MySQL索引优化
索引优化
1、单表索引优化
单表索引优化分析
范围值容易导致索引失效。
创建表
- 建表 SQL
CREATE TABLE IF NOT EXISTS article(
id INT(10) UNSIGNED NOT NULL PRIMARY KEY AUTO_INCREMENT,
author_id INT(10) UNSIGNED NOT NULL,
category_id INT(10) UNSIGNED NOT NULL,
views INT(10) UNSIGNED NOT NULL,
comments INT(10) UNSIGNED NOT NULL,
title VARCHAR(255) NOT NULL,
content TEXT NOT NULL
);
INSERT INTO article(author_id,category_id,views,comments,title,content)
VALUES
(1,1,1,1,'1','1'),
(2,2,2,2,'2','2'),
(1,1,3,3,'3','3');- 表中的测试数据
mysql> SELECT * FROM article;
+----+-----------+-------------+-------+----------+-------+---------+
| id | author_id | category_id | views | comments | title | content |
+----+-----------+-------------+-------+----------+-------+---------+
| 1 | 1 | 1 | 1 | 1 | 1 | 1 |
| 2 | 2 | 2 | 2 | 2 | 2 | 2 |
| 3 | 1 | 1 | 3 | 3 | 3 | 3 |
+----+-----------+-------------+-------+----------+-------+---------+
3 rows in set (0.00 sec)查询案例
- 查询category_id为1且comments 大于1的情况下，views最多的article_id。
mysql> SELECT id, author_id FROM article WHERE category_id = 1 AND comments > 1 ORDER BY views DESC LIMIT 1;
+----+-----------+
| id | author_id |
+----+-----------+
| 3 | 1 |
+----+-----------+
1 row in set (0.00 sec)- 此时 article 表中只有一个主键索引
mysql> SHOW INDEX FROM article;
+---------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| Table | Non_unique | Key_name | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | Index_comment |
+---------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| article | 0 | PRIMARY | 1 | id | A | 3 | NULL | NULL | | BTREE | | |
+---------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
1 row in set (0.00 sec)- 使用 explain 分析 SQL 语句的执行效率：
EXPLAIN SELECT id, author_id FROM article WHERE category_id = 1 AND comments > 1 ORDER BY views DESC LIMIT 1;
mysql> EXPLAIN SELECT id, author_id FROM article WHERE category_id = 1 AND comments > 1 ORDER BY views DESC LIMIT 1;
+----+-------------+---------+------+---------------+------+---------+------+------+-----------------------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+---------+------+---------------+------+---------+------+------+-----------------------------+
| 1 | SIMPLE | article | ALL | NULL | NULL | NULL | NULL | 3 | Using where; Using filesort |
+----+-------------+---------+------+---------------+------+---------+------+------+-----------------------------+
1 row in set (0.00 sec)- 结论：
- 很显然，type是ALL，即最坏的情况。
- Extra 里还出现了Using filesort，也是最坏的情况。
- 优化是必须的。
开始优化：新建索引
- 创建索引的 SQL 命令
# ALTER TABLE article ADD INDEX idx_article_ccv('category_id', 'comments', 'views');
create index idx_article_ccv on article(category_id, comments, views);- 在 category_id 列、comments 列和 views 列上建立联合索引
mysql> create index idx_article_ccv on article(category_id, comments, views);
Query OK, 0 rows affected (0.01 sec)
Records: 0 Duplicates: 0 Warnings: 0
mysql> SHOW INDEX FROM article;
+---------+------------+-----------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| Table | Non_unique | Key_name | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | Index_comment |
+---------+------------+-----------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| article | 0 | PRIMARY | 1 | id | A | 3 | NULL | NULL | | BTREE | | |
| article | 1 | idx_article_ccv | 1 | category_id | A | 3 | NULL | NULL | | BTREE | | |
| article | 1 | idx_article_ccv | 2 | comments | A | 3 | NULL | NULL | | BTREE | | |
| article | 1 | idx_article_ccv | 3 | views | A | 3 | NULL | NULL | | BTREE | | |
+---------+------------+-----------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
4 rows in set (0.00 sec)- 再次执行查询：type变成了range，这是可以忍受的。但是extra里使用Using filesort仍是无法接受的。
mysql> EXPLAIN SELECT id, author_id FROM article WHERE category_id = 1 AND comments > 1 ORDER BY views DESC LIMIT 1;
+----+-------------+---------+-------+-----------------+-----------------+---------+------+------+---------------------------------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+---------+-------+-----------------+-----------------+---------+------+------+---------------------------------------+
| 1 | SIMPLE | article | range | idx_article_ccv | idx_article_ccv | 8 | NULL | 1 | Using index condition; Using filesort |
+----+-------------+---------+-------+-----------------+-----------------+---------+------+------+---------------------------------------+
1 row in set (0.00 sec)-
分析：
-
但是我们已经建立了索引，为啥没用呢？
-
这是因为按照B+Tree索引的工作原理，先排序 category_id，如果遇到相同的 category_id 则再排序comments，如果遇到相同的 comments 则再排序 views。
-
当comments字段在联合索引里处于中间位置时，因为
comments>1条件是一个范围值（所谓 range），MySQL 无法利用索引再对后面的views部分进行检索，即 range 类型查询字段后面的索引无效。 -
将查询条件中的
comments > 1改为comments = 1，发现 Use filesort 神奇地消失了，从这点可以验证：范围后的索引会导致索引失效
-
mysql> EXPLAIN SELECT id, author_id FROM article WHERE category_id = 1 AND comments = 1 ORDER BY views DESC LIMIT 1;
+----+-------------+---------+------+-----------------+-----------------+---------+-------------+------+-------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+---------+------+-----------------+-----------------+---------+-------------+------+-------------+
| 1 | SIMPLE | article | ref | idx_article_ccv | idx_article_ccv | 8 | const,const | 1 | Using where |
+----+-------------+---------+------+-----------------+-----------------+---------+-------------+------+-------------+
1 row in set (0.00 sec)删除索引
- 删除索引的 SQL 指令
DROP INDEX idx_article_ccv ON article;- 删除刚才创建的 idx_article_ccv 索引
mysql> DROP INDEX idx_article_ccv ON article;
Query OK, 0 rows affected (0.00 sec)
Records: 0 Duplicates: 0 Warnings: 0
mysql> SHOW INDEX FROM article;
+---------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| Table | Non_unique | Key_name | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | Index_comment |
+---------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| article | 0 | PRIMARY | 1 | id | A | 3 | NULL | NULL | | BTREE | | |
+---------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
1 row in set (0.00 sec)再次创建索引
- 创建索引的 SQL 指令
# ALTER TABLE article ADD INDEX idx_article_ccv('category_id', 'views');
create index idx_article_ccv on article(category_id, views);- 由于 range 后（
comments > 1）的索引会失效，这次我们建立索引时，直接抛弃 comments 列，先利用 category_id 和 views 的联合索引查询所需要的数据，再从其中取出comments > 1的数据（我觉着应该是这样的）
mysql> create index idx_article_ccv on article(category_id, views);
Query OK, 0 rows affected (0.30 sec)
Records: 0 Duplicates: 0 Warnings: 0
mysql> SHOW INDEX FROM article;
+---------+------------+-----------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| Table | Non_unique | Key_name | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | Index_comment |
+---------+------------+-----------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| article | 0 | PRIMARY | 1 | id | A | 3 | NULL | NULL | | BTREE | | |
| article | 1 | idx_article_ccv | 1 | category_id | A | 3 | NULL | NULL | | BTREE | | |
| article | 1 | idx_article_ccv | 2 | views | A | 3 | NULL | NULL | | BTREE | | |
+---------+------------+-----------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
3 rows in set (0.00 sec)- 再次执行查询：可以看到，type变为了ref，Extra中的Using filesort也消失了，结果非常理想
ysql> EXPLAIN SELECT id, author_id FROM article WHERE category_id = 1 AND comments > 1 ORDER BY views DESC LIMIT 1;
+----+-------------+---------+------+-----------------+-----------------+---------+-------+------+-------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+---------+------+-----------------+-----------------+---------+-------+------+-------------+
| 1 | SIMPLE | article | ref | idx_article_ccv | idx_article_ccv | 4 | const | 2 | Using where |
+----+-------------+---------+------+-----------------+-----------------+---------+-------+------+-------------+
1 row in set (0.00 sec)- 为了不影响之后的测试，删除该表的 idx_article_ccv 索引
mysql> DROP INDEX idx_article_ccv ON article;
Query OK, 0 rows affected (0.05 sec)
Records: 0 Duplicates: 0 Warnings: 0
mysql> SHOW INDEX FROM article;
+---------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| Table | Non_unique | Key_name | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | Index_comment |
+---------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| article | 0 | PRIMARY | 1 | id | A | 3 | NULL | NULL | | BTREE | | |
+---------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
1 row in set (0.01 sec)2、两表索引优化
两表索引优化分析：主外键
主要是两表连接查询：用小表驱动大表，大表索引优化越大，越快。
创建表
- 建表 SQL
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 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)));- class 表中的测试数据
mysql> select * from class;
+----+------+
| id | card |
+----+------+
| 1 | 12 |
| 2 | 13 |
| 3 | 12 |
| 4 | 17 |
| 5 | 11 |
| 6 | 3 |
| 7 | 1 |
| 8 | 16 |
| 9 | 17 |
| 10 | 16 |
| 11 | 9 |
| 12 | 17 |
| 13 | 18 |
| 14 | 16 |
| 15 | 7 |
| 16 | 8 |
| 17 | 19 |
| 18 | 9 |
| 19 | 6 |
| 20 | 5 |
| 21 | 6 |
+----+------+
21 rows in set (0.00 sec)- book 表中的测试数据
mysql> select * from book;
+--------+------+
| bookid | card |
+--------+------+
| 1 | 16 |
| 2 | 1 |
| 3 | 17 |
| 4 | 3 |
| 5 | 20 |
| 6 | 12 |
| 7 | 18 |
| 8 | 13 |
| 9 | 13 |
| 10 | 4 |
| 11 | 1 |
| 12 | 13 |
| 13 | 20 |
| 14 | 20 |
| 15 | 1 |
| 16 | 2 |
| 17 | 9 |
| 18 | 16 |
| 19 | 14 |
| 20 | 2 |
+--------+------+
20 rows in set (0.00 sec)查询案例
- 实现两表的连接，连接条件是 class.card = book.card
mysql> SELECT * FROM class LEFT JOIN book ON class.card = book.card;
+----+------+--------+------+
| id | card | bookid | card |
+----+------+--------+------+
| 1 | 12 | 6 | 12 |
| 2 | 13 | 8 | 13 |
| 2 | 13 | 9 | 13 |
| 2 | 13 | 12 | 13 |
| 3 | 12 | 6 | 12 |
| 4 | 17 | 3 | 17 |
| 5 | 11 | NULL | NULL |
| 6 | 3 | 4 | 3 |
| 7 | 1 | 2 | 1 |
| 7 | 1 | 11 | 1 |
| 7 | 1 | 15 | 1 |
| 8 | 16 | 1 | 16 |
| 8 | 16 | 18 | 16 |
| 9 | 17 | 3 | 17 |
| 10 | 16 | 1 | 16 |
| 10 | 16 | 18 | 16 |
| 11 | 9 | 17 | 9 |
| 12 | 17 | 3 | 17 |
| 13 | 18 | 7 | 18 |
| 14 | 16 | 1 | 16 |
| 14 | 16 | 18 | 16 |
| 15 | 7 | NULL | NULL |
| 16 | 8 | NULL | NULL |
| 17 | 19 | NULL | NULL |
| 18 | 9 | 17 | 9 |
| 19 | 6 | NULL | NULL |
| 20 | 5 | NULL | NULL |
| 21 | 6 | NULL | NULL |
+----+------+--------+------+
28 rows in set (0.00 sec)- 使用 explain 分析 SQL 语句的性能，可以看到：驱动表是左表 class 表
mysql> EXPLAIN SELECT * FROM class LEFT JOIN book ON class.card = book.card;
+----+-------------+-------+------+---------------+------+---------+------+------+----------------------------------------------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+-------+------+---------------+------+---------+------+------+----------------------------------------------------+
| 1 | SIMPLE | class | ALL | NULL | NULL | NULL | NULL | 21 | NULL |
| 1 | SIMPLE | book | ALL | NULL | NULL | NULL | NULL | 20 | Using where; Using join buffer (Block Nested Loop) |
+----+-------------+-------+------+---------------+------+---------+------+------+----------------------------------------------------+
2 rows in set (0.00 sec)- 结论：
- type 有 All ，rows 为表中数据总行数，说明 class 和 book 进行了全表检索
- 即每次 class 表对 book 表进行左外连接时，都需要在 book 表中进行一次全表检索
添加索引：在右表添加索引
- 添加索引的 SQL 指令
ALTER TABLE 'book' ADD INDEX Y ('card');- 在 book 的 card 字段上添加索引
mysql> ALTER TABLE book ADD INDEX Y (card);
Query OK, 0 rows affected (0.30 sec)
Records: 0 Duplicates: 0 Warnings: 0
mysql> SHOW INDEX FROM book;
+-------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| Table | Non_unique | Key_name | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | Index_comment |
+-------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| book | 0 | PRIMARY | 1 | bookid | A | 20 | NULL | NULL | | BTREE | | |
| book | 1 | Y | 1 | card | A | 20 | NULL | NULL | | BTREE | | |
+-------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
2 rows in set (0.00 sec)- 测试结果：可以看到第二行的type变为了ref，rows也变成了优化比较明显。
mysql> EXPLAIN SELECT * FROM class LEFT JOIN book ON class.card = book.card;
+----+-------------+-------+------+---------------+------+---------+-----------------+------+-------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+-------+------+---------------+------+---------+-----------------+------+-------------+
| 1 | SIMPLE | class | ALL | NULL | NULL | NULL | NULL | 21 | NULL |
| 1 | SIMPLE | book | ref | Y | Y | 4 | db01.class.card | 1 | Using index |
+----+-------------+-------+------+---------------+------+---------+-----------------+------+-------------+
2 rows in set (0.00 sec)- 分析：
- 这是由左连接特性决定的。LEFT JOIN条件用于确定如何从右表搜索行，左边一定都有，所以右边是我们的关键点，一定需要建立索引。
- 左表连接右表，则需要拿着左表的数据去右表里面查，索引需要在右表中建立索引
添加索引：在右表添加索引
- 删除之前 book 表中的索引
DROP INDEX Y ON book;- 在 class 表的 card 字段上建立索引
ALTER TABLE class ADD INDEX X(card);- 再次执行左连接，凉凉~~~
mysql> EXPLAIN SELECT * FROM class LEFT JOIN book ON class.card = book.card;
+----+-------------+-------+-------+---------------+------+---------+------+------+----------------------------------------------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+-------+-------+---------------+------+---------+------+------+----------------------------------------------------+
| 1 | SIMPLE | class | index | NULL | X | 4 | NULL | 21 | Using index |
| 1 | SIMPLE | book | ALL | NULL | NULL | NULL | NULL | 20 | Using where; Using join buffer (Block Nested Loop) |
+----+-------------+-------+-------+---------------+------+---------+------+------+----------------------------------------------------+
2 rows in set (0.00 sec)- 别怕，我们来执行右连接：可以看到第二行的type变为了ref，rows也变成了优化比较明显。
mysql> EXPLAIN SELECT * FROM class RIGHT JOIN book ON class.card = book.card;
+----+-------------+-------+------+---------------+------+---------+----------------+------+-------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+-------+------+---------------+------+---------+----------------+------+-------------+
| 1 | SIMPLE | book | ALL | NULL | NULL | NULL | NULL | 20 | NULL |
| 1 | SIMPLE | class | ref | X | X | 4 | db01.book.card | 1 | Using index |
+----+-------------+-------+------+---------------+------+---------+----------------+------+-------------+
2 rows in set (0.00 sec)- 分析：
- 这是因为RIGHT JOIN条件用于确定如何从左表搜索行，右边一定都有，所以左边是我们的关键点，一定需要建立索引。
- class RIGHT JOIN book ：book 里面的数据一定存在于结果集中，我们需要拿着 book 表中的数据，去 class 表中搜索，所以索引需要建立在 class 表中
- 为了不影响之后的测试，删除该表的 idx_article_ccv 索引
mysql> DROP INDEX X ON class;
Query OK, 0 rows affected (0.04 sec)
Records: 0 Duplicates: 0 Warnings: 0
mysql> SHOW INDEX FROM class;
+-------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| Table | Non_unique | Key_name | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | Index_comment |
+-------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| class | 0 | PRIMARY | 1 | id | A | 21 | NULL | NULL | | BTREE | | |
+-------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
1 row in set (0.00 sec)3、三表索引优化
三表索引优化分析
创建表
- 建表 SQL
CREATE TABLE IF NOT EXISTS phone(
phoneid INT(10) UNSIGNED NOT NULL AUTO_INCREMENT,
card INT(10) UNSIGNED NOT NULL,
PRIMARY KEY(phoneid)
)ENGINE=INNODB;
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));
INSERT INTO phone(card) VALUES(FLOOR(1+(RAND()*20)));- phone 表中的测试数据
mysql> select * from phone;
+---------+------+
| phoneid | card |
+---------+------+
| 1 | 7 |
| 2 | 7 |
| 3 | 13 |
| 4 | 6 |
| 5 | 8 |
| 6 | 4 |
| 7 | 16 |
| 8 | 4 |
| 9 | 15 |
| 10 | 1 |
| 11 | 20 |
| 12 | 18 |
| 13 | 9 |
| 14 | 9 |
| 15 | 20 |
| 16 | 11 |
| 17 | 15 |
| 18 | 3 |
| 19 | 8 |
| 20 | 10 |
+---------+------+
20 rows in set (0.00 sec)查询案例
- 实现三表的连接查询：
mysql> SELECT * FROM class LEFT JOIN book ON class.card = book.card LEFT JOIN phone ON book.card = phone.card;
+----+------+--------+------+---------+------+
| id | card | bookid | card | phoneid | card |
+----+------+--------+------+---------+------+
| 2 | 13 | 8 | 13 | 3 | 13 |
| 2 | 13 | 9 | 13 | 3 | 13 |
| 2 | 13 | 12 | 13 | 3 | 13 |
| 8 | 16 | 1 | 16 | 7 | 16 |
| 10 | 16 | 1 | 16 | 7 | 16 |
| 14 | 16 | 1 | 16 | 7 | 16 |
| 8 | 16 | 18 | 16 | 7 | 16 |
| 10 | 16 | 18 | 16 | 7 | 16 |
| 14 | 16 | 18 | 16 | 7 | 16 |
| 7 | 1 | 2 | 1 | 10 | 1 |
| 7 | 1 | 11 | 1 | 10 | 1 |
| 7 | 1 | 15 | 1 | 10 | 1 |
| 13 | 18 | 7 | 18 | 12 | 18 |
| 11 | 9 | 17 | 9 | 13 | 9 |
| 18 | 9 | 17 | 9 | 13 | 9 |
| 11 | 9 | 17 | 9 | 14 | 9 |
| 18 | 9 | 17 | 9 | 14 | 9 |
| 6 | 3 | 4 | 3 | 18 | 3 |
| 4 | 17 | 3 | 17 | NULL | NULL |
| 9 | 17 | 3 | 17 | NULL | NULL |
| 12 | 17 | 3 | 17 | NULL | NULL |
| 1 | 12 | 6 | 12 | NULL | NULL |
| 3 | 12 | 6 | 12 | NULL | NULL |
| 5 | 11 | NULL | NULL | NULL | NULL |
| 15 | 7 | NULL | NULL | NULL | NULL |
| 16 | 8 | NULL | NULL | NULL | NULL |
| 17 | 19 | NULL | NULL | NULL | NULL |
| 19 | 6 | NULL | NULL | NULL | NULL |
| 20 | 5 | NULL | NULL | NULL | NULL |
| 21 | 6 | NULL | NULL | NULL | NULL |
+----+------+--------+------+---------+------+
30 rows in set (0.00 sec)- 使用 explain 分析 SQL 指令：
mysql> EXPLAIN SELECT * FROM class LEFT JOIN book ON class.card = book.card LEFT JOIN phone ON book.card = phone.card;
+----+-------------+-------+------+---------------+------+---------+------+------+----------------------------------------------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+-------+------+---------------+------+---------+------+------+----------------------------------------------------+
| 1 | SIMPLE | class | ALL | NULL | NULL | NULL | NULL | 21 | NULL |
| 1 | SIMPLE | book | ALL | NULL | NULL | NULL | NULL | 20 | Using where; Using join buffer (Block Nested Loop) |
| 1 | SIMPLE | phone | ALL | NULL | NULL | NULL | NULL | 20 | Using where; Using join buffer (Block Nested Loop) |
+----+-------------+-------+------+---------------+------+---------+------+------+----------------------------------------------------+
3 rows in set (0.00 sec)- 结论：
- type 有All ，rows 为表数据总行数，说明 class、 book 和 phone 表都进行了全表检索
- Extra 中 Using join buffer ，表明连接过程中使用了 join 缓冲区
创建索引
- 创建索引的 SQL 语句
ALTER TABLE book ADD INDEX Y (card);
ALTER TABLE phone ADD INDEX Z (card);- 进行 LEFT JOIN ，永远都在右表的字段上建立索引
mysql> ALTER TABLE book ADD INDEX Y (card);
Query OK, 0 rows affected (0.06 sec)
Records: 0 Duplicates: 0 Warnings: 0
mysql> SHOW INDEX FROM book;
+-------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| Table | Non_unique | Key_name | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | Index_comment |
+-------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| book | 0 | PRIMARY | 1 | bookid | A | 20 | NULL | NULL | | BTREE | | |
| book | 1 | Y | 1 | card | A | 20 | NULL | NULL | | BTREE | | |
+-------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
2 rows in set (0.00 sec)
mysql> ALTER TABLE phone ADD INDEX Z (card);
Query OK, 0 rows affected (0.05 sec)
Records: 0 Duplicates: 0 Warnings: 0
mysql> SHOW INDEX FROM phone;
+-------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| Table | Non_unique | Key_name | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | Index_comment |
+-------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| phone | 0 | PRIMARY | 1 | phoneid | A | 20 | NULL | NULL | | BTREE | | |
| phone | 1 | Z | 1 | card | A | 20 | NULL | NULL | | BTREE | | |
+-------+------------+----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
2 rows in set (0.00 sec)- 执行查询：后2行的type都是ref，且总rows优化很好，效果不错。因此索引最好设置在需要经常查询的字段中。
mysql> EXPLAIN SELECT * FROM class LEFT JOIN book ON class.card=book.card LEFT JOIN phone ON book.card = phone.card;
+----+-------------+-------+------+---------------+------+---------+-----------------+------+-------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+-------+------+---------------+------+---------+-----------------+------+-------------+
| 1 | SIMPLE | class | ALL | NULL | NULL | NULL | NULL | 21 | NULL |
| 1 | SIMPLE | book | ref | Y | Y | 4 | db01.class.card | 1 | Using index |
| 1 | SIMPLE | phone | ref | Z | Z | 4 | db01.book.card | 1 | Using index |
+----+-------------+-------+------+---------------+------+---------+-----------------+------+-------------+
3 rows in set (0.00 sec)Join 语句优化的结论
将 left join 看作是两层嵌套 for 循环
- 尽可能减少Join语句中的NestedLoop的循环总次数；
- 永远用小结果集驱动大的结果集（在大结果集中建立索引，在小结果集中遍历全表）；
- 优先优化NestedLoop的内层循环；
- 保证Join语句中被驱动表上Join条件字段已经被索引；
- 当无法保证被驱动表的Join条件字段被索引且内存资源充足的前提下，不要太吝惜JoinBuffer的设置；
我的理解
- 使用小表驱动大表，这就相当于外层 for 循环的次数少，内层 for 循环的次数多。
- 然后我们在大表中建立了索引，这样内层 for 循环的效率明显提高
- 综上，使用小表驱动大表，在大表中建立了索引
小表的每条都得遍历，用小表的id在大表中用索引查，所以快。
