Comments on: Top 84 MySQL Performance Tips

By: sergey

sergey — Tue, 30 Aug 2011 06:27:05 +0000

> Use Slow Query Log (always have it on!)

for what? use Slow Query Log for develop

>Don’t use COUNT * on Innodb tables for every search, do it a few times and/or summary tables, or if you need it for the total # of rows, use SQL_CALC_FOUND_ROWS and SELECT FOUND_ROWS()

Do not use SQL_CALC_FOUND_ROWS too, because it is slow (no much faster than COUNT).

> pare down cron scripts
you can solve some task with cron scripts, but at this time load increasing at server, its no good really

> Don’t use ORDER BY RAND() if you have > ~2K records
do not use ORDER BY RAND() never. You know range ID’s (last auto_increment, for ex.) and generate random value from script

By: Somnath Pawar

Somnath Pawar — Mon, 20 Jun 2011 11:04:23 +0000

Really helpful tips. Thanks.

By: Jerome

Jerome — Mon, 21 Feb 2011 10:24:29 +0000

Other tips:
- think about persistent connections
- use the right MySQL connector (e.g. PHP’s mysqli is far better than the older mysql)

By: John

John — Mon, 10 May 2010 20:22:55 +0000

Hands down, the best way to improve MySQL performance is PHP data caching. What I normally do is cache the data results from a MySQL query into a table called ‘cache’ that establishes a primary key associated with that specific data. If a user accesses that data, and the cache time hasn’t expired, I grab the data from the cache which is easily 2000 faster. (primary key lookups are a cinch)

By: Jay

Jay — Thu, 18 Mar 2010 18:04:13 +0000

Thanks. This is exactly what I was looking for.

By: macachelli

macachelli — Tue, 05 Jan 2010 00:28:21 +0000

“37. Use a clever key and ORDER BY instead of MAX”

what is a ‘clever key’?

is this just a column with a ’1′ to flag the max?

By: Jet Profiler

Jet Profiler — Thu, 01 Oct 2009 22:28:32 +0000

Thanks for a good list. Here are a few more tips:

-Use as short indexes as possible. Use smallint if you don’t need a whole int. When indexing hex data (eg md5 hash), you can store it as two BIGINTs.
-Delete data that you don’t need. Tables will grow, so monitor table sizes regularly. Use partitioning (or archive, or merge tables as mentioned above).
-Use SHOW PROFILES to get more details than EXPLAIN
-Monitor the heaviest queries, for example using Jet Profiler for MySQL

By: Avani

Avani — Wed, 28 Jan 2009 12:41:46 +0000

Its a totally helpfull material