优化GROUP BY语句
默认情况下，MySQL对所有GROUP BY col1，col2...的字段进行排序。这与在查询中指定ORDER BY col1,col2...类似。因此，如果显式包括一个包含相同的列的ORDER BY子句，则对MySQL的实际执行性能没有什么影响。 如果查询包括GROUP BY 但用户想要避免排序结果的消耗，则可以指定ORDER By NULL禁止排序，例如：
Java代码 
- explain select id, sum(moneys) from sales2 group by id \G
- explain select id, sum(moneys) from sales2 group by id order by null \G
explain select id, sum(moneys) from sales2 group by id \G explain select id, sum(moneys) from sales2 group by id order by null \G你可以通过比较发现第一条语句会比第二句在Extra:里面多了Using filesort.而恰恰filesort是最耗时的。
优化ORDER BY语句
在某些情况中，MySQL可以使用一个索引来满足ORDER BY子句，而不需要额外的排序。WHERE 条件和 ORDER BY使用相同的索引，并且ORDER BY的顺序和索引顺序相同，并且ORDER BY的字段都是升序或者都是降序。
例如：
Java代码
- SELECT * FROM t1 ORDER BY key_part1,key_part2,....:
- SELECT * FROM t1 WHERE key_part1 = 1 ORDER BY key_part1 DESC,key_part2 DESC;
- SELECT * FROM t1 ORDER BY key_part1 DESC, key_part2 DESC;
SELECT * FROM t1 ORDER BY key_part1,key_part2,....: SELECT * FROM t1 WHERE key_part1 = 1 ORDER BY key_part1 DESC,key_part2 DESC; SELECT * FROM t1 ORDER BY key_part1 DESC, key_part2 DESC;但是以下的情况不使用索引：
Java代码
- SELECT * FROM t1 ORDER BY key_part1 DESC, key_part2 ASC;
- --ORDER by的字段混合ASC 和 DESC
- SELECT * FROM t1 WHERE key2=constant ORDER BY key1;
- ----用于查询行的关键字与ORDER BY 中所使用的不相同
- SELECT * FROM t1 ORDER BY key1, key2;
- ----对不同的关键字使用ORDER BY
SELECT * FROM t1 ORDER BY key_part1 DESC, key_part2 ASC; --ORDER by的字段混合ASC 和 DESC SELECT * FROM t1 WHERE key2=constant ORDER BY key1; ----用于查询行的关键字与ORDER BY 中所使用的不相同 SELECT * FROM t1 ORDER BY key1, key2; ----对不同的关键字使用ORDER BY
Java代码
- mysql > explain select A . id , A . title , B . title from jos_content A left join jos_categories B on A . catid = B . id left join jos_sections C on A . sectionid = C . id order by B . id ;
- +----+-------------+-------+--------+---------------+---------+---------+-------------------------+-------+---------------------------------+
- | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
- +----+-------------+-------+--------+---------------+---------+---------+-------------------------+-------+---------------------------------+
- | 1 | SIMPLE | A | ALL | NULL | NULL | NULL | NULL | 46585 | Using temporary ; Using filesort |
- | 1 | SIMPLE | B | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . catid | 1 | |
- | 1 | SIMPLE | C | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . sectionid | 1 | Using index |
- +----+-------------+-------+--------+---------------+---------+---------+-------------------------+-------+---------------------------------+
- 3 rows in set ( 0.00 sec )
- mysql > explain select A . id , A . title , B . title from jos_content A left join jos_categories B on A . catid = B . id left join jos_sections C on A . sectionid = C . id order by A . id ;
- +----+-------------+-------+--------+---------------+---------+---------+-------------------------+-------+----------------+
- | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
- +----+-------------+-------+--------+---------------+---------+---------+-------------------------+-------+----------------+
- | 1 | SIMPLE | A | ALL | NULL | NULL | NULL | NULL | 46585 | Using filesort |
- | 1 | SIMPLE | B | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . catid | 1 | |
- | 1 | SIMPLE | C | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . sectionid | 1 | Using index |
- +----+-------------+-------+--------+---------------+---------+---------+-------------------------+-------+----------------+
mysql > explain select A . id , A . title , B . title from jos_content A left join jos_categories B on A . catid = B . id left join jos_sections C on A . sectionid = C . id order by B . id ; +----+-------------+-------+--------+---------------+---------+---------+-------------------------+-------+---------------------------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +----+-------------+-------+--------+---------------+---------+---------+-------------------------+-------+---------------------------------+ | 1 | SIMPLE | A | ALL | NULL | NULL | NULL | NULL | 46585 | Using temporary ; Using filesort | | 1 | SIMPLE | B | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . catid | 1 | | | 1 | SIMPLE | C | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . sectionid | 1 | Using index | +----+-------------+-------+--------+---------------+---------+---------+-------------------------+-------+---------------------------------+ 3 rows in set ( 0.00 sec )mysql > explain select A . id , A . title , B . title from jos_content A left join jos_categories B on A . catid = B . id left join jos_sections C on A . sectionid = C . id order by A . id ;
+----+-------------+-------+--------+---------------+---------+---------+-------------------------+-------+----------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+-------+--------+---------------+---------+---------+-------------------------+-------+----------------+
| 1 | SIMPLE | A | ALL | NULL | NULL | NULL | NULL | 46585 | Using filesort |
| 1 | SIMPLE | B | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . catid | 1 | |
| 1 | SIMPLE | C | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . sectionid | 1 | Using index |
+----+-------------+-------+--------+---------------+---------+---------+-------------------------+-------+----------------+
对于上面两条语句，只是修改了一下排序字段，而第一个使用了Using temporary，而第二个却没有。在日常的网站维护中，如果有Using temporary出现，说明需要做一些优化措施了。
而为什么第一个用了临时表，而第二个没有用呢？
因为如果有ORDER BY子句和一个不同的GROUP BY子句，或者如果ORDER BY或GROUP BY中的字段都来自其他的表而非连接顺序中的第一个表的话，就会创建一个临时表了。
那么，对于上面例子中的第一条语句，我们需要对jos_categories的id进行排序，可以将SQL做如下改动：
Java代码
- mysql > explain select B . id , B . title , A . title from jos_categories A left join jos_content B on A . id = B . catid left join jos_sections C on B . sectionid = C . id order by A . id ;
- +----+-------------+-------+--------+---------------+-----------+---------+-------------------------+------+----------------+
- | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
- +----+-------------+-------+--------+---------------+-----------+---------+-------------------------+------+----------------+
- | 1 | SIMPLE | A | ALL | NULL | NULL | NULL | NULL | 18 | Using filesort |
- | 1 | SIMPLE | B | ref | idx_catid | idx_catid | 4 | joomla_test . A . id | 3328 | |
- | 1 | SIMPLE | C | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . B . sectionid | 1 | Using index |
- +----+-------------+-------+--------+---------------+-----------+---------+-------------------------+------+----------------+
- 3 rows in set ( 0.00 sec )
mysql > explain select B . id , B . title , A . title from jos_categories A left join jos_content B on A . id = B . catid left join jos_sections C on B . sectionid = C . id order by A . id ; +----+-------------+-------+--------+---------------+-----------+---------+-------------------------+------+----------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +----+-------------+-------+--------+---------------+-----------+---------+-------------------------+------+----------------+ | 1 | SIMPLE | A | ALL | NULL | NULL | NULL | NULL | 18 | Using filesort | | 1 | SIMPLE | B | ref | idx_catid | idx_catid | 4 | joomla_test . A . id | 3328 | | | 1 | SIMPLE | C | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . B . sectionid | 1 | Using index | +----+-------------+-------+--------+---------------+-----------+---------+-------------------------+------+----------------+ 3 rows in set ( 0.00 sec )
这样我们发现，不会再有Using temporary了，而且在查询jos_content时，查询的记录明显有了数量级的降低，这是因为jos_content的idx_catid起了作用。
所以结论是：
尽量对第一个表的索引键进行排序，这样效率是高的。
我们还会发现，在排序的语句中都出现了Using filesort，字面意思可能会被理解为：使用文件进行排序或中文件中进行排序。实际上这是不正确的，这是一个让人产生误解的词语。
当我们试图对一个没有索引的字段进行排序时，就是filesoft。它跟文件没有任何关系，实际上是内部的一个快速排序。
然而，当我们回过头来再看上面运行过的一个SQL的时候会有以下发现：
Java代码
- mysql > explain select A . id , A . title , B . title from jos_content A , jos_categories B , jos_sections C where A . catid = B . id and A . sectionid = C . id order by C . id ;
- +----+-------------+-------+--------+-----------------------+-------------+---------+---------------------+-------+-------------+
- | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
- +----+-------------+-------+--------+-----------------------+-------------+---------+---------------------+-------+-------------+
- | 1 | SIMPLE | C | index | PRIMARY | PRIMARY | 4 | NULL | 1 | Using index |
- | 1 | SIMPLE | A | ref | idx_catid , idx_section | idx_section | 4 | joomla_test . C . id | 23293 | Using where |
- | 1 | SIMPLE | B | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . catid | 1 | Using where |
- +----+-------------+-------+--------+-----------------------+-------------+---------+---------------------+-------+-------------+
- 3 rows in set ( 0.00 sec )
mysql > explain select A . id , A . title , B . title from jos_content A , jos_categories B , jos_sections C where A . catid = B . id and A . sectionid = C . id order by C . id ; +----+-------------+-------+--------+-----------------------+-------------+---------+---------------------+-------+-------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +----+-------------+-------+--------+-----------------------+-------------+---------+---------------------+-------+-------------+ | 1 | SIMPLE | C | index | PRIMARY | PRIMARY | 4 | NULL | 1 | Using index | | 1 | SIMPLE | A | ref | idx_catid , idx_section | idx_section | 4 | joomla_test . C . id | 23293 | Using where | | 1 | SIMPLE | B | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . catid | 1 | Using where | +----+-------------+-------+--------+-----------------------+-------------+---------+---------------------+-------+-------------+ 3 rows in set ( 0.00 sec )
这是我们刚才运行过的一条语句，只是加了一个排序，而这条语句中C表的主键对排序起了作用，我们会发现Using filesort没有了。
而尽管在上面的语句中也是对第一个表的主键进行排序，却没有得到想要的效果（第一个表的主键没有用到），这是为什么呢？实际上以上运行过的所有left join的语句中，第一个表的索引都没有用到，尽管对第一个表的主键进行了排序也无济于事。不免有些奇怪！
于是我们继续测试了下一条SQL：
Java代码
- mysql > explain select A . id , A . title , B . title from jos_content A left join jos_categories B on A . catid = B . id left join jos_sections C on A . sectionid = C . id where A . id < 100 ;
- +----+-------------+-------+--------+----------------+---------+---------+-------------------------+------+-------------+
- | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
- +----+-------------+-------+--------+----------------+---------+---------+-------------------------+------+-------------+
- | 1 | SIMPLE | A | range | PRIMARY | PRIMARY | 4 | NULL | 90 | Using where |
- | 1 | SIMPLE | B | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . catid | 1 | |
- | 1 | SIMPLE | C | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . sectionid | 1 | Using index |
- +----+-------------+-------+--------+----------------+---------+---------+-------------------------+------+-------------+
- 3 rows in set ( 0.05 sec )
mysql > explain select A . id , A . title , B . title from jos_content A left join jos_categories B on A . catid = B . id left join jos_sections C on A . sectionid = C . id where A . id < 100 ; +----+-------------+-------+--------+----------------+---------+---------+-------------------------+------+-------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +----+-------------+-------+--------+----------------+---------+---------+-------------------------+------+-------------+ | 1 | SIMPLE | A | range | PRIMARY | PRIMARY | 4 | NULL | 90 | Using where | | 1 | SIMPLE | B | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . catid | 1 | | | 1 | SIMPLE | C | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . sectionid | 1 | Using index | +----+-------------+-------+--------+----------------+---------+---------+-------------------------+------+-------------+ 3 rows in set ( 0.05 sec )
然后，当再次进行排序操作的时候，Using filesoft也没有再出现
Java代码
- mysql > explain select A . id , A . title , B . title from jos_content A left join jos_categories B on A . catid = B . id left join jos_sections C on A . sectionid = C . id where A . id < 100 order by A . id ;
- +----+-------------+-------+--------+---------------+---------+---------+-------------------------+------+-------------+
- | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
- +----+-------------+-------+--------+---------------+---------+---------+-------------------------+------+-------------+
- | 1 | SIMPLE | A | range | PRIMARY | PRIMARY | 4 | NULL | 105 | Using where |
- | 1 | SIMPLE | B | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . catid | 1 | |
- | 1 | SIMPLE | C | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . sectionid | 1 | Using index |
- +----+-------------+-------+--------+---------------+---------+---------+-------------------------+------+-------------+
- 3 rows in set ( 0.00 sec )
mysql > explain select A . id , A . title , B . title from jos_content A left join jos_categories B on A . catid = B . id left join jos_sections C on A . sectionid = C . id where A . id < 100 order by A . id ; +----+-------------+-------+--------+---------------+---------+---------+-------------------------+------+-------------+ | id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra | +----+-------------+-------+--------+---------------+---------+---------+-------------------------+------+-------------+ | 1 | SIMPLE | A | range | PRIMARY | PRIMARY | 4 | NULL | 105 | Using where | | 1 | SIMPLE | B | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . catid | 1 | | | 1 | SIMPLE | C | eq_ref | PRIMARY | PRIMARY | 4 | joomla_test . A . sectionid | 1 | Using index | +----+-------------+-------+--------+---------------+---------+---------+-------------------------+------+-------------+ 3 rows in set ( 0.00 sec )
这个结果表明：对where条件里涉及到的字段，Mysql会使用索引进行搜索，而这个索引的使用也对排序的效率有很好的提升。
写了段程序测试了一下，分别让以下两个SQL语句执行200次：
Java代码
- select A . id , A . title , B . title from jos_content A left join jos_categories B on A . catid = B . id left join jos_sections C on A . sectionid = C . id
- select A . id , A . title , B . title from jos_content A , jos_categories B , jos_sections C where A . catid = B . id and A . sectionid = C . id
- select A . id , A . title , B . title from jos_content A left join jos_categories B on A . catid = B . id left join jos_sections C on A . sectionid = C . id order by rand () limit 10
- select A . id from jos_content A left join jos_categories B on B . id = A . catid left join jos_sections C on A . sectionid = C . id order by A . id
select A . id , A . title , B . title from jos_content A left join jos_categories B on A . catid = B . id left join jos_sections C on A . sectionid = C . id select A . id , A . title , B . title from jos_content A , jos_categories B , jos_sections C where A . catid = B . id and A . sectionid = C . id select A . id , A . title , B . title from jos_content A left join jos_categories B on A . catid = B . id left join jos_sections C on A . sectionid = C . id order by rand () limit 10 select A . id from jos_content A left join jos_categories B on B . id = A . catid left join jos_sections C on A . sectionid = C . id order by A . id
结果是第(1)条平均用时20s ，第(2)条平均用时44s ，第(3)条平均用时70s ，第(4)条平均用时2s 。而且假如我们用explain观察第(3)条语句的执行情况，会发现它创建了temporary表来进行排序。
综上所述，可以得出如下结论：
1. 对需要查询和排序的字段要加索引。
2. 在一定环境下，left join还是比普通连接查询效率要高，但是要尽量少地连接表，并且在做连接查询时注意观察索引是否起了作用。
3. 排序尽量对第一个表的索引字段进行，可以避免mysql创建临时表，这是非常耗资源的。
4. 对where条件里涉及到的字段，应适当地添加索引，这样会对排序操作有优化的作用。
5. 在做随机抽取数据的需求时，避免使用order by rand()，从上面的例子可以看出，这种是很浪费数据库资源的，在执行过程中用show processlist查看，会发现第(3)条有Copying to tmp table on disk。而对(3)和(4)的对比得知，如果要实现这个功能，最好另辟奚径，来减轻Mysql的压力。
6. 从第4点可以看出，如果说在分页时我们能先得到主键，再根据主键查询相关内容，也能得到查询的优化效果。通过国外《High Performance MySQL》专家组的测试可以看出，根据主键进行查询的类似“SELECT ... FROM... WHERE id = ...”的SQL语句（其中id为PRIMARYKEY），每秒钟能够处理10000次 以上的查询，而普通的SELECT查询每秒只能处理几十次到几百次 。涉及到分页的查询效率问题，网上的可用资源越来越多，查询功能也体现出了它的重要性。也便是sphinx、lucene这些第三方搜索引擎的用武之地了。
7. 在平时的作业中，可以打开Mysql的Slow queries功能，经常检查一下是哪些语句降低的Mysql的执行效率，并进行定期优化。
2018-06-11更新
准确的说“order by”后面要跟驱动表的字段，如果不清楚哪张表是驱动表，可以用“explain”执行SQL之后结果中第一条是哪张表驱动表就是哪张表。