On Tue, Nov 25, 2003 at 02:45:17PM -0500, Yevgeniy Miretskiy wrote: > No matter what bayes_expiry_max_db_size is set to (I tried anything from 100K to > 3Mil), > sa-learn reports, after running for quite some time: > bayes: couldn't find a good delta atime, need more token difference, skipping > expire.
It means that you have a whole bunch of tokens with an atime within the same range (if not the same value). Either they're all old, or they're all new. There's unfortunately not too many things you can do about that. The options are: 1) learn from more messages, such that the expire algorithm has some wiggle room 2) ditch the db and start over 3) manually screw with the db -- using the berkeley db tools you can dump the db, edit the dump, then import. immediately, run "sa-learn --force-expire". you'll want to set the magic expiry values to 0. warning: I've never tried this, but it should work in theory. Unintentionally, that would be my preference order as well. -- Randomly Generated Tagline: "I decry the current tendency to seek patents on algorithms. There are better ways to earn a living than to prevent other people from making use of one's contributions to computer science." - Donald E. Knuth
pgp00000.pgp
Description: PGP signature
