Re: New beta stats.

Karl J. Runge (runge@crl.com)
Sun, 25 May 1997 19:06:39 -0700


On Sun, 25 May 1997, "Douglas M. Waer" <dwaer@primenet.com> wrote:
>
> I'm still not convinced that the new beta stats are even close to correct.
> My own logs show that I did approx 49,000*2^20 keys, while the level III
> domain stats show that phx.primenet.com only processed 14000*2^20. What's
> the deal?

Sorry for the delay, this is related to something I already knew about, I
didn't have my cron job delete a gzip file, which led the program to
using the *previous* day's DNS hash table.

But you could have found the other phx.primenet.com's from their numerical
values:

14856 * 2^20 6 phx.primenet.com
12168 * 2^20 8 206.165.20
7328 * 2^20 6 206.165.21
9864 * 2^20 4 206.165.22
272 * 2^20 1 206.165.23
14984 * 2^20 1 206.165.25

Which is about 59,000*2^20 keys, a bit more than your estimate, hope it
is reasonable.

I will publish the stats using the updated DNS hash table in the next
half hour or so.

The good news is that the reverse DNS scheme is somewhat better now,
I've found a fast way to do that (just run 128 processes!). So now each
night the program tries to update *all* IP's that had not previously
been resolved in addition to the new IP's for the day. It is surprising
the number that if you keep trying will ultimately be resolved.

> --
> Doug Waer Finger dwaer@primenet.com for PGP key
> Arizona Rattler's web page: http://www.primenet.com/~dwaer/rattlers.html
>
> -------------------------------------------------------------------------
> "You lie so much you believe yourself..." - Metallica

It has been said that statisticians are liars, even w/o program bugs ;-)