I did a wandisplay from a pipeline 130 (4.6B)
that was having the problem. This shows the trace
when I did a telnet from 206.249.25.2 through a P-130 and PM2
to 205.164.176.32 port 25: (The telnet never
got through.)
[I've identified some TCP/IP packet and segment
information in the first packet for entertainment
purposes only.]
{SYN request}
XMIT-9:: 53 octets @ 25869C
[0000]: ff 03 00 3d c0 00 02 58 21 45 00 00 2c e3 be 40
[0010]: 00 7f 06 b2 4c ce f9 19 02 cd a4 b0 20 0c 37 00
Source IP ^---------^ ^---------^ Dest IP
[0020]: 19 23 ae e7 1b 00 00 00 00 60 02 20 00 fb 4b 00
^^ Destination Port 25 ^^ TCP Code Bits
[0030]: 00 02 04 05 b4
{SYN ACK}
RECV-9:: 53 octets @ 253764
[0000]: ff 03 00 3d c0 00 41 04 21 45 00 00 2c 5f fe 00
[0010]: 00 3f 06 b6 0d cd a4 b0 20 ce f9 19 02 00 19 0c
[0020]: 37 fc b6 81 7e 23 ae e7 1c 60 12 21 80 7f 21 00
[0030]: 00 02 04 02 18
{ACK}
XMIT-9:: 49 octets @ 25869C
[0000]: ff 03 00 3d c0 00 02 59 21 45 00 00 28 e4 be 40
[0010]: 00 7f 06 b1 50 ce f9 19 02 cd a4 b0 20 0c 37 00
[0020]: 19 23 ae e7 1c fc b6 81 7f 50 10 21 80 93 42 00
[0030]: 00
{FIN} WHAT's THIS Doing here?
RECV-9:: 69 octets @ 253D7C
[0000]: ff 03 00 3d c0 00 41 05 21 45 00 00 3c 60 0e 00
[0010]: 00 3f 06 b5 ed cd a4 b0 20 ce f9 19 02 0d b4 00
[0020]: 71 fc b7 a1 ea 00 00 00 00 a0 02 20 00 ed ae 00
[0030]: 00 02 04 02 18 01 03 03 00 01 01 08 0a 00 02 2e
[0040]: 6b 00 00 00 00
{RST ACK}
XMIT-9:: 49 octets @ 25869C
[0000]: ff 03 00 3d c0 00 02 5a 21 45 00 00 28 e6 be 00
[0010]: 00 7f 06 ef 50 ce f9 19 02 cd a4 b0 20 00 71 0d
[0020]: b4 00 00 00 00 fc b7 a1 eb 50 14 00 00 9d 47 00
[0030]: 00
{PUSH ACK}
RECV-9:: 147 octets @ 254394
[0000]: ff 03 00 3d c0 00 41 06 2f 45 00 00 8a 60 11 00
[0010]: 00 3f 0e b5 9c cd a4 b0 20 ce f9 19 02 00 19 0c
[0020]: 37 fc b6 81 7f 23 ae e7 1c 50 18 21 80 5b 0b 00
[0030]: 00 32 32 30 20 68 69 77 61 79 31 2e 65 78 69 74
[0040]: 31 30 39 2e 63 6f 6d 20 45 53 4d 54 50 20 53 65
[0050]: 6e 64 6d 61 69 6c 20 38 2e 37 2e 36 2f 38 2e 36
[0060]: 2e 39 20 72 65 61 64 79 20 61 74 20 54 68 75 2c
[0070]: 20 31 39 20 44 65 63 20 31 39 39 36 20 31 33 3a
[0080]: 35 37 3a 33 39 20 2d 30 35 30 30 20 28 45 53 54
[0090]: 29 0d 0a
The PUSH ACK repeats over and over until I abort my
telnet attempt.
Compare against the failed telnet in the other
direction as captured by the Portmaster (below.)
I upgraded the Pipeline to 4.6Cp11 and the same
thing happened.
I want to add that we've NEVER seen this happen
to P50's running 4.5B (or that the frequency of occurance
is so low that we don't even notice.)
This problem also occurs when the destination address
is the router itself. Ex: telnet from portmaster
to pipeline can hang with the same results.
I know this post is tedious, but there must be LOTS of you
out there trying to do what we're doing.
We NEED this fixed. I don't care who's fault it is, I
just want someone to work with us to diagnose, debug,
and fix this. In the past when we put in
trouble tickets with both companies we basically got
"It sounds like a [insert other company name here]
problem. Contact them."
-Ed
On Thu, 19 Dec 1996, Edward Jung wrote:
> We have had an on again off again problem with Ascend pipelines
> and the Portmaster2 ISDN. Basically, TCP sessions get lost.
>
(Failed request below. Telnet attempt from
205.164.176.34 to 206.84.156.131 port 80.
Note the SYN ACKs that keep coming from the
remote end. The PUSH ACKs occured when I tried to
send an HTTP GET request.)
TCP from 205.164.176.34.2484 to 206.84.156.131.80 seq FC697055, ack 0x0, win 8192, SYN
TCP from 206.84.156.131.80 to 205.164.176.34.2484 seq 96C26A1, ack 0xFC697056, win 8576, SYN ACK
TCP from 205.164.176.34.2484 to 206.84.156.131.80 seq FC697056, ack 0x96C26A2, win 8576, ACK
TCP from 206.84.156.131.80 to 205.164.176.34.2484 seq 96C26A1, ack 0xFC697056, win 8576, SYN ACK
TCP from 205.164.176.34.2484 to 206.84.156.131.80 seq FC697056, ack 0x96C26A2, win 8576, ACK
TCP from 205.164.176.34.2484 to 206.84.156.131.80 seq FC697056, ack 0x96C26A2, win 8576, PUSH ACK , 16 bytes
TCP from 205.164.176.34.2484 to 206.84.156.131.80 seq FC697056, ack 0x96C26A2, win 8576, PUSH ACK , 18 bytes
TCP from 206.84.156.131.80 to 205.164.176.34.2484 seq 96C26A1, ack 0xFC697056, win 8576, SYN ACK
TCP from 205.164.176.34.2484 to 206.84.156.131.80 seq FC697068, ack 0x96C26A2, win 8576, ACK
TCP from 205.164.176.34.2484 to 206.84.156.131.80 seq FC697056, ack 0x96C26A2, win 8576, PUSH ACK , 18 bytes
TCP from 205.164.176.34.2484 to 206.84.156.131.80 seq FC697056, ack 0x96C26A2, win 8576, PUSH ACK , 18 bytes
TCP from 206.84.156.131.80 to 205.164.176.34.2484 seq 96C26A1, ack 0xFC697056, win 8576, SYN ACK
TCP from 205.164.176.34.2484 to 206.84.156.131.80 seq FC697068, ack 0x96C26A2, win 8576, ACK
TCP from 205.164.176.34.2484 to 206.84.156.131.80 seq FC697068, ack 0x96C26A2, win 8576, FIN PUSH ACK , 5 by
tes
--------------------------------------------------------------
edward jung atlantic internet technologies, inc.
ejung@exit109.com 628 shrewsbury ave - red bank - nj
(908)758-0505, -0869 (FAX) http://www.exit109.com/