Has anyone made a suggestion or thought about ways to distribute databases which focus on fulltext indexes?
fulltext indexes do a good job of indexing a moderate amount of data, but when you get a lot of data to be indexed, the queries slow down significantly. I have an example table, with about 90 million rows.. and has a fulltext index on a varchar(100) field. A single-word query which would return approx 300k results takes an average of 15 seconds. A query with smaller results (~ 10k) can be as quick as 1 sec.. which I would consider acceptable. Has any thought about splitting the data into distributed files or even machines? ie: something as simple as 'words' starting with 'X' are split into a-h, i-p, q-z... or something more advanced? (maybe mysqld could automatically split results based on (#results per unique 'word' / desired # of 'split files/machines') Would such a system give any advantages to searching speed and concurrenct query scalability? I haven't looked at the fulltext internals.. so I don't know if such "query routing" could take place or not. If nothing else, does anyone else have experience with a table of this size or even larger? What kind of tuning have you done? Thanks, - Brian --------------------------------------------------------------------- Before posting, please check: http://www.mysql.com/manual.php (the manual) http://lists.mysql.com/ (the list archive) To request this thread, e-mail <[EMAIL PROTECTED]> To unsubscribe, e-mail <[EMAIL PROTECTED]> Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php --------------------------------------------------------------------- Before posting, please check: http://www.mysql.com/manual.php (the manual) http://lists.mysql.com/ (the list archive) To request this thread, e-mail <[EMAIL PROTECTED]> To unsubscribe, e-mail <[EMAIL PROTECTED]> Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php
