoss318@gmail.com

                      • June 1, 2007 at 3:10 pm #61388
                        Gena Gill
                        Participant

                          I just had the same problem, except I was sending messages from System A to Cloverleaf Engine A to Cloverleaf Engine B, then on to System B, which is a logistical nightmare, but that’s another story.  I’d forgotten to set the receiving thread on Engine B to send the ACK back.  Once I did that, the messages started going properly.

                      • Author
                        Replies
                      Viewing 8 reply threads
                      • The forum ‘Cloverleaf’ is closed to new topics and replies.