[RC5] More stats oddities

David McNett nugget at slacker.com
Fri Sep 25 22:37:02 EDT 1998


On 24-Sep-1998, Phillip Chambers wrote:
> Did anyone else find their stats strange yesterday? My keyrate went
> down from about 2000kk/sec to 23kk/sec! Is this making up for something
> we missed in the past?

If by "yesterday" you mean the 22nd (which given the recent delays in
lists mail, you might) then that's because we were unable to process half
the logs for 22-Sep due to our having run out of drive space on the log
volume in the keymaster.  We're in the process of collecting the various
keyserver (full proxy) logfiles which we will use to rebuild the missing
data, and the 22-Sep numbers will be "repaired" in a few days.

Let me also take this opportunity to make a shameless plug for a new
feature we've put into place:

http://distributed.net/cgi/dnet-finger.cgi

If you ever notice anything strange with stats, it's always a good idea
to drop by and check my plan.  Tiny hiccups that may not warrant a
full-blown [ANNOUNCE] posting will be found here, as well as a general
state of the statsbox rambling.

> Also - just a general question, how is the probability of 'wrapping it
> up before tomorrow' calculated. It seems to go up and down, I would
> have thought it should be on a downward trend?

This formula is pretty much exactly what you'd expect it to be.  I start
with the total remaining blocks to check (which gets smaller every day!)

Take yesterday for instance.  Of the 68,719,476,736 blocks in the
keyspace we've checked 1,761,259,764.  This leaves 66,958,216,972 blocks
to check before we've exhausted the entire keyspace.

Since yesterday we managed to do 11,550,773 blocks, that means we can
check all the remaining blocks (at yesterday's rate) in just a hair over
5,796 days.  (66,958,216,972 / 11,550,773 == 5796.86 and some change).

We will to 1/5,796th of the remaining keyspace tomorrow (or thereabouts)
and it logically follows that we'd have a 1 in 5796 chance of finding the
key during that time.

This is a far more useful (and motivational) figure than the more
commonly referenced "time to reach 100%" because it's very unlikely that
we'll get to 100%.  (in fact, the odds are 1 in 67,719,476,736 that we'll
have to) while it is very much more exciting to recognize that we /could/
be done tomorrow.

-- 
 ________________________________________________________________________
|David McNett      |To ensure privacy and data integrity this message has|
|nugget at slacker.com|been encrypted using dual rounds of ROT-13 encryption|
|Birmingham, AL USA|Please encrypt all important correspondence with PGP!|

--
To unsubscribe, send 'unsubscribe rc5' to majordomo at lists.distributed.net
rc5-digest subscribers replace rc5 with rc5-digest



More information about the rc5 mailing list