reduciendo el tiempo de ejecución de consultas mysql

SELECT mt.Test_ID,
       mtp.Test_Plan_Doc_No,
       mp.Group_Name,
       mp.Point_Name,
       mp.Limit_1,
       mp.Limit_2,
       mp.Datatype,
       mt.Start_Date,
       mv.Measurement_Value,
       mv.Status_Value
FROM   measurement_test mt
       INNER JOIN measurement_values mv
               ON mt.Test_ID = mv.Test_ID
       INNER JOIN measurement_point mp
               ON mv.Point_ID = mp.Point_ID
       INNER JOIN measurement_test_plan mtp
               ON mtp.Test_ID = mt.Test_ID
WHERE  mtp.Test_Plan_Doc_No IN ( 'test1', 'test2' )
       AND mp.Group_Name = 'gp_name'

Hola chicos.

The above is the query that i am executing to create a view.

But i have a problem as follows:

the measurement value table has around 82 million rows. the measurement point table has around say 5 million rows measurement test plan and measurement test tables have around 9000 - 100000 rows.

My problem is when i execute the above query the execution time is 8 minutes and the result that is generated has only 400 rows.

Is there any possible way to reduce the execution time of the above query?

NOTE : I am using the above query in my web page

preguntado el 28 de agosto de 12 a las 09:08

what are the rough duration of execution for the said query? -

have you created necessary indices -

@ianace : its around 8 minutes -

If you run explain on the query does it make use of your indexes? If not, can you now see some to add that will help? -

Publique la salida de SHOW CREATE TABLE xxx con xxx in (measurement_test, measurement_values, measurement_point, measurement_test_plan) -

3 Respuestas

First off, for an efficient join, you should keep the tables in increasing order of number of rows. This reduces the number of row scans drastically. So for your query, the join order should be measurement_test mt natural join measurement_point mp natural join measurement_values mv. Also ensure that the join columns have indexes defined on them, and they have exactly identical datatypes. char(15) y varchar(15) are considered similar in MySQL, but not char(15) y varchar(10).

Respondido 28 ago 12, 10:08

On 2nd thoughts, after going through a lot of other Stackoverflow posts and related articles on internet, I think that the order of the join columns is immaterial. Whatever be the way, it's always going to be a Cartesian product. The indexes and column types have to be taken care of, of course. - sexybestia

reference from you @comment you said u did not create any index on any table. So CREAR ÍNDICE on those column on which are used in ¿Dónde? y ON condición

in your case create index on 

mt.Test_ID, 
mv.Test_ID, 
mv.Point_ID,
mp.Point_ID, 
mtp.Test_ID ,
mp.group_name ,
mtp.Test_Plan_Doc_No

Here I Rewrite your Query:

SELECT mt.Test_ID,
       mtp.Test_Plan_Doc_No,
       mp.Group_Name,
       mp.Point_Name,
       mp.Limit_1,
       mp.Limit_2,
       mp.Datatype,
       mt.Start_Date,
       mv.Measurement_Value,
       mv.Status_Value
FROM   measurement_test mt
       INNER JOIN measurement_values mv USING(Test_ID)            
       INNER JOIN measurement_point mp USING (Point_ID)
       INNER JOIN measurement_test_plan mtp USING(Test_ID)
WHERE  mtp.Test_Plan_Doc_No IN ( 'test1', 'test2' )
       AND mp.Group_Name = 'gp_name'

Respondido 28 ago 12, 10:08

Make sure you have proper indexes for each column used in the WHERE clause.

You can use a query analyzer to see whether or not the indexes are used.

Respondido 28 ago 12, 09:08

No es la respuesta que estás buscando? Examinar otras preguntas etiquetadas or haz tu propia pregunta.