优化MySQL查询（慢）

Question

Someone can help me to optimize the following MySQL query?

SELECT SUM(ROUND(rp.distance,2)), t.name, MONTH(tr.assigned)
FROM tab1 tr
JOIN tab2 rp ON rp.robot=tr.robot 
JOIN tab3 t ON t.idTask=tr.idTask
GROUP BY t.name, MONTH(tr.assigned)

The query takes 20 seconds to execute. Without the GROUP BY down to 10 seconds. I tested with DISTINCT but I failed to improve the time. Any ideas?

Answer 1

try this query

SELECT SUM(distance), t.name, MONTH(tr.assigned)
FROM tab1 tr
JOIN tab2 rp ON rp.robot=tr.robot 
JOIN (select ROUND(rp.distance,2) as distance, a.idTask from tab3 a) t ON t.idTask=tr.idTask
GROUP BY t.name, MONTH(tr.assigned)

Answer 2

This is your query:

SELECT SUM(ROUND(rp.distance, 2)), t.name, MONTH(tr.assigned)
FROM tab1 tr JOIN
     tab2 rp
     ON rp.robot = tr.robot JOIN
     tab3 t
     ON t.idTask = tr.idTask
GROUP BY t.name, MONTH(tr.assigned);

It looks reasonable. So, before radically rewriting it, do you have the following indexes: tab1(robot, idtask, assigned) , tbl2(robot) , and tab3(idtask, name) .

Next, perhaps the volume of data is just too big. MySQL is pretty bad when it comes to group by -- and 10 seconds for a group by suggests a lot of data. Is there any way you can reduce the size of the data? What are the sizes of the tables? How large is the result set?

If you have lots of data, then 20 seconds may not be unreasonable.

Answer 3

Looks like you are grouping by string, so you need to add index on name column. You have to always use indexes when joining MySQL tables