使我的SQL查询更有效

时间:2013-01-30 08:48:44

标签: mysql sql optimization

我的数据库中有数百万个条目。

因为我正在使用3 !!内连接并放置大范围的日期,这个查询需要很长时间,比如几分钟。

有没有办法改进此查询并仍能获得相同的数据?

这是我的疑问:

SELECT 
    subscriptions.service_id, 
    service_send_type.usage, 
    service_send_type.price, SUM(IF(msg_t.status LIKE 'Success%', 1, 0)) AS s, 
    COUNT(1) AS t 
FROM (`subscriptions`) 
    INNER JOIN 
        `msg_t` ON `subscriptions`.`phone` = `msg_t`.`phone` AND subscriptions.id = msg_t.sub_id 
    INNER JOIN
        `msg` ON `msg`.`id` = `msg_t`.`msg_id` 
    INNER JOIN `service_send_type` ON `msg`.`service_id` = `service_send_type`.`service_id` 
        AND msg.sushi_service_id = service_send_type.sushi_service_id 
        AND msg.send_type = service_send_type.name 
        AND msg.service_id = subscriptions.service_id 

    WHERE 
        `subscriptions`.`service_id` IN (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16) 
        AND 
        `subscriptions`.`added` >= '2013-01-28 00:00:00' 
        AND 
        `subscriptions`.`added` <= '2013-01-28 23:59:59' 
        AND 
        `msg_t`.`send_time` >= '2013-01-28 00:00:00' 
        AND 
        `msg_t`.`send_time` <= '2013-01-28 23:59:59' 

    GROUP BY 
        `subscriptions`.`service_id`, 
        `service_send_type`.`usage`, 
        `service_send_type`.`price`

2 个答案:

答案 0 :(得分:2)

您可以尝试optimize your query by analyzing the Explain plan,包括index optimizing,如果您的桌子上仍然没有idnexes,这可能会非常有用。关于语句优化,请查看here

希望这有助于你进一步

答案 1 :(得分:2)

您是否始终需要IN查询?在您的查询中,IN子句可以替换为subscriptions.service_id < 17,这将产生巨大的差异。当您有离散值列表时,IN查询是合适的。

另请查看表上的索引。您应该在where条件和group by子句中的列上设置索引。