MySQL如何优化子查询性能

时间:2018-01-12 03:40:46

标签: mysql performance indexing

我刚创建了一个表格:

CREATE TABLE `table_test` (
`time` date NOT NULL,
`line_id` char(36) NOT NULL,
`location_id` char(36) NOT NULL,
`placement_id` char(36) NOT NULL,
`flight_id` char(36) NOT NULL,
`impressions` int(11) DEFAULT `0`,
PRIMARY KEY (`time`,`line_id`,`location_id`,`placement_id`,`flight_id`),
KEY `table_test_IDX` (`time`,`placement_id`,`line_id`,`impressions`) USING 
BTREE
) ENGINE=InnoDB DEFAULT CHARSET=utf8

然后当我尝试执行查询时:

SELECT 
  time,
  placement_id,
  line_id,
  SUM(impressions) AS totalImpress
FROM
  table_test
WHERE
  time BETWEEN '2017-11-01' AND '2017-11-30'
GROUP BY time , placement_id , line_id;

只花了大约0.28秒的执行时间。

但是当我试图执行时:

SELECT
  daily.`time`,
  daily.placement_id,
  daily.line_id,
  daily.totalImpress
FROM
(SELECT 
  time,
  placement_id,
  line_id,
  SUM(impressions) AS totalImpress
 FROM
  table_test
 WHERE
  time BETWEEN '2017-11-01' AND '2017-11-30'
 GROUP BY time , placement_id , line_id)
 daily

执行时耗时超过7秒。

我在这里做错了什么?

非常感谢。

1 个答案:

答案 0 :(得分:0)

执行时间并不总是效率的最佳指标。 (很奇怪,我知道。)MySQL允许查询缓存。这基本上意味着如果您重复运行相同的查询,它在后续运行中会明显加快。如果您想优化查询,我建议使用" EXPLAIN"在查询开头的命令。