At 12:13 PM 1/13/2007, Nurullah Akkaya wrote:
i have the following table in my database i am trying select postings
based on wordId whole table has around 200 million records.
DROP TABLE IF EXISTS `postingList`;
CREATE TABLE `postingList` (
`wordId` MEDIUMINT UNSIGNED NOT NULL,
`docId` INT UNSIGNED NOT NULL,
`docPosition` smallint UNSIGNED NOT NULL,
`bold` tinyint unsigned,
`header` tinyint unsigned,
`url` tinyint unsigned,
`anchorId` MEDIUMINT UNSIGNED,
`upperCase` tinyint unsigned,
INDEX (`wordId`),
INDEX (`docId`),
INDEX (`anchorId`)
) ENGINE=MyISAM DEFAULT CHARSET=utf8 COMMENT='list postings';
i read these in to my application using result set streaming since i
don't know how many i'll read i do not want to run out of memory i
have one thread reading results from database and two threds
processing the results read.
i can not spread reads across disks or use mysql cluster. what i was
wondering is if i squeeze those 3 tinyints into int and read just
that will it improve performance i only use mysql to store data all
calculations are done in memory are all the postings are read.
i have increased the key_buffer size but this problem occurs during
the first read so key_buffer wasn't very helpful.
Nurullah Akkaya
[EMAIL PROTECTED]
Nurullah,
If you're using MySQL 5.x, have you considered using a cursor?
http://dev.mysql.com/doc/refman/5.0/en/cursors.html
Mike
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
