In article <4F5EA586.CCCF6F71@Guy.com>, Virus@Guy.com says...
>
> Whoever wrote:
>
> > > So that can happen for any html content being requested by these
> > > infected PC's.
> >
> > Yes, it is simple for a web server to modify the displayed results
> > on the fly.
>
> A concept that went right over the heads of a lot of people here. Or at
> least the reason why you'd want to do that in this situation.
>
> > The problem that I'm having with understanding your scenario
> > is just how a DNS server will "tag" it's response to a specific
> > client
>
> Because the only clients hitting this DNS server are the ones infected
> with some specific malware. The PC's hitting this DNS server are part
> of a botnet that the fed's took down last year. They are the only PC's
> using a special DNS server that was set up to replace a malicious
> server.
>
> And because they're using this special server, the authorities and
> white-hats know the rate at which these computers are getting cleaned up
> because they monitor the traffic hitting this server. As machines get
> cleaned up, they stop using this special DNS server and they use
> what-ever is appropriate for them (their isp's server, etc).
I understood all of that perfectly well but it has nothing to do with
my question. Perhaps I didn't state it well. I'll try to reword it to
make it clearer.
As I understand it, you are describing two entirely separate
transactions using the internet. The first one is a request to a DNS
server to resolve a URL to an IP address. The IP address of the DNS
server itself is already known and set in the compromised computer. In
your example it was changed to 1.2.3.4 by the DNSChanger to form the
botnet. So the compromised computer sends a request to 1.2.3.4
(assumedly on port 53) to resolve the URL www.acme.com to an IP address.
The DNS server then returns 1.2.3.4 (in your example) as the IP address
for www.acme.com to the compromised computer. The compromised computer
then opens a completely separate request to 1.2.3.4 (assumedly on port
80) looking for the web server.
Here is where I'm having trouble understanding what you are suggesting.
How does that web server _know_ that this particular request is
expecting to receive the web page actually hosted at www.acme.com? You
seem to be suggesting that each response from the DNS server is somehow
"tagged" to identify the desired URL (www.acme.com) back to the
compromised computer. As far as I know, a DNS server cannot do that.
Even if you hacked the server to append such "tag data" onto the
response (i.e. "1.2.3.4/?host=www.acme.com") the compromised computer
wouldn't know what to do with the "extra" data and would not be able to
use it. Perhaps I'm wrong though. I don't know that much about the
internal workings of DNS clients and it's been a long time since I
looked over the RFC's for DNS resolution.
Even if you could do such a thing and get it to somehow work for web
pages, I have serious reservations about how other apps would react to
that solution. For instance, when you're using DNS to resolve for things
like time servers, IM servers, email servers, NNTP servers, update
servers of all sorts, etc. Do you just treat them all as if they were
web page address requests?
--
Whoever - but you can just call me who.
whoever@wherever.invalid


Reply With Quote
