MySQL高级 第9章关联查询优化

第9章关联查询优化

1. 建表语句

CREATE TABLE IF NOT EXISTS `class` ( `id` INT(10) UNSIGNED NOT NULL AUTO_INCREMENT, `card` INT(10) UNSIGNED NOT NULL, PRIMARY KEY (`id`) ); CREATE TABLE IF NOT EXISTS `book` ( `bookid` INT(10) UNSIGNED NOT NULL AUTO_INCREMENT, `card` INT(10) UNSIGNED NOT NULL, PRIMARY KEY (`bookid`) );   INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO class(card) VALUES(FLOOR(1 + (RAND() * 20)));   INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20))); INSERT INTO book(card) VALUES(FLOOR(1 + (RAND() * 20)));

 

2. 案例

2.1 left join

①EXPLAIN SELECT * FROM class LEFT JOIN book ON class.card = book.card;

②如何优化？在哪个表上建立索引？

ALTER TABLE `book` ADD INDEX  idx_card( `card`);

③删除book表的索引：drop index idx_card on book;

在class表上建立索引：alter table class add index idx_card(card);

结论：

①在优化关联查询时，只有在被驱动表上建立索引才有效！

②left join时，左侧的为驱动表，右侧为被驱动表！

 

2.2 inner join

①EXPLAIN SELECT * FROM book inner join class on class.card=book.card;

②两个查询字段调换顺序，发现结果也是一样的！

③在book表中，删除9条记录

④结论：inner join 时，mysql会自己帮你把小结果集的表选为驱动表。

⑤straight_join: 效果和inner join一样，但是会强制将左侧作为驱动表！

2.3 四个关联查询案例分析

EXPLAIN SELECT ed.name '人物',c.name '掌门' FROM (SELECT e.name,d.ceo from t_emp e LEFT JOIN t_dept d on e.deptid=d.id) ed  LEFT JOIN t_emp c on ed.ceo= c.id;

 

上述两个案例，第一个查询效率较高，且有优化的余地。第二个案例中，子查询作为被驱动表，由于子查询是虚表，无法建立索引，因此不能优化。

 

结论：

子查询尽量不要放在被驱动表，有可能使用不到索引；

left join时，尽量让实体表作为被驱动表。

 

本教程由尚硅谷教育大数据研究院出品，如需转载请注明来源。