Comments on: How is MySQL’s join performance these days? http://www.dbms2.com/2008/07/10/how-is-mysqls-join-performance-these-days/ Choices in data management and analysis Thu, 09 Feb 2012 16:57:09 +0000 hourly 1 http://wordpress.org/?v=3.0.3 By: Quora http://www.dbms2.com/2008/07/10/how-is-mysqls-join-performance-these-days/#comment-228568 Quora Thu, 23 Jun 2011 09:59:53 +0000 http://www.dbms2.com/?p=461#comment-228568 <strong>Why do people typically choose PostgreSQL over MySQL?...</strong> First off, I believe that whatever you're doing, you'll probably be fine using either. That said, I'm generally more frustrated when using MySQL than when using PG. Here's a sample of the problems I've encountered from using both MySQL and PG. I h... Why do people typically choose PostgreSQL over MySQL?…

First off, I believe that whatever you’re doing, you’ll probably be fine using either. That said, I’m generally more frustrated when using MySQL than when using PG. Here’s a sample of the problems I’ve encountered from using both MySQL and PG. I h…

]]> By: Quora http://www.dbms2.com/2008/07/10/how-is-mysqls-join-performance-these-days/#comment-228567 Quora Thu, 23 Jun 2011 09:48:59 +0000 http://www.dbms2.com/?p=461#comment-228567 <strong>What are the advantages and disadvantages of using PostgreSQL over MySQL?...</strong> First off, I believe that whatever you're doing, you'll probably be fine using either. That said, I'm generally more frustrated when using MySQL than when using PG. Here's a sample of the problems I've encountered from using both MySQL and PG. I h... What are the advantages and disadvantages of using PostgreSQL over MySQL?…

First off, I believe that whatever you’re doing, you’ll probably be fine using either. That said, I’m generally more frustrated when using MySQL than when using PG. Here’s a sample of the problems I’ve encountered from using both MySQL and PG. I h…

]]> By: shamim http://www.dbms2.com/2008/07/10/how-is-mysqls-join-performance-these-days/#comment-91268 shamim Mon, 21 Jul 2008 18:35:59 +0000 http://www.dbms2.com/?p=461#comment-91268 Really true. i am satisfied reading your article. I can do well. if you get me scholarship. i want to do Msc in software Engineering. i have strong eagernesses,but nobody want help me in this world. Really true. i am satisfied reading your article.
I can do well. if you get me scholarship. i want to do Msc in software Engineering. i have strong eagernesses,but nobody want help me in this world.

]]> By: Curt Monash http://www.dbms2.com/2008/07/10/how-is-mysqls-join-performance-these-days/#comment-90690 Curt Monash Tue, 15 Jul 2008 15:47:25 +0000 http://www.dbms2.com/?p=461#comment-90690 And since it's open source, you can write your own modules even beyond the default cases for doing so. Duh. Thanks! CAM And since it’s open source, you can write your own modules even beyond the default cases for doing so. Duh.

Thanks!

CAM

]]> By: Victoria Eastwood http://www.dbms2.com/2008/07/10/how-is-mysqls-join-performance-these-days/#comment-90682 Victoria Eastwood Tue, 15 Jul 2008 12:48:30 +0000 http://www.dbms2.com/?p=461#comment-90682 If you implement only the storage engine interface then you would using the MySQL optimizer. This interface is row oriented and expects indices, which are used among other things to determine the costing. Since Brighthouse is both column oriented and uses the knowledge grid instead of indices, this type of optimizer is not a good fit for us. So although we have implemented this interface and use it in rare cases, we have written our own optimizer to take advantage of the knowledge grid during query execution. If you implement only the storage engine interface then you would using the MySQL optimizer. This interface is row oriented and expects indices, which are used among other things to determine the costing.

Since Brighthouse is both column oriented and uses the knowledge grid instead of indices, this type of optimizer is not a good fit for us. So although we have implemented this interface and use it in rare cases, we have written our own optimizer to take advantage of the knowledge grid during query execution.

]]> By: Curt Monash http://www.dbms2.com/2008/07/10/how-is-mysqls-join-performance-these-days/#comment-90671 Curt Monash Mon, 14 Jul 2008 16:35:06 +0000 http://www.dbms2.com/?p=461#comment-90671 Victoria (or anybody), I probably knew the answer to this, but I've forgotten: When you write your own MySQL storage engine, what happens with the optimizer? Do you do one from scratch? Does MySQL have a basic cost-based optimizer for which you write your own cost functions? Thanks, CAM Victoria (or anybody),

I probably knew the answer to this, but I’ve forgotten:

When you write your own MySQL storage engine, what happens with the optimizer? Do you do one from scratch? Does MySQL have a basic cost-based optimizer for which you write your own cost functions?

Thanks,

CAM

]]> By: Victoria Eastwood http://www.dbms2.com/2008/07/10/how-is-mysqls-join-performance-these-days/#comment-90670 Victoria Eastwood Mon, 14 Jul 2008 15:26:23 +0000 http://www.dbms2.com/?p=461#comment-90670 Some storage engines, including Infobright’s, implement their own join algorithms. Infobright implements two join techniques: nested loop and sort-merge. We are also in the process of implementing hash-join. Since we don’t use indices but rather the Knowledge Grid, our implementation of these techniques is different and the performance dynamic much improved. We also implement subqueries using the Knowledge Grid; so we significantly out perform other MySQL storage engines on large datasets. And our clients consistently use joins and subqueries in their queries. Some storage engines, including Infobright’s, implement their own join algorithms. Infobright implements two join techniques: nested loop and sort-merge. We are also in the process of implementing hash-join. Since we don’t use indices but rather the Knowledge Grid, our implementation of these techniques is different and the performance dynamic much improved.

We also implement subqueries using the Knowledge Grid; so we significantly out perform other MySQL storage engines on large datasets. And our clients consistently use joins and subqueries in their queries.

]]> By: Justin Swanhart http://www.dbms2.com/2008/07/10/how-is-mysqls-join-performance-these-days/#comment-90476 Justin Swanhart Sat, 12 Jul 2008 06:55:35 +0000 http://www.dbms2.com/?p=461#comment-90476 KFDB (the storage engine used in the Kickfire appliances) supports multiple types of joins beyond nested loops, but only for tables backed by the KFDB storage engine. Kickfire supports hash joins as well as additional join methods that leverage hardware acceleration. Kickfire also features accelerated joins to specialized indexes in addition to the performance benefits garnered by vertical column storage. Join performance may be orders of magnitude faster than other storage engines when using specialized join algorithms. KFDB (the storage engine used in the Kickfire appliances) supports multiple types of joins beyond nested loops, but only for tables backed by the KFDB storage engine.

Kickfire supports hash joins as well as additional join methods that leverage hardware acceleration. Kickfire also features accelerated joins to specialized indexes in addition to the performance benefits garnered by vertical column storage. Join performance may be orders of magnitude faster than other storage engines when using specialized join algorithms.

]]>