MySQL Performance Optimization

MySQL Performance MySQL Performance OptimizationOptimization

Presenter: Avishek Kumar SharmaMindfire Solutions Date: 16/04/2014

OCP MySQL 5.0 - Oracle Certified Professional

Connect Me :Facebook: https://www.facebook.com/avishekkumar.sharma.5LinkedIn: http://in.linkedin.com/pub/avishek-kumar-sharma/31/798/509Twitter: https://twitter.com/sharma_avishekGoogle+: https://plus.google.com/103775319893123886681/posts

Blog:http://avisheksharma.wordpress.com

Contact Me :Email: [email protected] / [email protected]: mfsi_avisheks

About Me

3

Today's Talk

1. Choose Correct engine2. Data Types and Schema Design 3. Understanding Query Execution plan4. Profiling5. Role of Indexing6. Optimizing GROUP BY/ORDER BY7. Covering index8. Scaling data9. Conclusion

4

Choose Correct engine“You should use InnoDB for your tables unless you have a compelling need to use a different engine” - High Performance MySQL by Peter Zaitsev

MyISAM vs InnoDBMyISAM vs InnoDB

InnoDB: Transaction Support, Crash-safeRow Level lockingMix for mix of Updates and Select Statements

MyISAM:Best for Read-heavy applicationsDoesn't scale very well when there are a lot of writes.

Memory:Best for small sized temp summary table, low latency, non-persistent

InnoDB also support “Full Text Search” in MySQL 5.6

5

Data Types and Schema Design

1. Add `id` to each table ( id INT(11) UNSIGNED PRIMARY KEY AUTO_INCREMENT)

2. Smaller sizes are usually better

3. Have a discussion on NULL VS blank VS 0. (choose not null for KEY columns).

4. CHAR vs VARCHAR ( Space Vs Performance)

6

Data Types and Schema Design

5. ENUM : For storing strings values which have fixed and small sample space use ENUM data type

6. Identifier for table joins should be of the same data type to improve performance by reducing type conversion.

7. Normalization vs Denormalization (Faster Joins Vs Redundencies)

7

Have a look what procedure analyse() 's saying.

MySQLClint(DB:as_01):select id from dochead procedure analyse()\G

*************************** 1. row *************************** Field_name: as_01.dochead.id

Min_value: 1 Max_value: 11294 Min_length: 1

Max_length: 5 Empties_or_zeros: 0

Nulls: 0 Avg_value_or_avg_length: 5678.8494 Std: 3270.5730 Optimal_fieldtype: SMALLINT(5) UNSIGNED NOT NULL

1 row in set (0.01 sec)

This is what MySQL suggesting the datatype to be

8

Query Execution Plan

Using the EXPLAIN keyword can give you insight on what MySQL is doing to execute your query.

Eg: explain select col1 from <table_name> where col2=123 group by col3 order by col3;

9

Explain some field values

Some of the important columns we gonna consider are:

● type: const, eq_ref, ref, fulltext, range, index, ALL(sort order by speed)

● possible_keys: all the candicates for the query

● key: selected index (could be more than one)

● key_len: index length (agg for composite indexes)

● rows: no of approximate rows traversal

● extra: using filesort, using index, using temporary,using where

10

SET Profiling=ON

Profiling helps us analysing query resources● mysql> SET profiling=1;

● mysql> run your query....

● mysql> SHOW profiles;

● mysql> SHOW profile [CPU] FOR QUERY 1;

11

Role of Indexing

Index optimization is perhaps the most powerful way to improve query performance.

When performance problems occur:

– Add indexes

– Rewrite your queries

– Or both

Do you need to fetch data (often on disk) ?

– If the index contains the data, you don't

– If you don't, your query is covered by an index (=index-only query)

12

Isolating the Column

MySQL generally can’t use indexes on columns unless the columns are isolated in the query. “Isolating” the column means it should not be part of an expression or be inside a function in the query.

Wrong: – mysql> SELECT actor_id FROM sakila.actor WHERE actor_id + 1 = 5;

Correct: – mysql> SELECT actor_id FROM sakila.actor WHERE actor_id = 4;

13

Keep queries deterministic(Query Cache)

Since the return result of the function can change, MySQL decides to disable query caching for that query.

// query cache does NOT work

$r = mysql_query("SELECT username FROM user WHERE signup_date >= CURDATE()");

// query cache works!

$today = date("Y-m-d");

$r = mysql_query("SELECT username FROM user WHERE signup_date >= '$today'");

NB: This applies to all non-deterministic functions like NOW() and RAND() etc

14

Column Prefix Indexes

● Smaller is usually better for obvious reasons

● Allow index BLOB/TEXT columns

CREATE TABLE user_address(

`id` INT UNSIGNED

...

...

`address` varchar(255)

PRIMARY KEY(`id`)

)Eg: CREATE INDEX part_of_address ON user_address(address(10));

15

Choose correct prefix length

● Check the full selectivity of your column.

● Check selectivity of several prefix lengths.

16

Here we go

● Choose the prefix length having almost the same selectivity with the full column selectivity. In the example, prefix length 7 selectivity is almost the same with the full column selectivity so we will used it as the index.

● Create prefix indexes.

17

Optimize GROUP BY/ORDER BY

● GROUP BY and ORDER BY queries do post-retrieval work

● GROUP BY compels to create “Temp Table”

● and ORDER BY “File Sort”

● Indexing can help get rid of this work.

18

Ignore GROUP BY sorting

● GROUP BY does implicit sortingMySQLClint(DB:as_01)>

EXPLAIN

SELECT DISTINCT c.id,legal_name

FROM contractor c

JOIN dochead dh ON c.id=dh.contractor_to

WHERE

dh.system_id=623

AND invoice.doc_type=”invoice”

GROUP BY c.id

ORDER BY null\G

19

Check EXPLAIN's Extra

*************************** 1. row ***************************

id: 1

type: ref

....

possible_keys: dochead_contractor_to,dochead_system_id

key: dochead_system_id

key_len: 4

ref: const

rows: 306

Extra: Using where; Using temporary; Using filesort

20

Check EXPLAIN's Changes

*************************** 1. row ***************************

id: 1

type: ref

....

....

possible_keys: dochead_contractor_to,dochead_system_id

key: dochead_system_id

key_len: 4

ref: const

rows: 306

Extra: Using where; Using temporary; Using filesort

Using file sort will be vanished if we dont bother for ordering

21

Sometimes derived tables help

● Logic is less amount of data rows has to be traverse

● Filter results by cheapest first

● Limitation is Indexes cannot be used, and it has to be less row size.

22

Covering Index

● Indexes need to be designed for the whole query, not just the WHERE clause. Contains (or “covers”) all the data needed to satisfy a query.

Query with traditional index:Query with traditional index:

– Get right rows with indexGet right rows with index

– Get data from rowsGet data from rows

– Send data back to clientSend data back to client

Index-covered query:Index-covered query:

– Get right rows with indexGet right rows with index

– Get data from rowsGet data from rows

– Send data back to clientSend data back to client

23

Case Study

SELECT sum(value)SELECT sum(value)

FROM Table1FROM Table1

WHERE item_id=? AND category_id = ?WHERE item_id=? AND category_id = ?

GROUP BY customer_id;GROUP BY customer_id;

ALTER TABLE Table1ALTER TABLE Table1

ADD INDEX t1_index(item_id,category_id,customer_id,value)ADD INDEX t1_index(item_id,category_id,customer_id,value)

24

Check if used covering index

mysql> EXPLAIN SELECT sum(value)

FROM Table1

WHERE item_id=? AND category_id = ?

GROUP BY customer_id;

*************************** 1. row ***************************

table: Table1

.....

possible_keys: t1_index

key: t1_index

Extra: Using where;Using index

This signs that query used covering index.

25

Column Order in Covering Index

1. Const or equality

WHERE a = 5

2. Range

WHERE a = 5 AND b > 5

3. Order or Group By

WHERE a = 5 AND b > 5 GROUP BY c

WHERE a = 5 AND b > 5 ORDER BY c DESC

4. Select

count(d), sum(d)

26

Lets put all together(Scaling)

● Analyse explain and keep on trying/improving– STRAIGHT_JOIN,

– FORCE INDEX/USE INDEX,

– IGNORE INDEX

● Memory/Disk/CPU/Network

● Upgrade MySQL Server

● Move to Amazon RDS– I strongly recommend, have personal experience

27

Thank you :)Thank you :)

28

https://www.facebook.com/MindfireSolutions

http://www.linkedin.com/company/mindfire-solutions

http://twitter.com/mindfires

www.mindfiresolutions.com

Contact Us @