Wireshark-users: Re: [Wireshark-users] Sectionalizing Network Loss with TCP

From: Tamás Varga <Tamas.Varga@xxxxxxxxxxxx>
Date: Mon, 2 May 2011 17:31:13 +0200
 
Hi Barry,
You may find the above paper useful for understanding the challenges of TCP packet loss calculation.
cheers,
 Tamas


From: wireshark-users-bounces@xxxxxxxxxxxxx [mailto:wireshark-users-bounces@xxxxxxxxxxxxx] On Behalf Of Barry Constantine
Sent: Friday, April 29, 2011 15:34
To: wireshark-users@xxxxxxxxxxxxx
Subject: [Wireshark-users] Sectionalizing Network Loss with TCP

Hello,
 
I am trying to develop a technique to isolate packet loss between networks (i.e. Network A versus Network B loss diagnosis).
 
I've experimented with the common technique that identifies whether a retransmiitted packet has duplicate sequence numbers in the capture file.  The results have been OK, but with acuracy that is not good enough.  Under certain conditions (i.e. when entire windows are not retransmitted), it works well.
 
Has anyone on this list developed a technique using tshark + scripting?
 
Thanks,
Barry