通过多个内部联接加速SQL查询

Question

我有以下查询：

SELECT COUNT(DISTINCT(`person_id`)) as `count`, `mc_office`.`name` as `office_name`
FROM `aft_people` 
                      INNER JOIN `aft_offices` 
                      ON `aft_people`.`lc`=`aft_offices`.`id`
                      INNER JOIN `aft_offices` as `mc_office`
                      ON `aft_offices`.`parent_id` = `mc_office`.`id`
                      INNER JOIN `aft_constant_maps` as `constant_maps`
                      ON `aft_people`.`person_id` = `constant_maps`.`representable_id`
                      WHERE `constant_maps`.`constant_id` IN (741)
                      GROUP BY `mc_office`.`name`

表aft_people有1M条记录，而aft_constant_maps有大约500万条记录。 字段上有索引

• aft_people.person_id
• aft_constant_maps.representable_id
• aft_constant_maps.constant_id

该查询确实非常缓慢，有时甚至根本无法加载。 我需要此查询在不到10秒的时间内执行。

如果您需要更多信息来帮助我，请告诉我。

Answer 1

对于此查询（我刚刚重新排列了联接的位置，以便可以更好地关注它们）：

SELECT COUNT(DISTINCT(`person_id`)) as `count`,
       `mc_office`.`name` as `office_name`
from `aft_people` INNER JOIN 
     `aft_constant_maps` as `constant_maps`
     ON `aft_people`.`person_id` = `constant_maps`.`representable_id` INNER JOIN
     `aft_offices` 
     ON `aft_people`.`lc` = `aft_offices`.`id` INNER JOIN
     `aft_offices` as `mc_office`
     ON `aft_offices`.`parent_id` = `mc_office`.`id` 
WHERE `constant_maps`.`constant_id` IN (741)
GROUP BY `mc_office`.`name`

最好的索引是： constant_maps(constant_id, representable_id) aft_people(person_id, lc) constant_maps(constant_id, representable_id) ， aft_people(person_id, lc)和aft_offices(id, parent_id) 。 这些可能有助于加快查询速度。