Re: Caching a sniffer

["Ian Latter" <Ian.Latter () mq edu au>]

> > While there's no way to be sure-sure ... you can get into your
> > local LAN segment and send ICMP(/whatever) requests to the
> > correct L3 address with the wrong L2 address and see if you
> > get a response; this will show you if hosts/devices are listening
> > promiscuously (which makes for a good starting point).
>
> Not necessarily?
>
> I thought that depended on the ip stack implementation.

Not sure what you're driving at .. do you mean you can't use the
same test on all stack implementations or that this test just won't 
work on all stack implementations?

One of the links sent through before had a link to a good read 
on the variations of the theme required for three specific 
implementations. From Tim's message you get this link;
    http://seclists.org/lists/focus-ids/2004/Feb/0028.html

In turn, gives you this link;
    http://www.securiteam.com/tools/AntiSniff_-
         _find_sniffers_on_your_local_network.html   [wr-wr-wrapped]

There they discuss NetBSD, Linux and Windows detection.

The assumption that I'm skirting around is that the sniffer is on an
existing host (pc/server/etc) .. and as such its not well prepared 
for the task; ie - that it is capable of being actively probed (that
it will respond).

I think the original post / first response included a reference to a 
site being physically accessed ... I guess that's when good 
physical access controls/records/etc become valuable.

As I said, its a good starting point (better than looking at a wiring
closet and your watch, and working out the latest time you can
order pizza).



--
Ian Latter
Internet and Networking Security Officer
Macquarie University

_______________________________________________
Full-Disclosure - We believe in it.
Charter: http://lists.netsys.com/full-disclosure-charter.html