One of the serious limitations in the fast index creation feature introduced in the InnoDB plugin is that it only works when indexes are explicitly created using ALTER TABLE or CREATE INDEX. Peter has already blogged about it before, here I’ll just briefly reiterate other cases that might benefit from that feature: when ALTER TABLE [...]
Shard-Query EC2 images available
May 11, 2011 By 2 Comments
Infobright and InnoDB AMI images are now available There are now demonstration AMI images for Shard-Query. Each image comes pre-loaded with the data used in the previous Shard-Query blog post. The data in the each image is split into 20 “shards”. This blog post will refer to an EC2 instances as a node from here [...]
Recover BLOB fields
July 1, 2010 By 4 Comments
For a long time long types like BLOB, TEXT were not supported by Percona InnoDB Recovery Tool. The reason consists in a special way InnoDB stores BLOBs. An InnoDB table is stored in a clustered index called PRIMARY. It must exist even if a user hasn’t defined the primary index. The PRIMARY index pages are [...]
A workaround for the performance problems of TEMPTABLE views
May 19, 2010 By 7 Comments
MySQL supports two different algorithms for views: the MERGE algorithm and the TEMPTABLE algorithm. These two algorithms differ greatly. A view which uses the MERGE algorithm can merge filter conditions into the view query itself. This has significant performance advantages over TEMPTABLE views. A view which uses the TEMPTABLE algorithm will have to compute the [...]
Analyzing air traffic performance with InfoBright and MonetDB
October 2, 2009 By 50 Comments
Accidentally me and Baron played with InfoBright (see http://www.mysqlperformanceblog.com/2009/09/29/quick-comparison-of-myisam-infobright-and-monetdb/) this week. And following Baron’s example I also run the same load against MonetDB. Reading comments to Baron’s post I tied to load the same data to LucidDB, but I was not successful in this. I tried to analyze a bigger dataset and I took public [...]
Multi Column indexes vs Index Merge
September 19, 2009 By 31 Comments
The mistake I commonly see among MySQL users is how indexes are created. Quite commonly people just index individual columns as they are referenced in where clause thinking this is the optimal indexing strategy. For example if I would have something like AGE=18 AND STATE=’CA’ they would create 2 separate indexes on AGE and STATE [...]
High-Performance Click Analysis with MySQL
December 22, 2008 By 13 Comments
We have a lot of customers who do click analysis, site analytics, search engine marketing, online advertising, user behavior analysis, and many similar types of work. The first thing these have in common is that they’re generally some kind of loggable event. The next characteristic of a lot of these systems (real or planned) is [...]
Picking datatype for STATUS fields
August 9, 2008 By 24 Comments
Quite commonly in the applications you would need to use some kind of “status” field – status of order – “new”, “confirmed”, “in production”, “shipped” status of job, message etc. People use variety of ways to handle them often without giving enough thought to the choice which can cause problems later. Perhaps worst, though quite [...]
Testing InnoDB “Barracuda” format with compression
April 23, 2008 By 3 Comments
New features of InnoDB – compression format and fast index creation sound so promising so I spent some time to research time and sizes on data we have on our production. The schema of one of shards is
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 | CREATE TABLE `article87` ( `id` bigint(20) unsigned NOT NULL, `ext_key` varchar(32) NOT NULL, `site_id` int(10) unsigned NOT NULL, `forum_id` int(10) unsigned NOT NULL, `thread_id` varchar(255) CHARACTER SET latin1 NOT NULL, `published` timestamp NOT NULL DEFAULT '0000-00-00 00:00:00', `crawled` timestamp NOT NULL DEFAULT '0000-00-00 00:00:00', `subject` varchar(255) NOT NULL, `title` varchar(255) NOT NULL, `url` varchar(255) NOT NULL, `num_links` smallint(6) NOT NULL, `links_in` int(10) unsigned NOT NULL, `cache_author` varchar(255) NOT NULL, `cache_site` varchar(255) DEFAULT NULL, `anchor` varchar(255) NOT NULL, `isthread` tinyint(3) unsigned NOT NULL, `author_id` int(10) unsigned NOT NULL, `inserted` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP, `fromfile` varchar(255) NOT NULL, `language_id` tinyint(3) unsigned NOT NULL, `encoding` varchar(255) NOT NULL, `warning` mediumtext NOT NULL, `is_thread_start` tinyint(3) unsigned NOT NULL, `source` mediumint(8) unsigned NOT NULL, `hash` char(32) NOT NULL, `mod_is` tinyint(3) unsigned NOT NULL DEFAULT '0', `is_adult` tinyint(3) unsigned NOT NULL DEFAULT '0', `bodyuc` mediumtext NOT NULL, PRIMARY KEY (`id`), KEY `ext_key` (`ext_key`), KEY `forum_id` (`forum_id`,`thread_id`,`published`), KEY `site_id` (`site_id`,`published`), KEY `hash` (`hash`), KEY `forum_id_2` (`forum_id`,`is_thread_start`,`published`), KEY `published` (`published`), KEY `inserted` (`inserted`), KEY `forum_id_3` (`forum_id`,`thread_id`,`is_thread_start`), KEY `site_id_2` (`site_id`,`author_id`) ) ENGINE=InnoDB DEFAULT CHARSET=latin1; |
Enum Fields VS Varchar VS Int + Joined table: What is Faster?
January 24, 2008 By 30 Comments
Really often in customers’ application we can see a huge tables with varchar/char fields, with small sets of possible values. These are “state”, “gender”, “status”, “weapon_type”, etc, etc. Frequently we suggest to change such fields to use ENUM column type, but is it really necessary (from performance standpoint)? In this post I’d like to present [...]
