Wireshark-dev: Re: [Wireshark-dev] Redback Lawful Intercept Dissector
From: "Michael A. McCartney" <[email protected]>
Date: Thu, 10 Apr 2008 10:27:50 -0500
Florian,

There is a existing bug ticket open for this:

http://bugs.wireshark.org/bugzilla/show_bug.cgi?id=2376

Thanks-Mike


Florian Lohoff wrote:
On Thu, Apr 10, 2008 at 10:21:28AM -0400, Jeff Morriss wrote:
Andrew Feren wrote:
I've recently started getting a number of false positive hits from the new
Redback Lawful Intercept heuristic.  I was going to try and tighten up the
heuristic a bit, but I can't find any sort of protocol specification.

Basically I use some protocols that start with a 32 bit version number. However since the version numers are all well below 65,535 the first two
bytes are always 0.  The Redback heuristic sees this as an end of header
marker and returns true.

My thought was to return false if the first avptype is an end of header
marker, but without a protocol spec I can't be sure that this is actually an
invalid redback packet.

Anyone have any more details?
The dissector came in via http://bugs.wireshark.org/bugzilla/show_bug.cgi?id=2320
I'm not sure if Florian is a member of this list or not.  Florian, can 
you provide some pointers?  (What about the Wiki page I asked for after 
checking in the dissector?)
    
I thought about packets beeing all zero after the patch got added 
and that might end up beeing taken by the redbackli dissector
accidentally.

I'll try to cook up a patch tonight which checks for the existance of some "essential" avp's ...
Basically the protocol is non published and i reverse engineered it
from traces. Its a packet header for forwarding lawful intercept traffic
from a RedBack Smartedge Router to some device which passes the traffic
onto some government bodies. To differentiate the different lawful
intercept session one can either use a "label" and/or a "lawful intercept
id". At least one of those two and a sequence number should be present
before an "eoh" avp ...

Attached a simple trace - the traffic is artificial which is the cause
for the udp packet encapsulated being broken ...

Flo
------------------------------------------------------------------------
_______________________________________________
Wireshark-dev mailing list
[email protected]
http://www.wireshark.org/mailman/listinfo/wireshark-dev