结合业务特性巧用临时表

拿到某开发sql如下

SELECT  p.products_id FROM products AS p
JOIN products_to_categories AS pc USING(products_id)
JOIN categories AS c USING(categories_id)
JOIN products_realtime_quantity AS prq ON prq.sku_or_poa = p.products_model
WHERE products_status =1 AND categories_status =1   AND prq.msg != 'Temporary out stock.'
ORDER BY p.products_date_added DESC LIMIT 10

一般看到这种sql,在where中只有status类似的字段(可选择性非常低,数据两极分化非常明显)而且需要order by的语句,我们就应该想到使用force index(order_by_column)来进行优化.

explian

+----+-------------+-------+--------+------------------------------+-------------------+---------+--------------------------------+--------+----------------------------------------------+
| id | select_type | table | type   | possible_keys                | key               | key_len | ref                            | rows   | Extra                                        |
+----+-------------+-------+--------+------------------------------+-------------------+---------+--------------------------------+--------+----------------------------------------------+
|  1 | SIMPLE      | pc    | index  | PRIMARY,categories_id        | PRIMARY           | 8       | NULL                           | 1009510 | Using index; Using temporary; Using filesort |
|  1 | SIMPLE      | p     | eq_ref | PRIMARY,products_model       | PRIMARY           | 4       | banggood_work.pc.products_id   |      1 | Using where                                  |
|  1 | SIMPLE      | c     | eq_ref | PRIMARY                      | PRIMARY           | 4       | banggood_work.pc.categories_id |      1 | Using where                                  |
|  1 | SIMPLE      | prq   | ref    | ix_prg_sku_or_poa,ix_prq_msg | ix_prg_sku_or_poa | 152     | banggood_work.p.products_model |      1 | Using where                                  |
+----+-------------+-------+--------+------------------------------+-------------------+---------+--------------------------------+--------+----------------------------------------------+

发现mysql优化器选择了pc表的主键,虽然使用了索引,但是进行了全索引扫描，效果还是不理想!

强制使用force index后,explain

EXPLAIN
    -> SELECT  p.products_id FROM products AS p   FORCE INDEX(products_date_added)
    -> JOIN products_to_categories AS pc USING(products_id)
    -> JOIN categories AS c USING(categories_id)
    -> JOIN products_realtime_quantity AS prq ON prq.sku_or_poa = p.products_model
    -> WHERE products_status =1 AND categories_status =1   AND prq.msg != 'Temporary out stock.'
    -> ORDER BY p.products_date_added DESC LIMIT 10
    -> ;
+----+-------------+-------+--------+------------------------------+---------------------+---------+--------------------------------+------+-------------+
| id | select_type | table | type   | possible_keys                | key                 | key_len | ref                            | rows | Extra       |
+----+-------------+-------+--------+------------------------------+---------------------+---------+--------------------------------+------+-------------+
|  1 | SIMPLE      | p     | index  | NULL                         | products_date_added | 8       | NULL                           |    1 | Using where |
|  1 | SIMPLE      | prq   | ref    | ix_prg_sku_or_poa,ix_prq_msg | ix_prg_sku_or_poa   | 152     | banggood_work.p.products_model |    1 | Using where |
|  1 | SIMPLE      | pc    | ref    | PRIMARY,categories_id        | PRIMARY             | 4       | banggood_work.p.products_id    | 1009 | Using index |
|  1 | SIMPLE      | c     | eq_ref | PRIMARY                      | PRIMARY             | 4       | banggood_work.pc.categories_id |    1 | Using where |
+----+-------------+-------+--------+------------------------------+---------------------+---------+--------------------------------+------+-------------+

发现索引已经变成productsw_date_added,而执行时间，前者是2s,后者是0.003s.

这是我们一贯的优化方法,但是我们可以根据sql语句的特性和业务特性,结合临时表进行一些淫邪的优化，虽然并不通用，但是可以开阔sql优化者的思维。

我们可以看到这条语句是需要根据产品添加时间拿取符合(products_status =1 AND categories_status =1   AND prq.msg != 'Temporary out stock.')条件的10个最新上架产品.而我们知道，最新上架的产品一般状态都是不可能马上下架,而且对应的类别id也是可用,而且库存也是充足的(要不然何必上架),这个特性站到了99.9%以上.所以，我们利用这个特性，先从产品表中找出不带任何条件的200个产品,放到临时表,然后再用临时表结果集,和拿取条件进行匹配,取出最新的10条.

(200条是一个参考值，根据各自的逻辑特性来取)

sql如下

SELECT DISTINCT p.products_id FROM
(SELECT products_id,products_model,products_status,products_date_added
    FROM products
    ORDER BY products_date_added DESC LIMIT 200
) AS p
JOIN products_to_categories AS pc USING(products_id)
JOIN categories AS c USING(categories_id)
JOIN products_realtime_quantity AS prq ON prq.sku_or_poa = p.products_model
WHERE products_status =1 AND categories_status =1   AND prq.msg != 'Temporary out stock.'
ORDER BY products_date_added DESC LIMIT 10;

explain后发现

|  1 | PRIMARY     |  | ALL    | NULL                         | NULL                | NULL    | NULL                           |  200 | Using where; Using temporary; Using filesort |
|  1 | PRIMARY     | prq        | ref    | ix_prg_sku_or_poa,ix_prq_msg | ix_prg_sku_or_poa   | 152     | p.products_model               |    1 | Using where; Distinct                        |
|  1 | PRIMARY     | pc         | ref    | PRIMARY,categories_id        | PRIMARY             | 4       | p.products_id                  | 1009 | Using index; Distinct                        |
|  1 | PRIMARY     | c          | eq_ref | PRIMARY                      | PRIMARY             | 4       | banggood_work.pc.categories_id |    1 | Using where; Distinct                        |
|  2 | DERIVED     | products   | index  | NULL                         | products_date_added | 8       | NULL                           |  200 |                                              |

我们可以看到，已经利用products表中的products_date_added字段排序取出200条,整个sql语句变成一个只有200行的临时表驱动查询了，性能相对于原来的语句,提高上百倍！

执行时间大约是0.02s(可能比force index略慢)。