Re: [NSE] script idea: identify ports behind a NAT

["DePriest, Jason R." <jrdepriest () gmail com>]

On Wed, Mar 17, 2010 at 10:56 AM, jah <> wrote:
> On 17/03/2010 14:31, Ron wrote:
> > I just had an idea for a useful script that I don't really have time to write. Maybe somebody else does?
> >
> > Basically, identify and group which ports on a NAT point at the same computer. So, if I have port 22, 80, and 443 
> > forwarded to computer A, and 21, 445 forwarded to computer B, I'd like to be able to tell that.
> >
> > There are probably a few different ways, and it really comes down to the same techniques used for OS fingerprinting 
> > (and some limited intelligence), but I think the easiest way initially would be to look at the IPIDs, especially for 
> > incremental systems.
> >
> > Any other ideas?
> Doug Hoyte created a patch for Nmap which introduced a scan type he
> named Qscan. It did detection based on timing (grouping ports by similar
> round-trip times) and worked well.  The patch was never integrated and
> Marek Majkowski suggested it might be a job for NSE back in '07:
> http://seclists.org/nmap-dev/2007/q3/63
>
> It's definitely a good idea.  Someone's just got to write it...
>
> jah

I was manually tweaking the Qscan patch for each version of nmap up to
around 4.50 at which point I changed jobs and stopped having a need
for it.

Strangely enough, I am now back at the job I had before I changed jobs.

I like Qscan and would also be happy to see it officially incorporated.

I am not volunteering to write it mainly because I have no idea how to
do so in a portable way, but hopefully someone else will.

-Jason
_______________________________________________
Sent through the nmap-dev mailing list
http://cgi.insecure.org/mailman/listinfo/nmap-dev
Archived at http://seclists.org/nmap-dev/