ANNOUNCEMENT: Live Wireshark University & Allegro Packets online APAC Wireshark Training Session
July 17th, 2024 | 10:00am-11:55am SGT (UTC+8) | Online

Ethereal-dev: Re: [Ethereal-dev] decode NetMeeting packet

Note: This archive is from the project's previous web site, ethereal.com. This list is no longer active.

From: Guy Harris <guy@xxxxxxxxxxxx>
Date: Mon, 25 Aug 2003 19:41:43 -0700

On Monday, August 25, 2003, at 7:25 PM, li wrote:

I use Ethereal to decode NetMeeting packet, NetMeeting always send TPKT header first which contain the PDU length, then wait for the ACK, after get the ACK, it will send the H.225 signalling to establish the connection.
 
Client              Server
------------------------> send TPKT header
<----------------------- send ACK
------------------------>send H.225 Setup(contain PDU)
 
Ethereal decode first packet "Desegmented TCP", and merge the third packet with the first packet, which mean insert TPKT header, then decode it as:

	[incomplete image, which didn't say anything about a "Fragment Offset"]

but the third packet's "Fragment Offset" is 0, why insert this header?

The header is inserted because the TPKT dissector considers a TPKT packet to consist of the TPKT header and the TPKT payload. There is no guarantee that you will get the TPKT header in one frame and the TPKT payload in the next frame - you could get them both in one frame, or you could get the TPKT header and part of the TPKT payload in one frame and the rest of the TPKT payload in subsequent frames. You could even, conceivably, get part of the TPKT header in one frame and the rest of the TPKT header, and the TPKT payload, in subsequent frames.

Therefore, the TPKT dissector reassembles the TPKT header and payload as a single unit; this means that the TPKT header and payload will be dissected in the third frame.