tcptrace-bugs Potential rexmit undercount

From: Slava Kritov (Slava.kritov@veritas.com)
Date: 08/04/04

Next message: Purchasing: "tcptrace-bugs New Toner Cartridge Summer Sale."
Previous message: sunmedia@tiger.odn.ne.jp: "tcptrace-bugs $BL$>5Bz9-9p"((J $B3J0B!*EEOC(B$B2CF~8"HNGd!u9b2AGc<h(B"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

Message-ID: <B78ECA1C29CAD4119E0C0008C7866678096225B1@mtvxch08.veritas.com>
From: Slava Kritov <Slava.kritov@veritas.com>
Subject: tcptrace-bugs Potential rexmit undercount
Date: Wed, 4 Aug 2004 19:16:06 -0700

Hello,

I've been using tcptrace to analyze

[sun240, solaris8]-[1G eth]-[Compaq
NIST]-[1G eth]-[sun240, solaris8]

Nist was set to 100 Mbps, 10 ms and

I was capturing either on linux or
irregularity:

For all packets coming from remote
capture) I would see dups - but not

Please consider results of analyzing
marked by ***):

1 arg remaining, starting with 'eth2100-1l
Ostermann's tcptrace -- version 6.6.1

80648 packets seen, 80648 TCP packets
elapsed wallclock time: 0:00:00.368683,
trace file elapsed time: 0:01:00.232135
TCP connection info:
2 TCP connections traced:
TCP connection 1:
      host a:
      host b:
      complete
      first packet:
      last packet:
      elapsed time:
      total
      filename:
   a->b: b->a:
     total packets:
     ack pkts sent:
     pure acks sent:
     sack pkts sent:
     dsack pkts sent:
     max sack blks/ack:
     unique bytes sent:
     actual data pkts:
     actual data bytes:
     rexmt data pkts:
     rexmt data bytes:
     zwnd probe pkts:
     zwnd probe bytes:
     outoforder pkts:
     pushed data pkts:
     SYN/FIN pkts sent:
     req sack:
     sacks sent:
     urgent data pkts:
pkts
     urgent data bytes:
bytes
     mss requested:
bytes
     max segm size:
bytes
     min segm size:
bytes
     avg segm size:
bytes
     max win adv:
bytes
     min win adv:
bytes
     zero win adv:
times
     avg win adv:
bytes
     initial window:
bytes
     initial window:
pkts
     ttl stream length:
bytes
     missed data:
bytes
     truncated data:
bytes
     truncated packets:
pkts
     data xmit time:
secs
     idletime max:
     hardware dups:
segs
*** even if I use no_hw_dups option
left column ***

**
suspect!
throughput:

     RTT samples:
     RTT min:
     RTT max:
     RTT avg:
     RTT stdev:

RTT from 3WHS:

     RTT full_sz smpls:
     RTT full_sz min:
     RTT full_sz max:
     RTT full_sz avg:
     RTT full_sz stdev:

     post-loss acks:
        For
       &nb
       &nb
        of
     ambiguous acks:
     RTT min (last):
     RTT max (last):
     RTT avg (last):
     RTT sdv (last):
     segs cum acked:
     duplicate acks:
     triple dupacks:
     max # retrans:
     min retr time:
     max retr time:
     avg retr time:
     sdv retr time:
================================
TCP connection 2:
      host c:
      host d:
      complete
      first packet:
      last packet:
      elapsed time:
      total
      filename:
   c->d: d->c:
     total packets:
     ack pkts sent:
     pure acks sent:
     sack pkts sent:
     dsack pkts sent:
     max sack blks/ack:
     unique bytes sent:
     actual data pkts:
     actual data bytes:
     rexmt data pkts:
*** This seems to be way more than expected

     rexmt data bytes:
     zwnd probe pkts:
     zwnd probe bytes:
     outoforder pkts:
     pushed data pkts:
     SYN/FIN pkts sent:
     req sack:
     sacks sent:
     urgent data pkts:
pkts
     urgent data bytes:
bytes
     mss requested:
bytes
     max segm size:
bytes
     min segm size:
bytes
     avg segm size:
bytes
     max win adv:
bytes
     min win adv:
bytes
     zero win adv:
times
     avg win adv:
bytes
     initial window:
bytes
     initial window:
pkts
     ttl stream length:
bytes
     missed data:
bytes
     truncated data:
bytes
     truncated packets:
pkts
     data xmit time:
secs
     idletime max:
     hardware dups:
segs
*** again - very strange stats above

**
suspect!
throughput:

     RTT samples:
     RTT min:
     RTT max:
     RTT avg:
     RTT stdev:

RTT from 3WHS:

     RTT full_sz smpls:
     RTT full_sz min:
     RTT full_sz max:
     RTT full_sz avg:
     RTT full_sz stdev:

Thanks a lot

Slava

Next message: Previous message:
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

This archive was generated by Purchasing: "tcptrace-bugs New Toner Cartridge Summer Sale." sunmedia@tiger.odn.ne.jp: "tcptrace-bugs $BL$>5Bz9-9p"((J $B3J0B!*EEOC(B$B2CF~8"HNGd!u9b2AGc<h(B" /www.hypermail.org/">hypermail 2.1.7