> I am writing a serverside application in Java which needs to query/search > 200 000 rows and update affected records once every 2 minutes. > As performance/speed is of importance here - and this Table only contains > 5-10 Columns of Integer numbers, I figured that using a HEAP type of table > would be the fastest.. (right?) > The imaginary system this will run on, is a Dual CPU P4/1,5Ghz or similar > with 2Gb RAM.
Really, you will have no problem with that kind of workload. Just keep MyISAM (the default) and it's ok. You should of course tune MySQL a little (with 2 GB Ram you can set at least 256 MB for key_buffer_size, which is the most important setting). Keep in mind that MySQL doesn't cache the data by itself, it just uses the filesystem cache, so don't fill all of your RAM with the other apps. Run "top" from time to time, and check that "cached" memory is fairly used ; on my MySQL server with 2 GB RAM, the cached memory is currently ~800 MB (but my database is bigger than yours ;-PP). And don't forget to run EXPLAIN on your queries to check they use indexes correctly ;-)) However, given the figures you told us, I don't think you should worry very much.... --------------------------------------------------------------------- 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
