Re: 2 forwarded messages...DNSEXT discussion of Day and Kaminsky

On Wed, Feb 25, 2009 at 12:53:03AM -0500, Dean Anderson wrote:

> > With stock dnscache, each of the X+N active queries opens a new port. So
> > the chances of hitting a port which belongs to dnscache and matches our
> > Q is X/(X+N+O). The chances of hitting the queryid correctly for that
> > port are 2^-16. So
> > 
> >   P = X / (X+N+O) * 2^-16
> 
> How did you arrive at that formula?

It is the AND of two independent probabilities, given in the paragraph
above.

> (%i55) n: 0; x: 200; o: 100;
> (%o55) 0
> (%o56) 200
> (%o57) 100
> (%i58) p: x / (x + n + o) * 2^16;
> (%o58) 131072/3
> (%i59) float(p);
> (%o59) 43690.66666666666
> 
> Hmm. I thought probabilities ranged from 0 to 1.

It would probably help if you typed my formula correctly (hint, the
probability of guessing the query id is not the nonsensical 2^16, but
rather 2^-16).

> After I get my distribution of DNScache done, I'll post what I think to
> be the correct math on the site.

I look forward to seeing it.

-Peff