Ethereal-dev: [Ethereal-dev] Wrong decoding of tektronix K15 Gb file

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 This list is no longer active.

From: "Yann Samama" <ysamama@xxxxxxxxxx>
Date: Wed, 10 Aug 2005 19:14:19 +0200
Hi All!


I was further testing the reading of the K15 file format by Ethereal
when I encountered the following issues with a GPRS Gb (frame relay)
trace (see enclosed file gb_merlin.rf5) :

1/ The numbers of packets and the packets themselves seen by Ethereal
and the K15 viewer are different.
 => K15 viewer sees 257 packets (from 7 to 263, see enclosed file
k15.jpg).
 => Ethereal sees 147 packets (from 1 to 147, see enclosed file
ethereal.jpg).
 => First analysis shows that the first packet to differ (caution : I
checked only the packets' length, so maybe it's sooner?) is #21 in
Ethereal.

2/ The statistics for GSM/A-Interface DTAP/GPRS Session Management
wrongly displays 0 for all counters from time to time.
 => I wanted to check the session management (create, activate, delete,
etc. : filter set to "gsm_a.dtap_msg_sm_type") counters.
 => Sometimes the stats are OK, sometimes all counters are set to 0.

3/ The start timestamp read by Ethereal differs from the one read by the
K15 viewer.
 => K15 : 03/08/2005 22:50:20,907,373
 => Ethereal : 04/08/2005 00:50:20,907,373
 => So there is a 2 hours shift ; since those traces were taken in a
country which time is GMT+2, I would suggest that there is some field
holding the shift value wrongly taken into account?


If Luis or someone else could have a look at it (especially item 1), it
would be very nice ; I am willing to help and can provide several other
traces if needed.


Thanks in advance,

Yann.

Attachment: gb_merlin.rf5
Description: gb_merlin.rf5

Attachment: k15.jpg
Description: k15.jpg

Attachment: ethereal.jpg
Description: ethereal.jpg