Educause Security Discussion mailing list archives

Re: IRC, IM Proxy Implementations


From: Mark Wilson <wilsodm () AUBURN EDU>
Date: Wed, 8 Sep 2004 08:52:04 -0500

Well, 100% of the boxes we have found with this sig have been
compromised. However, your point is well taken.  I would be interested
in others experiences.

JAzoff () UAMAIL ALBANY EDU 9/8/2004 8:38:37 AM >>>
Mark Wilson wrote:
Concerning port 113, regular scans of our network for port 113 has
uncovered many bots.  One "tool" you may wish to use is expect.  I
have
written an expect script that telnets into port 113 and performs a
<CR>
to get the familiar:

spawn telnet 131.204.x.x 113
Trying 131.204.x.x ...
Connected to 131.204.x.x.
Escape character is '^]'.

 : USERID : UNIX : ggdmlnfa
^]
This confirms PC is Bot-ed.

After scanning port 113, dump the IPs (with port 113 open) to a
file.
The expect script reads the IP file to "automate" the process.

Mark Wilson
GCIA, CISSP #53153
Network Security Specialist
Auburn University
(334) 844-9347

This is very similar to what my script does, I wrote a python wrapper
to
nmap, and then a module called "banners" which connects to each port
and
sends \n\n, then reads in the response.  Then another module has a
list
of bad banners.  Any host with a bad banner, gets its port disabled,
and
a ticket created.

Question for you though :-)  Right now the "USERID : UNIX" is not set
as
a bad banner, as I wasn't sure if any ligitimate irc client's ident
server had that signature.  Has using that criteria picked up any
false
positives for you?

--
-- Justin Azoff
-- Network Performance Analyst

**********
Participation and subscription information for this EDUCAUSE Discussion
Group discussion list can be found at http://www.educause.edu/cg/.

**********
Participation and subscription information for this EDUCAUSE Discussion Group discussion list can be found at 
http://www.educause.edu/cg/.

Current thread: