CREATE TABLE bayes_token ( PRIMARY KEY (id, token), INDEX bayes_token_idx1 (token), INDEX bayes_token_idx2 (id, atime) ) TYPE=MyISAM;
PRIMARY for `id` and `token` should not have INDEX for `id` and `token` added, too.
Why not? IIRC the three indexes above makes perfect sense. Like this: "WHERE id=xxx AND token=xxx" will use the primary index. "WHERE token=xxx" will use the bayes_token_idx1 index. "WHERE id=xxx AND atime=xxx" will use the bayes_token_idx2 index. Again IIRC, the clause "WHERE token=xxx" should be faster with the existance of the "bayes_token_idx1" index than without it. Or is it simply that the MySQL bayse store module never queries with "token" as the first column in a WHERE clause? I might of course completely misremenber this, so rather than trust me I'd suggest reading about index optimization or something like that in the current MySQL documentation. :-) Regards /Jonas -- Jonas Eckerman, FSDB & Fruktträdet http://whatever.frukt.org/ http://www.fsdb.org/ http://www.frukt.org/
