Speeding up a MySQL query [closed]

Question

Closed. This question does not meet Stack Overflow guidelines. It is not currently accepting answers.

---

This question appears to be off-topic because it lacks sufficient information to diagnose the problem. Describe your problem in more detail or include a minimal example in the question itself.

Closed 9 years ago.

Improve this question

I have this query:

SELECT
    FIELD1, FIELD2, FIELD3, FIELD4, FIELD5
FROM
    MYTABLE
WHERE
    FIELD1 = .... AND
    FIELD2 = .... AND
    FIELD3 = ....
GROUP BY
    FIELD4
;

Note that the only thing I changed are the names of the fields and the table. My query does have 5 fields in the SELECT statement and 1 table in the FROM section. The WHERE and the GROUP BY sections also looks exactly like that: 3 fields with equals condition.

The fields FIELD1, FIELD2 and FIELD3 are both PKs and indexes.

The problem: The query takes almost 8 seconds to finish (and it returns around 30 rows).

The question: Is there any way I can speed up the query, that isn't server related (increasing cache, running the database in a more powerful server, etc...)?

EDIT:

Fields declaration:

FIELD1: VARCHAR(2)
FIELD2: VARCHAR(3)
FIELD3: VARCHAR(7)
FIELD4: VARCHAR(4)
FIELD5: VARCHAR(5)

EDIT 2:

Output of SHOW INDEX: http://ibin.co/1LSSwvr0CVpX

EDIT 3:

Output of EXPLAIN:

select_type: SIMPLE
table: MYTABLE
type: ALL
possible_keys:
key:
key_len:
ref:
rows: 2081601
Extra: Using where; Using temporary; Using filesort

EDIT 4:

Example data:

Chicago = 1 (neighborhoods 001, 002, 003) (neighborhood zone 1, 2, 3) (postal code 10, 20, 30)
NY = 2 (neighborhoods 001, 002, 003) (neighborhood zone 1, 2, 3)(postal code 10, 20, 30)
Los Angeles = 3 (neighborhoods 001, 002, 003) (neighborhood zone 1, 2, 3)(postal code 10, 20, 30)

Data

1    001    1    10
1    002    1    10
1    002    1    20
1    002    2    20
1    003    1    30
2    001    1    10
2    002    1    20
3    001    1    10
3    002    1    10
3    002    2    20

What I do:

Search all postal codes that match TOWN == 1 (Chicago) AND Neighborhood == 001 AND ZONE == 1 and then GROUP BY the postal code. I hope this makes it clear.

Answer 1

problem seems with your searching method just inclosed your values like field1='value' as these are varchar and you are using with out quotes. By this your query will use index and you get performance.

Note: There is no need to use index as mysql do it its own.

Answer 2

With the information now given

1. The table contains streets along with their postal codes, zone, neighbourhood and town.
2. You are looking up postal codes and are not interested in the streets (so you wouldn't select FIELD5 in your example as I understand it now).
3. You can not change the table design, but you can add indexes.

you should have an index on TOWN + NEIGHBORHOOD + ZONE + PC. This contains all information you need and points to all streets associated. The dbms should be clever enough to notice that no table access is needed as all data required is present in the index. That should speed up your query significantly.

Answer 3

1. you mention that you use indexes, make sure an index contains all fields you mention in your select

2. it can be that it returns only 30 rows, but it has to go through many more rows due to

   GROUP BY FIELD1

so you might want to "tighten" your query by adding "LIMIT XX", "SELECT DISTINCT FIELD 1" nad such

Answer 4

I suggest you to take a look to implement a partitioning strategy. For big table it's the only way to significantly improve performance.

Read here: https://dev.mysql.com/doc/refman/5.1/en/partitioning.html