Loading ...
Sorry, an error occurred while loading the content.

Re: [BPQ32] Update - BPQ32 & RMS Packet failure

Expand Messages
  • WS7I
    I am seeing that message as well, activity timeout-disconnecting on the RF side, I think that it is important to tune the various packet parameters a bit. I
    Message 1 of 8 , Oct 8, 2008
    • 0 Attachment
      I am seeing that message as well, "activity timeout-disconnecting" on the RF side, I think that it is important to tune the various packet parameters a bit.  I haven't done this yet but will be playing with it this weekend.
       
      I also have restricted the BBS value so that it doesn't propogate as I found that when they are doing nearby node connects on RF it might well add another layer on Level 2.  Need to examine this and make some changes.  MaxFrames on my setup might well be set too high.
       
      Jay

      On Wed, Oct 8, 2008 at 5:59 AM, Jeff Thomas - WA4ZKO <wa4zko@...> wrote:

      I'm also noticing that often when I'm pulling a large (say 50k)
      message in from RMS Packet via the BPQ switch on RF that I don't get
      all the message. It looks like it cuts off prematurely with a message
      saying along the lines of "Activity timeout - disconnecting". The RF
      connection doesn't get stuck in RMS like it often does when I'm
      pulling in a message on RF and things glitch (as previously described).

      Almost like AGWtoBPQ and RMS Packet just don't hold the session open
      long enough on a long transfer and something times out prematurely???
      Just a big guess though.

      I've never seen this happen on a telnet connect into the switch and on
      into RMS. Maybe the telnet connect pulls the message in fast enough
      that the issue doesn't manifest itself? starting to think the problem
      is in AGWtoBPQ, but who knows...the RMS software is pretty beta itself.

      For those that haven't "stress" tested their systems with messages of
      a few dozen kb, might want to test things.

      Jeff
      WA4ZKO




      --
      Jay WS7I

    • John Wiseman
      Hi Jeff, This could be a flow control problem between RMS Packet, AGWtoBPQ and BPQ32. Could you check the BPQ stats? If it shows a low current buffer count
      Message 2 of 8 , Oct 8, 2008
      • 0 Attachment
        Hi Jeff,
         
        This could be a flow control problem between RMS Packet, AGWtoBPQ and BPQ32. Could you check the BPQ stats? If it shows a low current buffer count during a transfer, or a low minimum buffers, even if the current count recovers, then it needs investigating.
         
        73, John G8BPQ
         
        -----Original Message-----
        From: BPQ32@yahoogroups.com [mailto:BPQ32@yahoogroups.com]On Behalf Of Jeff Thomas - WA4ZKO
        Sent: 08 October 2008 14:00
        To: BPQ32@yahoogroups.com
        Subject: [BPQ32] Update - BPQ32 & RMS Packet failures

        I'm also noticing that often when I'm pulling a large (say 50k)
        message in from RMS Packet via the BPQ switch on RF that I don't get
        all the message. It looks like it cuts off prematurely with a message
        saying along the lines of "Activity timeout - disconnecting" . The RF
        connection doesn't get stuck in RMS like it often does when I'm
        pulling in a message on RF and things glitch (as previously described).

        Almost like AGWtoBPQ and RMS Packet just don't hold the session open
        long enough on a long transfer and something times out prematurely? ??
        Just a big guess though.

        I've never seen this happen on a telnet connect into the switch and on
        into RMS. Maybe the telnet connect pulls the message in fast enough
        that the issue doesn't manifest itself? starting to think the problem
        is in AGWtoBPQ, but who knows...the RMS software is pretty beta itself.

        For those that haven't "stress" tested their systems with messages of
        a few dozen kb, might want to test things.

        Jeff
        WA4ZKO

      • Jeff Thomas - WA4ZKO
        Jay, Thanks! Now I feel better that someone else is seeing similar. Be real curious to see what your weekend testing reveals, particularly on the high speed rf
        Message 3 of 8 , Oct 8, 2008
        • 0 Attachment
          Jay,

          Thanks! Now I feel better that someone else is seeing similar. Be real
          curious to see what your weekend testing reveals, particularly on the
          high speed rf ports you've got.

          If you can, shove some multi-kb (i've been using some 50k and 100k
          test msgs) into things via RF and pull the same back out on RF and see
          if things trip up on you after a few messages.

          I've been tempted to tweak things better for 9.6k, but I was worried
          that slamming even more data at it faster and in larger frames might
          aggravate things. The defaults look geared toward 1200 baud, which is
          sensible. Figured I'd refrain from perf' tuning till after the bugs
          are worked out using fairly conservative values. Maxframe and paclen
          clearly could use tweaking for the higher rf link speeds.

          Jeff
          WA4ZKO

          PS - Great to see more folks at least testing out the higher speeds
          for a packet application that can clearly benefit from them.

          --- In BPQ32@yahoogroups.com, WS7I <ws7ik7tj@...> wrote:
          >
          > I am seeing that message as well, "activity timeout-disconnecting"
          on the RF
          > side, I think that it is important to tune the various packet
          parameters a
          > bit. I haven't done this yet but will be playing with it this weekend.
          >
          > I also have restricted the BBS value so that it doesn't propogate as
          I found
          > that when they are doing nearby node connects on RF it might well add
          > another layer on Level 2. Need to examine this and make some changes.
          > MaxFrames on my setup might well be set too high.
          >
          > Jay
          >
          > On Wed, Oct 8, 2008 at 5:59 AM, Jeff Thomas - WA4ZKO <wa4zko@...>wrote:
          >
          > > I'm also noticing that often when I'm pulling a large (say 50k)
          > > message in from RMS Packet via the BPQ switch on RF that I don't get
          > > all the message. It looks like it cuts off prematurely with a message
          > > saying along the lines of "Activity timeout - disconnecting". The RF
          > > connection doesn't get stuck in RMS like it often does when I'm
          > > pulling in a message on RF and things glitch (as previously
          described).
          > >
          > > Almost like AGWtoBPQ and RMS Packet just don't hold the session open
          > > long enough on a long transfer and something times out prematurely???
          > > Just a big guess though.
          > >
          > > I've never seen this happen on a telnet connect into the switch and on
          > > into RMS. Maybe the telnet connect pulls the message in fast enough
          > > that the issue doesn't manifest itself? starting to think the problem
          > > is in AGWtoBPQ, but who knows...the RMS software is pretty beta
          itself.
          > >
          > > For those that haven't "stress" tested their systems with messages of
          > > a few dozen kb, might want to test things.
          > >
          > > Jeff
          > > WA4ZKO
          > >
          > >
          > >
          >
          >
          >
          > --
          > Jay WS7I
          >
        • Jeff Thomas - WA4ZKO
          Only took a couple rf originated 100k messages to trip it up this time, usually takes a few more ;-) Here s some logs/stats... ... s WILKYU:WA4ZKO-5} Time
          Message 4 of 8 , Oct 8, 2008
          • 0 Attachment
            Only took a couple rf originated 100k messages to trip it up this
            time, usually takes a few more ;-)

            Here's some logs/stats...

            ------ stats on fresh restart -----------

            s
            WILKYU:WA4ZKO-5}
            Time active (mins) 0
            Semaphore Get/Rel/Clashes 293 292 0
            Buffers:Max/Cur/Min/Out/Wait 614 613 612 0 0
            Known Nodes 2
            L4 Connects Sent/Rxed 0 0
            L4 Frames TX/RX/Resent/Reseq 0 0 0 0
            L3 Frames Relayed 0
            Port 1 Port 2
            L2 Frames Digied 0 0
            L2 Frames Heard 0 0
            L2 Frames Rxed 0 0
            L2 Frames Sent 0 0
            L2 Timeouts 0 0
            REJ Frames Rxed 0 0
            RX out of Seq 0 0
            L2 Resequenced 0 0
            Undrun/Poll T-O 0 0
            RX Overruns 0 0
            RX CRC Errors 0 0
            FRMRs Sent 0 0
            FRMRs Received 0 0
            Frames abandoned 0 0
            Link Active % 0 0 0 0


            ----------- after sending a 100k msg via RF>BPQ>RMS no errors ---------

            s
            WILKYU:WA4ZKO-5}
            Time active (mins) 24
            Semaphore Get/Rel/Clashes 43906 43905 3719
            Buffers:Max/Cur/Min/Out/Wait 614 613 604 0 0
            Known Nodes 2
            L4 Connects Sent/Rxed 0 0
            L4 Frames TX/RX/Resent/Reseq 0 0 0 0
            L3 Frames Relayed 0
            Port 1 Port 2
            L2 Frames Digied 0 0
            L2 Frames Heard 1475 0
            L2 Frames Rxed 1475 0
            L2 Frames Sent 832 3
            L2 Timeouts 2 0
            REJ Frames Rxed 0 0
            RX out of Seq 121 0
            L2 Resequenced 0 0
            Undrun/Poll T-O 0 0
            RX Overruns 0 0
            RX CRC Errors 0 0
            FRMRs Sent 0 0
            FRMRs Received 0 0
            Frames abandoned 0 0
            Link Active % 0 0 0 0

            ----------- after sending another 100k msg that tripped up RMS ---------

            s
            WILKYU:WA4ZKO-5}
            Time active (mins) 47
            Semaphore Get/Rel/Clashes 86552 86551 7371
            Buffers:Max/Cur/Min/Out/Wait 614 613 603 0 0
            Known Nodes 2
            L4 Connects Sent/Rxed 0 0
            L4 Frames TX/RX/Resent/Reseq 0 0 0 0
            L3 Frames Relayed 0
            Port 1 Port 2
            L2 Frames Digied 0 0
            L2 Frames Heard 2952 0
            L2 Frames Rxed 2950 0
            L2 Frames Sent 1642 6
            L2 Timeouts 3 0
            REJ Frames Rxed 0 0
            RX out of Seq 233 0
            L2 Resequenced 0 0
            Undrun/Poll T-O 0 0
            RX Overruns 0 0
            RX CRC Errors 0 0
            FRMRs Sent 0 0
            FRMRs Received 0 0
            Frames abandoned 0 0
            Link Active % 0 0 0 0

            -------------------------------------------------------

            From the RMS Packet Exceptions log:

            2008/10/08 16:42:57 1.1.4.0 [AGW OnDataIn] Index was out of range.
            Must be non-negative and less than the size of the collection.
            Parameter name: index
            2008/10/08 17:01:32 1.1.4.0 [AGW OnDataIn] Object reference not set to
            an instance of an object.
            2008/10/08 17:01:32 1.1.4.0 [AGW OnDataIn] Object reference not set to
            an instance of an object.
            2008/10/08 17:01:32 1.1.4.0 [AGW OnDataIn] Object reference not set to
            an instance of an object.
            2008/10/08 17:01:32 1.1.4.0 ServerAGWPacket.ProcessOutboundQueue:
            Object reference not set to an instance of an object.
            2008/10/08 17:01:33 1.1.4.0 ServerAGWPacket.ProcessOutboundQueue:
            Object reference not set to an instance of an object.
            2008/10/08 17:01:33 1.1.4.0 [AGW OnDataIn] Object reference not set to
            an instance of an object.
            2008/10/08 17:01:33 1.1.4.0 ServerAGWPacket.ProcessOutboundQueue:
            Object reference not set to an instance of an object.
            2008/10/08 17:01:34 1.1.4.0 [AGW OnDataIn] Object reference not set to
            an instance of an object.
            2008/10/08 17:01:34 1.1.4.0 [AGW OnDataIn] Object reference not set to
            an instance of an object.
            2008/10/08 17:01:34 1.1.4.0 [AGW OnDataIn] Object reference not set to
            an instance of an object.
            2008/10/08 17:01:35 1.1.4.0 [AGW OnDataIn] Object reference not set to
            an instance of an object.
            2008/10/08 17:01:42 1.1.4.0 [AGW OnDataIn] Object reference not set to
            an instance of an object.
            2008/10/08 17:01:42 1.1.4.0 [AGW OnDataIn] Object reference not set to
            an instance of an object.

            and lots more of the same AGW OnDataIN errors.

            --------------------------------------------------------------

            From the RMS TCP events log file:

            2008/10/08 17:01:32 [10] upper case "W." Notice also the space
            between SEDAN and Wx.
            2008/10/08 17:06:32 [10] Disconnected

            ----------------------------------------------------------------

            From the RMS Packet Events log file:

            2008/10/08 17:01:24 [10] ed feature.
            2008/10/08 17:01:24 [10] the transparent mode).
            2008/10/08 17:01:27 [10] udrate changes.

            then nothing else. Which is odd.

            I'm also noticing that the "channel events" windows in RMS do not
            show all traffic being sent in, just bits and pieces of it...normal or
            not?? This is the case whether things error or not, so may not be
            related?

            Thanks for looking into this.

            73
            Jeff
            WA4ZKO








            --- In BPQ32@yahoogroups.com, "John Wiseman" <john.wiseman@...> wrote:
            >
            > Hi Jeff,
            >
            > This could be a flow control problem between RMS Packet, AGWtoBPQ
            and BPQ32.
            > Could you check the BPQ stats? If it shows a low current buffer
            count during
            > a transfer, or a low minimum buffers, even if the current count
            recovers,
            > then it needs investigating.
            >
            > 73, John G8BPQ
            >
          • John Wiseman
            Ok Jeff. There is no indication of a buffers problem, and the link looks amazingly good (3 timeouts out of 1642 frames). Also it looks as if all, or virtually
            Message 5 of 8 , Oct 8, 2008
            • 0 Attachment
              Ok Jeff. There is no indication of a buffers problem, and the link looks amazingly good (3 timeouts out of 1642 frames). Also it looks as if all, or virtually all, the data was received on the second transmission (the send and receive counts after the second transfer are almost exactly twice those after the first). It would be interesting to know if the failure was reported right at the end of the transmission, and it the pattern is repeatable.
               
              It's impossible to be sure without knowing how RMS Packet works, but the errors in the log seem to indicate that it has lost track of the connection - the AGW OnDataIn messages probably indicate data being received when there is no valid session. This could be an error in RMS Packet, but could equally be due to AGWtoBPQ erroneously reporting that the connection had closed.
               
              I'm not sure where to go from here. I could add code to AGWtoBPQ to trace the interface with RMS Packet, but that would generate a huge volume of data, and take an age to analyse. It would be interesting to know if anyone is successfully sending large messages, and also if there is a size below which it always works. It would help to know if it fails as often sending 10 x 10K messages as sending one 100k. That would indicate whether the problem is related to volume or time of an individual connection, or just a random event.
               
              73,
              John
               
               
               
              -----Original Message-----
              From: BPQ32@yahoogroups.com [mailto:BPQ32@yahoogroups.com]On Behalf Of Jeff Thomas - WA4ZKO
              Sent: 08 October 2008 18:33
              To: BPQ32@yahoogroups.com
              Subject: [BPQ32] Some stats/logs - BPQ32 & RMS Packet failures

              Only took a couple rf originated 100k messages to trip it up this
              time, usually takes a few more ;-)

              Here's some logs/stats.. .

              ------ stats on fresh restart -----------

              s
              WILKYU:WA4ZKO- 5}
              Time active (mins) 0
              Semaphore Get/Rel/Clashes 293 292 0
              Buffers:Max/ Cur/Min/Out/ Wait 614 613 612 0 0
              Known Nodes 2
              L4 Connects Sent/Rxed 0 0
              L4 Frames TX/RX/Resent/ Reseq 0 0 0 0
              L3 Frames Relayed 0
              Port 1 Port 2
              L2 Frames Digied 0 0
              L2 Frames Heard 0 0
              L2 Frames Rxed 0 0
              L2 Frames Sent 0 0
              L2 Timeouts 0 0
              REJ Frames Rxed 0 0
              RX out of Seq 0 0
              L2 Resequenced 0 0
              Undrun/Poll T-O 0 0
              RX Overruns 0 0
              RX CRC Errors 0 0
              FRMRs Sent 0 0
              FRMRs Received 0 0
              Frames abandoned 0 0
              Link Active % 0 0 0 0

              ----------- after sending a 100k msg via RF>BPQ>RMS no errors ---------

              s
              WILKYU:WA4ZKO- 5}
              Time active (mins) 24
              Semaphore Get/Rel/Clashes 43906 43905 3719
              Buffers:Max/ Cur/Min/Out/ Wait 614 613 604 0 0
              Known Nodes 2
              L4 Connects Sent/Rxed 0 0
              L4 Frames TX/RX/Resent/ Reseq 0 0 0 0
              L3 Frames Relayed 0
              Port 1 Port 2
              L2 Frames Digied 0 0
              L2 Frames Heard 1475 0
              L2 Frames Rxed 1475 0
              L2 Frames Sent 832 3
              L2 Timeouts 2 0
              REJ Frames Rxed 0 0
              RX out of Seq 121 0
              L2 Resequenced 0 0
              Undrun/Poll T-O 0 0
              RX Overruns 0 0
              RX CRC Errors 0 0
              FRMRs Sent 0 0
              FRMRs Received 0 0
              Frames abandoned 0 0
              Link Active % 0 0 0 0

              ----------- after sending another 100k msg that tripped up RMS ---------

              s
              WILKYU:WA4ZKO- 5}
              Time active (mins) 47
              Semaphore Get/Rel/Clashes 86552 86551 7371
              Buffers:Max/ Cur/Min/Out/ Wait 614 613 603 0 0
              Known Nodes 2
              L4 Connects Sent/Rxed 0 0
              L4 Frames TX/RX/Resent/ Reseq 0 0 0 0
              L3 Frames Relayed 0
              Port 1 Port 2
              L2 Frames Digied 0 0
              L2 Frames Heard 2952 0
              L2 Frames Rxed 2950 0
              L2 Frames Sent 1642 6
              L2 Timeouts 3 0
              REJ Frames Rxed 0 0
              RX out of Seq 233 0
              L2 Resequenced 0 0
              Undrun/Poll T-O 0 0
              RX Overruns 0 0
              RX CRC Errors 0 0
              FRMRs Sent 0 0
              FRMRs Received 0 0
              Frames abandoned 0 0
              Link Active % 0 0 0 0

              ------------ --------- --------- --------- --------- -------

              From the RMS Packet Exceptions log:

              2008/10/08 16:42:57 1.1.4.0 [AGW OnDataIn] Index was out of range.
              Must be non-negative and less than the size of the collection.
              Parameter name: index
              2008/10/08 17:01:32 1.1.4.0 [AGW OnDataIn] Object reference not set to
              an instance of an object.
              2008/10/08 17:01:32 1.1.4.0 [AGW OnDataIn] Object reference not set to
              an instance of an object.
              2008/10/08 17:01:32 1.1.4.0 [AGW OnDataIn] Object reference not set to
              an instance of an object.
              2008/10/08 17:01:32 1.1.4.0 ServerAGWPacket. ProcessOutboundQ ueue:
              Object reference not set to an instance of an object.
              2008/10/08 17:01:33 1.1.4.0 ServerAGWPacket. ProcessOutboundQ ueue:
              Object reference not set to an instance of an object.
              2008/10/08 17:01:33 1.1.4.0 [AGW OnDataIn] Object reference not set to
              an instance of an object.
              2008/10/08 17:01:33 1.1.4.0 ServerAGWPacket. ProcessOutboundQ ueue:
              Object reference not set to an instance of an object.
              2008/10/08 17:01:34 1.1.4.0 [AGW OnDataIn] Object reference not set to
              an instance of an object.
              2008/10/08 17:01:34 1.1.4.0 [AGW OnDataIn] Object reference not set to
              an instance of an object.
              2008/10/08 17:01:34 1.1.4.0 [AGW OnDataIn] Object reference not set to
              an instance of an object.
              2008/10/08 17:01:35 1.1.4.0 [AGW OnDataIn] Object reference not set to
              an instance of an object.
              2008/10/08 17:01:42 1.1.4.0 [AGW OnDataIn] Object reference not set to
              an instance of an object.
              2008/10/08 17:01:42 1.1.4.0 [AGW OnDataIn] Object reference not set to
              an instance of an object.

              and lots more of the same AGW OnDataIN errors.

              ------------ --------- --------- --------- --------- --------- -

              From the RMS TCP events log file:

              2008/10/08 17:01:32 [10] upper case "W." Notice also the space
              between SEDAN and Wx.
              2008/10/08 17:06:32 [10] Disconnected

              ------------ --------- --------- --------- --------- --------- -

              From the RMS Packet Events log file:

              2008/10/08 17:01:24 [10] ed feature.
              2008/10/08 17:01:24 [10] the transparent mode).
              2008/10/08 17:01:27 [10] udrate changes.

              then nothing else. Which is odd.

              I'm also noticing that the "channel events" windows in RMS do not
              show all traffic being sent in, just bits and pieces of it...normal or
              not?? This is the case whether things error or not, so may not be
              related?

              Thanks for looking into this.

              73
              Jeff
              WA4ZKO

              --- In BPQ32@yahoogroups. com, "John Wiseman" <john.wiseman@ ...> wrote:
              >
              > Hi Jeff,
              >
              > This could be a flow control problem between RMS Packet, AGWtoBPQ
              and BPQ32.
              > Could you check the BPQ stats? If it shows a low current buffer
              count during
              > a transfer, or a low minimum buffers, even if the current count
              recovers,
              > then it needs investigating.
              >
              > 73, John G8BPQ
              >

            • David Wagner WA2DXQ
              John. If you want info on RMS you should talk to Rick KN6KB , If you want his personal e-mail I ll send it to you in private e-mail. Send to
              Message 6 of 8 , Oct 8, 2008
              • 0 Attachment
                John.
                If you want info on RMS you should talk to Rick KN6KB , If you want his personal e-mail
                 
                I'll send it to you in private e-mail.
                 
                Send to wa2dxq@...
                 
                I expect he can explain what RMS errors might mean.
                 
                Dave
                WA2DXQ
                RMS Ft. Lauderdale
                 
                 
                ----- Original Message -----
                Sent: Wednesday, October 08, 2008 15:54
                Subject: RE: [BPQ32] Some stats/logs - BPQ32 & RMS Packet failures

                Ok Jeff. There is no indication of a buffers problem, and the link looks amazingly good (3 timeouts out of 1642 frames). Also it looks as if all, or virtually all, the data was received on the second transmission (the send and receive counts after the second transfer are almost exactly twice those after the first). It would be interesting to know if the failure was reported right at the end of the transmission, and it the pattern is repeatable.
                 
                It's impossible to be sure without knowing how RMS Packet works, but the errors in the log seem to indicate that it has lost track of the connection - the AGW OnDataIn messages probably indicate data being received when there is no valid session. This could be an error in RMS Packet, but could equally be due to AGWtoBPQ erroneously reporting that the connection had closed.
                 
                I'm not sure where to go from here. I could add code to AGWtoBPQ to trace the interface with RMS Packet, but that would generate a huge volume of data, and take an age to analyse. It would be interesting to know if anyone is successfully sending large messages, and also if there is a size below which it always works. It would help to know if it fails as often sending 10 x 10K messages as sending one 100k. That would indicate whether the problem is related to volume or time of an individual connection, or just a random event.
                 
                73,
                John
                 
                 
                 
                -----Original Message-----
                From: BPQ32@yahoogroups. com [mailto:BPQ32@ yahoogroups. com]On Behalf Of Jeff Thomas - WA4ZKO
                Sent: 08 October 2008 18:33
                To: BPQ32@yahoogroups. com
                Subject: [BPQ32] Some stats/logs - BPQ32 & RMS Packet failures

                Only took a couple rf originated 100k messages to trip it up this
                time, usually takes a few more ;-)

                Here's some logs/stats.. .

                ------ stats on fresh restart -----------

                s
                WILKYU:WA4ZKO- 5}
                Time active (mins) 0
                Semaphore Get/Rel/Clashes 293 292 0
                Buffers:Max/ Cur/Min/Out/ Wait 614 613 612 0 0
                Known Nodes 2
                L4 Connects Sent/Rxed 0 0
                L4 Frames TX/RX/Resent/ Reseq 0 0 0 0
                L3 Frames Relayed 0
                Port 1 Port 2
                L2 Frames Digied 0 0
                L2 Frames Heard 0 0
                L2 Frames Rxed 0 0
                L2 Frames Sent 0 0
                L2 Timeouts 0 0
                REJ Frames Rxed 0 0
                RX out of Seq 0 0
                L2 Resequenced 0 0
                Undrun/Poll T-O 0 0
                RX Overruns 0 0
                RX CRC Errors 0 0
                FRMRs Sent 0 0
                FRMRs Received 0 0
                Frames abandoned 0 0
                Link Active % 0 0 0 0

                ----------- after sending a 100k msg via RF>BPQ>RMS no errors ---------

                s
                WILKYU:WA4ZKO- 5}
                Time active (mins) 24
                Semaphore Get/Rel/Clashes 43906 43905 3719
                Buffers:Max/ Cur/Min/Out/ Wait 614 613 604 0 0
                Known Nodes 2
                L4 Connects Sent/Rxed 0 0
                L4 Frames TX/RX/Resent/ Reseq 0 0 0 0
                L3 Frames Relayed 0
                Port 1 Port 2
                L2 Frames Digied 0 0
                L2 Frames Heard 1475 0
                L2 Frames Rxed 1475 0
                L2 Frames Sent 832 3
                L2 Timeouts 2 0
                REJ Frames Rxed 0 0
                RX out of Seq 121 0
                L2 Resequenced 0 0
                Undrun/Poll T-O 0 0
                RX Overruns 0 0
                RX CRC Errors 0 0
                FRMRs Sent 0 0
                FRMRs Received 0 0
                Frames abandoned 0 0
                Link Active % 0 0 0 0

                ----------- after sending another 100k msg that tripped up RMS ---------

                s
                WILKYU:WA4ZKO- 5}
                Time active (mins) 47
                Semaphore Get/Rel/Clashes 86552 86551 7371
                Buffers:Max/ Cur/Min/Out/ Wait 614 613 603 0 0
                Known Nodes 2
                L4 Connects Sent/Rxed 0 0
                L4 Frames TX/RX/Resent/ Reseq 0 0 0 0
                L3 Frames Relayed 0
                Port 1 Port 2
                L2 Frames Digied 0 0
                L2 Frames Heard 2952 0
                L2 Frames Rxed 2950 0
                L2 Frames Sent 1642 6
                L2 Timeouts 3 0
                REJ Frames Rxed 0 0
                RX out of Seq 233 0
                L2 Resequenced 0 0
                Undrun/Poll T-O 0 0
                RX Overruns 0 0
                RX CRC Errors 0 0
                FRMRs Sent 0 0
                FRMRs Received 0 0
                Frames abandoned 0 0
                Link Active % 0 0 0 0

                ------------ --------- --------- --------- --------- -------

                From the RMS Packet Exceptions log:

                2008/10/08 16:42:57 1.1.4.0 [AGW OnDataIn] Index was out of range.
                Must be non-negative and less than the size of the collection.
                Parameter name: index
                2008/10/08 17:01:32 1.1.4.0 [AGW OnDataIn] Object reference not set to
                an instance of an object.
                2008/10/08 17:01:32 1.1.4.0 [AGW OnDataIn] Object reference not set to
                an instance of an object.
                2008/10/08 17:01:32 1.1.4.0 [AGW OnDataIn] Object reference not set to
                an instance of an object.
                2008/10/08 17:01:32 1.1.4.0 ServerAGWPacket. ProcessOutboundQ ueue:
                Object reference not set to an instance of an object.
                2008/10/08 17:01:33 1.1.4.0 ServerAGWPacket. ProcessOutboundQ ueue:
                Object reference not set to an instance of an object.
                2008/10/08 17:01:33 1.1.4.0 [AGW OnDataIn] Object reference not set to
                an instance of an object.
                2008/10/08 17:01:33 1.1.4.0 ServerAGWPacket. ProcessOutboundQ ueue:
                Object reference not set to an instance of an object.
                2008/10/08 17:01:34 1.1.4.0 [AGW OnDataIn] Object reference not set to
                an instance of an object.
                2008/10/08 17:01:34 1.1.4.0 [AGW OnDataIn] Object reference not set to
                an instance of an object.
                2008/10/08 17:01:34 1.1.4.0 [AGW OnDataIn] Object reference not set to
                an instance of an object.
                2008/10/08 17:01:35 1.1.4.0 [AGW OnDataIn] Object reference not set to
                an instance of an object.
                2008/10/08 17:01:42 1.1.4.0 [AGW OnDataIn] Object reference not set to
                an instance of an object.
                2008/10/08 17:01:42 1.1.4.0 [AGW OnDataIn] Object reference not set to
                an instance of an object.

                and lots more of the same AGW OnDataIN errors.

                ------------ --------- --------- --------- --------- --------- -

                From the RMS TCP events log file:

                2008/10/08 17:01:32 [10] upper case "W." Notice also the space
                between SEDAN and Wx.
                2008/10/08 17:06:32 [10] Disconnected

                ------------ --------- --------- --------- --------- --------- -

                From the RMS Packet Events log file:

                2008/10/08 17:01:24 [10] ed feature.
                2008/10/08 17:01:24 [10] the transparent mode).
                2008/10/08 17:01:27 [10] udrate changes.

                then nothing else. Which is odd.

                I'm also noticing that the "channel events" windows in RMS do not
                show all traffic being sent in, just bits and pieces of it...normal or
                not?? This is the case whether things error or not, so may not be
                related?

                Thanks for looking into this.

                73
                Jeff
                WA4ZKO

                --- In BPQ32@yahoogroups. com, "John Wiseman" <john.wiseman@ ...> wrote:
                >
                > Hi Jeff,
                >
                > This could be a flow control problem between RMS Packet, AGWtoBPQ
                and BPQ32.
                > Could you check the BPQ stats? If it shows a low current buffer
                count during
                > a transfer, or a low minimum buffers, even if the current count
                recovers,
                > then it needs investigating.
                >
                > 73, John G8BPQ
                >

              • Jeff Thomas - WA4ZKO
                John thanks for taking a look at it. Yeap, it seems to run along nicely till the glitch strikes. I ll run some more tests when I get a chance and likely drop
                Message 7 of 8 , Oct 8, 2008
                • 0 Attachment
                  John thanks for taking a look at it. Yeap, it seems to run along
                  nicely till the glitch strikes. I'll run some more tests when I get a
                  chance and likely drop my "bug report" onto the RMS mailing list. I'm
                  also curious to see what Jay WS7I comes up with this weekend in his
                  testing as I think he's seen similar glitches and has some high speed
                  ports on his setup.

                  I'm no expert, you don't want me writing code(grin), but I tend to
                  think that RMS is randomly choking (for lack of a better word) on the
                  data stream from BPQ, especially on the larger transfers. Maybe a
                  delay on the telnet side of things causes a "data backlog" that in the
                  right moment can trigger it. Buffering problem in RMS?

                  I don't think I've ever seen it happen on small messages, but I could
                  be wrong. I'll create some more test messages of various sizes and
                  throw them at it. I'd also like to examine the time frames involved,
                  but I think it's random and only when it's under a bit more stress
                  from a non-1.2k rf feed..but I can't prove that.

                  The RMS Packet software is pretty much beta stuff from what I gather.
                  New versions are being released fairly frequently. I know at one time
                  one of the RMS developers was on this list and was working on getting
                  a native option for connecting to BPQ. You can see that it's been
                  started, but not ready yet (grayed out). Maybe the AGW interface code
                  in it is currently just a loose framework that is far from complete
                  and that works for light duty stuff, but still needs a lot of "real
                  world" debugging done on it? Maybe just timing issues in the code that
                  don't even show up at 1200 baud (where the bulk of folks are), but who
                  knows.

                  John, nice work on BPQ32! I was originally looking at using FLexNet,
                  but it looks to be stagnated and no longer supported by the original
                  developers. Probably my only feature request for BPQ32 would be a way
                  to log with date/time stamp all the switch traffic to daily log files.
                  Maybe a log to disk option sysops could enable in the terminal program
                  as it would be a good start as traffic on it is already time stamped.
                  Thanks for your efforts on BPQ32.

                  73
                  Jeff
                  WA4ZKO
                Your message has been successfully submitted and would be delivered to recipients shortly.