Comments on: How to improve subqueries derived tables performance

By: Blake

Blake — Wed, 18 Jan 2012 20:33:38 +0000

Check this out… new in version 5.6 – http://dev.mysql.com/doc/refman/5.6/en/from-clause-subquery-optimization.html

By: Matt

Matt — Wed, 26 May 2010 23:10:43 +0000

Subqueries materialize in v6 and I think they moved that into 5.4 when 6 was ditched. I’m still running v6 and it works great.

By: PHP-help » Log Buffer #182, a Carnival of the Vanities for DBAs

PHP-help » Log Buffer #182, a Carnival of the Vanities for DBAs — Fri, 12 Mar 2010 22:48:26 +0000

[...] Venu Anuganti gives you tips on one of the most common MySQL frustrations: optimizing subqueries in how to improve subqueries derived tables performance. Justin Swanhart posts the way in which he Gets Linux performance information from your MySQL [...]

By: Venu Anuganti Blog » When indexes are created in internal temporary tables

Venu Anuganti Blog » When indexes are created in internal temporary tables — Mon, 08 Mar 2010 23:30:47 +0000

[...] strcmp: .. and here I thought my answe… [...]

By: strcmp

strcmp — Mon, 08 Mar 2010 08:36:54 +0000

.. and here I thought my answers made it clear I understood the post the same way.

By: Venu Anuganti

Venu Anuganti — Mon, 08 Mar 2010 01:56:14 +0000

strcmp,

the point is not about regular table indexes; all tables has indexes and they are used right. If one needs to re-write; we have enough ways to re-write and optimize the query.

But the main theme of the post is to talk about benefit of adding index on intermediate temp results and allowing an option todo so.

By: Andrii

Andrii — Sun, 07 Mar 2010 12:17:43 +0000

I bet you should first calculate sums in table `aggrpt`, then join table `sub`.
What is execution time then?

By: strcmp

strcmp — Sun, 07 Mar 2010 10:02:37 +0000

If you want the optimizer to start adding indexes for subquery results automatically, why not start automatically adding indexes to normal tables? There must be reasons why adding indexes is left to the humans usually.

In case of intermediate tables the arguments for your solution would be “the table is deleted afterwards, so no indexes aggregate”, “creating an index is slow, but it would be even slower without the index” and “all rows are created by this statement, so there is no unknown content”.

I have seen many people run into exactly the same situation, most think: There is an optimizer, so it will do this optimally, don’t you think. …