June 19, 2013

Post: How number of columns affects performance ?

to be slower than tables with short rows. I was interested to check… ) ENGINE=…onmysql [localhost] {msandbox} (test) > explain select max(t1+0) from t99v1; +—-+————-+——-+——-+—————+——+———+——+———-+————-+ | id | select_type | table | typeto how

Post: Connecting orphaned .ibd files

ENGINE=InnoDB DEFAULT CHARSET=latin1 Please note field SPACE. For table actor it is equal to 15: mysql> selectTYPE 2 current defined typesTo refrain from such error ibdconnect does check if the space_id is already used. It will refuse to

Post: Falcon Storage Engine Design Review

on multiple CPU systems you need some spinning done, at least for some typesto see how small indexes really are. [-] No “using index” Falcon always have to checkMySQL Storage Engine level for MySQL – reading rows in their physical order can be applied to most storage engines…support for SELECT FOR …

Post: Using CHAR keys for joins, how much is the overhead ?

…) ENGINE=MyISAM DEFAULT CHARSET=utf8 mysql> selecton t1.i=t2.j \G *************************** 1. row *************************** id: 1 select_type: SIMPLE table: t1 typeto 3.5 seconds which is just 20% slower than integer based join for Innodb. Finally I decided to checkto integer based join. So how

Post: Find and remove duplicate indexes

MySQL needs to examine more query plans. The storage engine needs toto explain the different types of duplicate indexes and how to find and remove them. Duplicate keys onSELECT * FROM t WHERE name=’kahxailo’ AND i > 100000; With a index on

Post: MySQL 6.0 vs 5.1 in TPC-H queries

ENGINE=MyISAM DEFAULT CHARSET=latin1 with 59986052 rows So this query takes 32.27 sec toselect_type | table | typeto 1200 finally. Probably we need to check how really MRR works. What else worth to mention here is that MySQL

Post: MySQL Slow query log in the table

MySQL 5.1 get MySQL slow query log logged in mysqlto change table to other storage engineto get new queries in the end you can do SELECT * FROM (SELECTto MyISAM, disable concurrent inserts and run long reporting query on the log file query execution is still proceeds. I have not checked

Post: Hijacking Innodb Foreign Keys

on rename of ‘./test/child’ to ‘./test/#sql2-496-40a5′ (errno: 152)” in many cases and foreign keys are still handled on storage enginemysql> set foreign_key_checks=1; Query OK, 0 rows affected (0.00 sec) mysql> selecthow to

Post: Logging Deadlock errors

ENGINE INNODB STATUS and log the new deadlocks to a file or to a table that we can later review. Let’s see howcheck the deadlock information in our table: mysql> select…_typeon every node and configure it to store logs on a centralized database. In MySQL 5.6 will be possible to log all deadlocks on

Post: Distributed Set Processing with Shard-Query

… you want but not how to get it. Most distributed engines work with rows. They …to distribute work over any type of compute resource which speaks SQL, but right now only MySQLselected for a UNIQUE CHECK * on the base table: `origin_airport_id` * storage node result set merge optimization enabled: ON