mysql单表查询语句优化
浏览数：26 /
时间：2015年06月12日
Mysql语句优化
范例1：优化语句SELECT * FROM `tbl_order_buy_eta` WHERE `id_order`=1843
#通过explain分析语句结果如下 mysql> explain SELECT * FROM `tbl_order_buy_eta` WHERE `id_order`=1843\G *************************** 1. row *************************** id: 1 select_type: SIMPLE table: tbl_order_buy_eta type: ALL possible_keys: NULL key: NULL key_len: NULL ref: NULL rows: 1592 Extra: Using where 1 row in set (0.00 sec) #从上面我们能看出该语句没有使用任何索引，查询到结果扫描了1592行。 #查看表索引 mysql> show index from tbl_order_buy_eta\G *************************** 1. row *************************** Table: tbl_order_buy_eta Non_unique: 0 Key_name: PRIMARY Seq_in_index: 1 Column_name: id Collation: A Cardinality: 1592 Sub_part: NULL Packed: NULL Null: Index_type: BTREE Comment: Index_comment: 1 row in set (0.00 sec) #结果显示该表没有任何索引的存在 #我们在id_order列上创建索引 mysql> create index index_id_order on tbl_order_buy_eta(id_order); Query OK, 0 rows affected (0.29 sec) Records: 0 Duplicates: 0 Warnings: 0 mysql> show index from tbl_order_buy_eta\G *************************** 1. row *************************** Table: tbl_order_buy_eta Non_unique: 0 Key_name: PRIMARY Seq_in_index: 1 Column_name: id Collation: A Cardinality: 1592 Sub_part: NULL Packed: NULL Null: Index_type: BTREE Comment: Index_comment: *************************** 2. row *************************** Table: tbl_order_buy_eta Non_unique: 1 Key_name: index_id_order Seq_in_index: 1 Column_name: id_order Collation: A Cardinality: 1592 Sub_part: NULL Packed: NULL Null: Index_type: BTREE Comment: Index_comment: 2 rows in set (0.00 sec) #这一行是我们刚创建的索引 #再重新执行该查询语句，看看查询结果 mysql> explain SELECT * FROM `tbl_order_buy_eta` WHERE `id_order`=1843\G *************************** 1. row *************************** id: 1 select_type: SIMPLE table: tbl_order_buy_eta type: ref possible_keys: index_id_order key: index_id_order key_len: 4 ref: const rows: 1 Extra: NULL 1 row in set (0.03 sec) #添加索引后查询语句走的索引，扫描了1行就得到结果了
范例2：优化语句SELECT * FROM `tbl_order_buy` WHERE (`id_order`=1989) AND (`pay_status`=0) AND (`finish_status`=0);
#通过explain分析语句 mysql> explain SELECT * FROM `tbl_order_buy` WHERE (`id_order`=1989) AND (`pay_status`=0) AND (`finish_status`=0); +----+-------------+---------------+------+---------------+------+---------+------+------+-------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +----+-------------+---------------+------+---------------+------+---------+------+------+-------------+ | 1 | SIMPLE | tbl_order_buy | ALL | NULL | NULL | NULL | NULL | 1592 | Using where | +----+-------------+---------------+------+---------------+------+---------+------+------+-------------+ 1 row in set (0.00 sec) #该表中没有任何索引，查询语句走的是全表扫，一共扫描1592行 #创建索引 mysql> create index tbl_id_pay_finish on tbl_order_buy(id_order,pay_status,finish_status); Query OK, 0 rows affected (0.59 sec) Records: 0 Duplicates: 0 Warnings: 0 mysql> show index from tbl_order_buy; +---------------+------------+-------------------+--------------+---------------+-----------+-------------+----------+--------+------+------------+---------+---------------+ | Table | Non_unique | Key_name | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | Index_comment | +---------------+------------+-------------------+--------------+---------------+-----------+-------------+----------+--------+------+------------+---------+---------------+ | tbl_order_buy | 0 | PRIMARY | 1 | id | A | 1592 | NULL | NULL | | BTREE | | | | tbl_order_buy | 1 | tbl_id_pay_finish | 1 | id_order | A | 1592 | NULL | NULL | | BTREE | | | | tbl_order_buy | 1 | tbl_id_pay_finish | 2 | pay_status | A | 1592 | NULL | NULL | YES | BTREE | | | | tbl_order_buy | 1 | tbl_id_pay_finish | 3 | finish_status | A | 1592 | NULL | NULL | YES | BTREE | | | +---------------+------------+-------------------+--------------+---------------+-----------+-------------+----------+--------+------+------------+---------+---------------+ 4 rows in set (0.00 sec) #再次分析sql语句 mysql> explain SELECT * FROM `tbl_order_buy` WHERE (`id_order`=1989) AND (`pay_status`=0) AND (`finish_status`=0); +----+-------------+---------------+------+-------------------+-------------------+---------+-------------------+------+-------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +----+-------------+---------------+------+-------------------+-------------------+---------+-------------------+------+-------+ | 1 | SIMPLE | tbl_order_buy | ref | tbl_id_pay_finish | tbl_id_pay_finish | 14 | const,const,const | 1 | NULL | +----+-------------+---------------+------+-------------------+-------------------+---------+-------------------+------+-------+ 1 row in set (0.06 sec) #可以看到只扫描了1行就得到结果了
范例3：优化语句SELECT * FROM `tbl_order_vendor_item_variation` WHERE `id_order`=1989;
#使用explain分析语句 mysql> explain SELECT * FROM `tbl_order_vendor_item_variation` WHERE `id_order`=1989; +----+-------------+---------------------------------+------+---------------+------+---------+------+------+-------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +----+-------------+---------------------------------+------+---------------+------+---------+------+------+-------------+ | 1 | SIMPLE | tbl_order_vendor_item_variation | ALL | NULL | NULL | NULL | NULL | 2581 | Using where | +----+-------------+---------------------------------+------+---------------+------+---------+------+------+-------------+ 1 row in set (0.00 sec) #结果显示没有索引，走的是全表扫，一共扫描2581行 #创建索引 mysql> create index tbl_order_vendor_item_variation_id_order on tbl_order_vendor_item_variation(id_order); Query OK, 0 rows affected (0.36 sec) Records: 0 Duplicates: 0 Warnings: 0 #重新分析sql语句 mysql> explain SELECT * FROM `tbl_order_vendor_item_variation` WHERE `id_order`=1989\G *************************** 1. row *************************** id: 1 select_type: SIMPLE table: tbl_order_vendor_item_variation type: ref possible_keys: tbl_order_vendor_item_variation_id_order key: tbl_order_vendor_item_variation_id_order key_len: 5 ref: const rows: 1 Extra: NULL 1 row in set (0.00 sec) #sql语句走的是刚创建的索引，共扫描1行
本文出自 “ly36843运维” 博客，请务必保留此出处http://ly36843.blog.51cto.com/3120113/1640906
郑重声明：本站内容如果来自互联网及其他传播媒体，其版权均属原媒体及文章作者所有。转载目的在于传递更多信息及用于网络分享，并不代表本站赞同其观点和对其真实性负责，也不构成任何其他建议。
