MySQL Partitioning

Introduction of MySQL Partitioning

MySQL Partitioning is used to improve performance and reduce the cost of storing a large amount of data. By using partitioning, we are splitting the tables, indexes, and index-organized tables into smaller pieces by which queries can run faster. The partitioning can be done in two major forms:

• Horizontal Partitioning
• Vertical Partitioning.

Horizontal partitioning is a partition that will divide the table rows into multiple partitions. It doesn’t have any attribute missing in the partitions. Vertical partitioning is to partition a table row into multiple partitions.

In vertical partitioning, most referenced columns in one table and the rest of the columns that are not frequently referenced will be stored in another partition. Partition support has to be provided to MySQL. To enable we do it by option DWITH_PARTITION_STORAGE_ENGINE

Syntax of MySQL Partitioning

Below is the syntax for the partition declaration: –

<Create table definition>
<table options>
<partition options>

The partition options are specified as below: –

partition_options:

PARTITION BY
{ [LINEAR] HASH(expr)
| [LINEAR] KEY(column_list)
| RANGE(expr)
| LIST(expr) }
[PARTITIONS num] [SUBPARTITION BY
{ [LINEAR] HASH(expr)
| [LINEAR] KEY(column_list) }
[SUBPARTITIONS num] ]

Types of Partitioning

Different Partition types are mentioned below: –

1. LIST Partitioning
2. RANGE Partitioning
3. COLUMNS Partitioning
4. HASH Partitioning
5. KEY Partitioning
6. Sub partitioning

Now let us see in detail the partitions: –

1. LIST Partitioning

LIST Partitioning allows partitioning the data based on the set of values defined at the table creation time.

Syntax:

PARTITON BY LIST( <expression> )
(
PARTITION <partition_name1> VALUES IN ( <Desired partition1 values>)
PARTITION <partition_name2> VALUES IN ( <Desired partition2 values>)
.
.
PARTITION <partition_nameN> VALUES IN ( <Desired partitionN values>)
)

2. RANGE Partitioning

RANGE Partitioning allows to the partition of the data based on the RANGE. The specified RANGE should be contiguous.

Syntax: 

PARTITON BY RANGE ( <expression> )
(
PARTITION <partition_name1> VALUES LESS THAN( <Desired partition1 Date>)
PARTITION <partition_name2> VALUES LESS THAN ( <Desired partition2 Date>)
.
.
PARTITION <partition_nameN> VALUES LESS THAN ( <Desired partitionNDate>)
)

3. COLUMN Partitioning

COLUMN Partitioning can be done based on multiple columns. Therefore, we have two types of COLUMN Partitioning.

• RANGE COLUMNS Partitioning and
• LIST COLUMNS Partitioning.

Range Columns

It is similar to the RANGE partition. Here RANGE COLUMNS accepts a list of one or more columns as the partition key.

Syntax:

PARTITON BY RANGE COLUMNS( <column list> )
(
PARTITION <partition_name1> VALUES LESS THAN ( <Desired value list >)
PARTITION <partition_name2> VALUES LESS THAN ( <Desired value list >)
.
.
PARTITION <partition_nameN> VALUES LESS THAN ( <Desired value list >)
)

List Columns

It is similar to the LIST partition. Here LIST COLUMNS accepts a list of one or more columns as the partition key.

Syntax:

PARTITON BY LIST COLUMNS ( <column list> )
(
PARTITION <partition_name1> VALUES LESS THAN ( <Desired value list >)
PARTITION <partition_name2> VALUES LESS THAN ( <Desired value list >)
.
.
PARTITION <partition_nameN> VALUES LESS THAN ( <Desired value list >)
)

4. HASH Partitioning

Here in HASH, Partitioning the partition is done based on the column value and the number of partition.

Syntax:

PARTITION BY HASH ( <expression> )
PARTITIONS <NUM Clause>;

5. KEY Partitioning

KEY Partition is similar to HASH partition. The MySQL server will do the Hashing function for key partition.

Syntax:

PARTITION BY KEY( <expression> )
PARTITIONS <NUM Clause>;

6. SUB Partitioning

Sub partition is to further partition the partition table.

Syntax:

PARTITON BY LIST <partition type1>( <expression> )
SUBPARTITION BY <partition type2>( <expression> )
SUBPARTITION <num clause>

How does MySQL Partitioning Works?

Now let us see how the partition works in the tables: –

1. RANGE Partitioning

Now let us create the table and see how partition works on it:

Code:

CREATE TABLE MONTHLY_SALES
(
SALES_NO INT,
SALES_DATE timestamp,
CUST_CODE INT,
TOTAL_AMOUNT int
)
PARTITION BY RANGE( UNIX_TIMESTAMP(SALES_DATE))
(
PARTITION PT1 VALUES LESS THAN (UNIX_TIMESTAMP('2020-04-01')),
PARTITION PT2 VALUES LESS THAN (UNIX_TIMESTAMP('2020-05-01')),
PARTITION PT3 VALUES LESS THAN (UNIX_TIMESTAMP('2020-06-01'))
);

2. LIST Partitioning

Now let us create the table and see how partition works on it:

Code:

CREATE TABLE MONTHLY_SALES_LIST
(
SALES_NO INT,
SALES_DATE timestamp,
CUST_CODE INT,
TOTAL_AMOUNT int
)
PARTITION BY LIST ( CUST_CODE )
(
PARTITION PT1 VALUES IN( 1,2,3,4 ),
PARTITION PT2 VALUES IN ( 5,6,7,8 ),
PARTITION PT3 VALUES IN ( 9,10, 11, 12)
);

3. RANGE COLUMNS Partitioning

Now let us create the table and see how partition works on it:

Code:

CREATE TABLE SALES_RANGE_COLUMNS
(
SALES_NO INT,
CUST_CODE INT,
SALES_NAME VARCHAR(20)
)
PARTITION BY RANGE COLUMNS ( SALES_NO, CUST_CODE, SALES_NAME )
(
PARTITION PT1 VALUES LESS THAN ( 1, 2, 'LUX' ),
PARTITION PT2 VALUES LESS THAN ( 2, 4, 'Paper' ),
PARTITION PT3 VALUES LESS THAN ( 3, 5, 'Pen' )
);

4. LIST COLUMNS Partitioning

Now let us create the table and see how partition works on it:

Code:

CREATE TABLE SALES_LIST _COLUMNS
(
SALES_NO INT,
CUST_CODE INT,
SALES_NAME VARCHAR(20)
)
PARTITION BY LIST COLUMNS ( SALES_NO )
(
PARTITION PT1 VALUES IN ( 1, 2, 3 ),
PARTITION PT2 VALUES IN ( 4, 5, 6 ),
PARTITION PT3 VALUES IN ( 7, 8, 9 )
);

5. Key Partition

Now let us create the table and see how partition works on it:

Code:

CREATE TABLE KEY_PARTITION
(
ID INT,
NAME VARCHAR(10),
LOCATION VARCHAR(20)
)
PARTITION BY KEY( ID)
PARTITIONS 2;

6. HASH Partition

Now let us create the table and see how partition works on it:

Code:

CREATE TABLE MONTHLY_SALES_HASH
(
SALES_NO INT,
SALES_DATE timestamp,
CUST_CODE INT,
TOTAL_AMOUNT int
)
PARTITION BY HASH ( SALES_NO)
PARTITIONS 3;

7. SUB Partitioning

Now let us create the table and see how partition works on it:

Code:

CREATE TABLE MONTHLY_SALES_SUB_PARTITION
(
sale_NO INT,
sale_date DATE,
cust_codeVARCHAR(15),
AMOUNT int
)
PARTITION BY RANGE(YEAR(sale_date) )
SUBPARTITION BY HASH(TO_DAYS(sale_date))
SUBPARTITIONS 4 (
PARTITION pt0 VALUES LESS THAN (1990),
PARTITION pt1 VALUES LESS THAN (2000),
PARTITION pt2 VALUES LESS THAN (2010),
PARTITION pt3 VALUES LESS THAN MAXVALUE
);

Recommended Articles

This is a guide to MySQL Partitioning. Here we also discuss the introduction and how does mysql partitioning works? Along with types of partitioning. You may also have a look at the following articles to learn more –

1. MySQL IF Function
2. MySQL UUID
3. MySQL Merge
4. Toad for MySQL