正 文:
1, limit语句的优化 。
常见的limit语句的形式为:LIMIT m,n;随之偏移量m的增大,limit语句的执行效率也跟着下降。所以,优化limit的原则是尽量不要使用偏移量m,将limit m,n转换为limit n的形式,万一非要使用偏移量m,也要m尽可能的小。
现在,从表items表中,找出10000之后的10条记录。一般的查找方法如下:
mysql> explain select ItemID,ItemName,Code from items limit 10000,10;
+----+-------------+-------+------+---------------+------+---------+------+-------+-------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------+
| 1 | SImpLE | items | ALL | NULL | NULL | NULL | NULL | 22116 | |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------+
1 row in set (0.00 sec)
索引类型为ALL,而且是全表扫描,影响的行数整张表的记录总数。可见是对全行数据进行偏移。
一个简单的技巧是对索引数据进行偏移,然后将索引数据与全行数据内联,得到所需的列。语句如下:
mysql> explain select ItemID from items INNER JOIN (select ItemID from items order by ItemID LIMIT 10000,10) AS items2 USING(ItemID);
+----+-------------+------------+--------+---------------+---------+---------+---------------+-------+-------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+------------+--------+---------------+---------+---------+---------------+-------+-------------+
| 1 | PRIMARY | | ALL | NULL | NULL | NULL | NULL | 10 | |
| 1 | PRIMARY | items | eq_ref | PRIMARY | PRIMARY | 4 | items2.ItemID | 1 | Using index |
| 2 | DERIVED | items | index | NULL | PRIMARY | 4 | NULL | 10010 | Using index |
+----+-------------+------------+--------+---------------+---------+---------+---------------+-------+-------------+
3 rows in set (0.00 sec)
上述查询影响的rows书只有10010,查询中对索引ItemID进行了排序。
还有一种查找方法,使用子查询,转换为limit n形式。如下
mysql> explain select ItemID,ItemName,Code from items where ItemID >= 10000 order by ItemID limit 10;
+----+-------------+-------+-------+---------------+---------+---------+------+-------+-------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+-------+-------+---------------+---------+---------+------+-------+-------------+
| 1 | SImpLE | items | range | PRIMARY | PRIMARY | 4 | NULL | 13563 | Using where |
+----+-------------+-------+-------+---------------+---------+---------+------+-------+-------------+
1 row in set (0.00 sec)
如果可以确定记录的具体位置,也可以使用between….and…来达到效果
mysql> explain select ItemID,ItemName,Code FROM items where ItemID between 10000 and 10010 order by ItemID;
+----+-------------+-------+-------+---------------+---------+---------+------+------+-------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+-------+-------+---------------+---------+---------+------+------+-------------+
| 1 | SImpLE | items | range | PRIMARY | PRIMARY | 4 | NULL | 8 | Using where |
+----+-------------+-------+-------+---------------+---------+---------+------+------+-------------+
2 count语句的优化
count语句不好做优化,只有写特例,在没有使用where语句的count查询,非常快。也就是select count(*) from items
现在,查询items表中,ItemID大于2的所有记录。sql语句如下:
+----+-------------+-------+-------+---------------+---------+---------+------+------+--------------------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+-------+-------+---------------+---------+---------+------+------+--------------------------+
| 1 | SImpLE | items | index | PRIMARY | PRIMARY | 4 | NULL | 22114 | Using where; Using index |
+----+-------------+-------+-------+---------------+---------+---------+------+------+--------------------------+
1 row in set (0.00 sec)
对比如下查询
mysql> explain select (select count(*) from items) - count(*) from items where ItemID <= 2;
+----+-------------+-------+-------+---------------+---------+---------+------+------+------------------------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+-------+-------+---------------+---------+---------+------+------+------------------------------+
| 1 | PRIMARY | items | range | PRIMARY | PRIMARY | 4 | NULL | 1 | Using where; Using index |
| 2 | SUBQUERY | NULL | NULL | NULL | NULL | NULL | NULL | NULL | Select tables optimized away |
+----+-------------+-------+-------+---------------+---------+---------+------+------+------------------------------+
2 rows in set (0.00 sec)
第一句查询影响了>2的所以行,第二句查询影响了<=2的所以记录。关键区别在(select count(*) from items) - count(*)来计算行数和where条件的不同。
还有一个有趣的地方,如果要统计同一列中不同值的记录数。如items表里产品上架(1)和下架(0)的不同数量。有两种查找方法,一种是使用count,一种是使用sum函数。
mysql> explain select count(isActive = '1' or null) as up,count(isActive = '0' or null) as down from items;
+----+-------------+-------+------+---------------+------+---------+------+-------+-------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------+
| 1 | SImpLE | items | ALL | NULL | NULL | NULL | NULL | 22116 | |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------+
1 row in set (0.00 sec)
mysql> explain select sum(if(isActive='1',1,0)) as up,sum(if(isActive='0',1,0)) as down from items;
+----+-------------+-------+------+---------------+------+---------+------+-------+-------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------+
| 1 | SImpLE | items | ALL | NULL | NULL | NULL | NULL | 22116 | |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------+
1 row in set (0.00 sec)