On Fri, Jan 30, 2004 at 12:27:53PM -0500, Adam Denenberg wrote: > debug: bayes: expiry check keep size, 75% of max: 750000 > debug: bayes: token count: 2203679, final goal reduction size: 1453679 > debug: bayes: First pass? Current: 1075482160, Last: 1075479068, atime: > 1382400, count: 1019, newdelta: 969, ratio: 1426.5740922473 > debug: bayes: Can't use estimation method for expiry, something fishy, > calculating optimal atime delta (first pass) > > debug: bayes: atime token reduction > debug: bayes: ======== =============== > debug: bayes: 43200 2209986 > debug: bayes: 86400 2209986 > debug: bayes: 172800 2209932 > debug: bayes: 345600 2209932 > debug: bayes: 691200 2209932 > debug: bayes: 1382400 2209932 > debug: bayes: 2764800 2209849 > debug: bayes: 5529600 2209849 > debug: bayes: 11059200 2209849 > debug: bayes: 22118400 2209849 > > any more insight as to the problem based on this new 2.63 output?
Yep. The vast majority of your tokens are over 256 days old which means you need to do more learning before an expire can occur. SA wants to expire 1.4m tokens, but even going out to 256 days (which would reduce the least number of tokens) causes 2.2m to expire. Interestingly, you only seem to have ~140 tokens learned that are newer than 256 days. -- Randomly Generated Tagline: Borg Burgers.. How you like them is irrelivant.
pgp00000.pgp
Description: PGP signature
