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

BPQ32 & RMS Packet failures

Expand Messages
  • Jeff Thomas - WA4ZKO
    Yeah, this probably should go onto the RMS Packet list too, but since I d say the bulk of the folks using this combination are on here, figured I d try this
    Message 1 of 1 , Oct 6, 2008
    • 0 Attachment
      Yeah, this probably should go onto the RMS Packet list too, but since
      I'd say the bulk of the folks using this combination are on here,
      figured I'd try this list first.

      Here's what I've got. Windows Svr 2003 running the latest version of
      BPQ32 v4.10e, RMS Packet v1.1.4.0, a 19.2 serial port with an AEA
      PK-96 running RF at 9600 baud on UHF, an AX-IP port (that I've done
      nothing with yet), and using AGWtoBPQ v1.1.0 on port 8000. RMS Packet
      is setup to use AGWPE on a remote computer, but actually just set for
      127.0.0.1 (local loopback) and all works fine for a bit. Nothing
      really odd about this setup, best I can tell.

      The BPQ node/switch seems fine, but I've been pushing and pulling some
      test emails of various sizes (30k, 100k) though the RMS Packet system.
      It will seem fine for a bit, then suddenly (no rhyme or reason, same
      test message) it will just stop passing data into RMS, RMS will time
      out on telnet. If I try to disconnect the link for the incoming packet
      connection it will just set there. All I can do is exit RMS to clear
      it. RMS comes right back up, so the BPQ side apparently hasn't
      crashed. After RMS "dies" the BPQ node is still usable, but attempts
      to connect to the RMS system via RF just comes up with "connected" and
      it stalls there if I don't exit RMS and bring it back up (duh).

      It seems to only do this on outgoing messages, I can pull messages in
      all day and have yet to see this happen.

      Looking at the RMS logs I see this in the packet exceptions log:

      2008/10/06 03:16:56 1.1.4.0 [AGW OnDataIn] Object reference not set to
      an instance of an object.
      2008/10/06 03:16:56 1.1.4.0 [AGW OnDataIn] Object reference not set to
      an instance of an object.
      2008/10/06 03:16:56 1.1.4.0 ServerAGWPacket.ProcessOutboundQueue:
      Object reference not set to an instance of an object.
      2008/10/06 03:16:57 1.1.4.0 ServerAGWPacket.ProcessOutboundQueue:
      Object reference not set to an instance of an object.
      2008/10/06 03:16:57 1.1.4.0 ServerAGWPacket.ProcessOutboundQueue:
      Object reference not set to an instance of an object.
      2008/10/06 03:16:58 1.1.4.0 [AGW OnDataIn] Object reference not set to
      an instance of an object.
      2008/10/06 03:16:58 1.1.4.0 [AGW OnDataIn] Object reference not set to
      an instance of an object.
      2008/10/06 03:16:59 1.1.4.0 [AGW OnDataIn] Object reference not set to
      an instance of an object.
      2008/10/06 03:16:59 1.1.4.0 [AGW OnDataIn] Object reference not set to
      an instance of an object.
      2008/10/06 03:17:04 1.1.4.0 [AGW OnDataIn] Object reference not set to
      an instance of an object.
      2008/10/06 03:17:04 1.1.4.0 [AGW OnDataIn] Object reference not set to
      an instance of an object.
      2008/10/06 03:17:05 1.1.4.0 [AGW OnDataIn] Object reference not set to
      an instance of an object.
      AND A BOATLOAD MORE of the same.

      Since I'm using a USB to serial adapter, I questioned how solid it
      was. So I took BPQ down and setup RMS to talk direct to the PK-96 TNC
      over the com port. The error didn't come up, seems solid.

      I haven't messed with any of the rf timings (paclen, frack, maxframe)
      which clearly could use some tweaks for throughput at 9600 baud, but
      figured I'd stay with the more conservative settings for now. I think
      TXDelay is the only thing I changed (set at a pretty moderate 200ms
      for now) which the rig handles fine.

      I haven't had a chance to test it at 1200 baud versus 9600 and see if
      that made a difference...pushing such test messages thru at 1.2k
      wouldn't be much fun, grin.

      Might just be in my head, but it seems like the problem is not as bad
      with the latest RMS Packet v1.1.4.0 versus the prior version, but is
      still there.

      Any ideas? Something on the RMS side? Something on the AGWtoBPQ side?

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