Metasploit mailing list archives

ie_unsafe_scripting.rb exploit module


From: hdm at metasploit.com (H D Moore)
Date: Wed, 17 Dec 2008 13:39:06 -0600

On Wednesday 17 December 2008, natron wrote:
So you have to know the server name. ?What are our options?

1) Just scan localhost for default apps running on default ports and
ignore external servers. ?(Think workstation management apps, virus
scan consoles, stuff like that.)

I agree that localhost should be included in every test, regardless of how 
we do this next part.

2) Discover through unknown external methods (like identifying their
naming scheme through some webserver information disclosure, then
generating a list of permutations... or a compromised DNS server) and
have the mod import a file.

Makes sense, lets punt this to the user and let them specify a file 
containing a list of hosts to try.

3) Pre-populate a list of guessed naming schemes.

I think we should include a default file with common server names.

How do you propose we do 3)? ?That doesn't sound easy or very
successful. ?In most environments I see, the naming schemes are all
over the map.

A few naming schemes seem really common and its something to start with at 
least. To get the ball rolling, I would suggest using a few base names and 
then permuting them based on common naming conventions:

server<suffix>
webserver<suffix>
mailserver<suffix>
client<suffix>
user<suffix>
printer<suffix>
backup<suffix>
mail<suffix>
web<suffix>
www<suffix>
intranet
hr<suffix>

With the suffix being something like:
0-9, 00-99, A-Z, AA-ZZ, -old, -new, etc

So the question becomes, at what number of permutations does that list 
become infeasible? 

-HD







Current thread: