From: C Chan (cch.ccie@gmail.com)
Date: Sat Nov 03 2007 - 23:47:38 ART
Dear Expert,
Assume that I am using R2 to telnet R1. It is common that sudden TCP
disconnection happens in the WAN cloud between R1 and R2. Whenever the
disconnection happens, I found that it is difficult to get back R2 terminal
screen (already telnet to R1 and got halts). I have tried to use "ctrl-c",
"ctrl-x", "ctrl-z", "ctrl-^" to break the session. None of them works.
When I turn on "debug ip tcp trans" on R2 before sudden TCP disconnection
happens. I found that the TCP session will retry 16 times before declaring
session closed automatically. However, it already takes 7 mins before
getting R2 terminal back on my control.
Thanks for your advice in advance.
R28_TS#2
[Resuming connection 2 to r2 ... ]
SENT -> ESTAB [41396 -> 10.0.0.1(23)]
*Nov 3 13:05:43 UTC: TCP: tcb 66DE5D08 connection to 10.0.0.1:23, peer MSS
536, MSS is 536
*Nov 3 13:05:43 UTC: TCB66DE5D08 connected to 10.0.0.1.23
*Nov 3 13:07:47 UTC: TCP0: Data repacketized, seq 1979629618, sent 10 byte
*Nov 3 13:07:47 UTC: 155.1.23.2:41396 <---> 10.0.0.1:23 congestion window
changes
*Nov 3 13:07:47 UTC: cwnd from 573 to 536, ssthresh from 65535 to 1072
*Nov 3 13:07:47 UTC: TCP0: timeout #1 - timeout is 2146 ms, seq 1979629618
*Nov 3 13:07:49 UTC: TCP0: Data repacketized, seq 1979629618, sent 20 byte
*Nov 3 13:07:49 UTC: TCP0: timeout #2 - timeout is 4292 ms, seq 1979629618
*Nov 3 13:07:54 UTC: TCP0: Data repacketized, seq 1979629618, sent 36 byte
*Nov 3 13:07:54 UTC: TCP0: timeout #3 - timeout is 8584 ms, seq 1979629618
*Nov 3 13:08:02 UTC: TCP0: Data repacketized, seq 1979629618, sent 48 byte
*Nov 3 13:08:02 UTC: TCP0: timeout #4 - timeout is 17168 ms, seq 1979629618
*Nov 3 13:08:20 UTC: TCP0: Data repacketized, seq 1979629618, sent 64 byte
*Nov 3 13:08:20 UTC: TCP0: timeout #5 - timeout is 34336 ms, seq 1979629618
*Nov 3 13:08:54 UTC: TCP0: timeout #6 - timeout is 34336 ms, seq 1979629618
*Nov 3 13:09:28 UTC: TCP0: timeout #7 - timeout is 34336 ms, seq 1979629618
*Nov 3 13:10:03 UTC: TCP0: timeout #8 - timeout is 34336 ms, seq 1979629618
*Nov 3 13:10:37 UTC: TCP0: timeout #9 - timeout is 34336 ms, seq 1979629618
*Nov 3 13:11:11 UTC: TCP0: timeout #10 - timeout is 34336 ms, seq
1979629618
*Nov 3 13:11:46 UTC: TCP0: timeout #11 - timeout is 34336 ms, seq
1979629618
*Nov 3 13:12:20 UTC: TCP0: timeout #12 - timeout is 34336 ms, seq
1979629618
*Nov 3 13:12:54 UTC: TCP0: timeout #13 - timeout is 34336 ms, seq
1979629618
*Nov 3 13:13:29 UTC: TCP0: timeout #14 - timeout is 34336 ms, seq
1979629618
*Nov 3 13:14:03 UTC: TCP0: timeout #15 - timeout is 34336 ms, seq
1979629618
*Nov 3 13:14:37 UTC: TCP0: timeout #16 - timeout is 34336 ms, seq
1979629618
[Connection to 10.0.0.1 closed by foreign host]
R2#
*Nov 3 13:24:18 UTC: TCP0: Data repacketized, seq 3363863064, sent 63 byte
*Nov 3 13:24:18 UTC: TCP0: state was ESTAB -> CLOSED [13562 -> 10.0.0.1
(23)]
*Nov 3 13:24:18 UTC: TCB 0x67139504 destroyed
Regards,
cch
This archive was generated by hypermail 2.1.4 : Sat Dec 01 2007 - 06:37:28 ART