This article mainly introduces MySQL table partitioning, such as what is table partitioning, why do we need to partition a table, and detailed explanations of the 4 types of table partitioning, etc. Friends who need it can refer to the following
1. What is table partitioning
In layman's terms, table partitioning is to divide a large table into several small tables according to conditions. MySQL5.1 began to support data table partitioning.
For example, if there are more than 6 million records in a user table, the table can be partitioned according to the storage date, or the table can be partitioned according to the location. Of course, it can also be divided according to other conditions.
Second, why to partition the table
In order to improve the scalability of large tables and tables with various access patterns, manageability and improve database efficiency.
Some of the advantages of partitions include:
1) More data can be stored than a single disk or file system partition.
2) For those data that have lost the meaning of preservation, it is usually easy to delete those data by deleting the partitions related to those data. Conversely, in some cases, the process of adding new data can be conveniently implemented by adding a new partition specifically for that new data. Other advantages generally associated with partitioning include those listed below. These features in MySQL partitioning are not currently implemented, but are high on our priority list; we hope to include these features in the production release of 5.1.
3), some queries can be greatly optimized, mainly because the data that satisfies a given WHERE statement can only be stored in one or more partitions, so that there is no need to find other remaining partitions when searching. Because partitions can be modified after a partitioned table is created, data can be reorganized to improve the efficiency of commonly used queries when the partitioning scheme is first configured without doing so.
4) Queries involving aggregate functions such as SUM() and COUNT() can be easily processed in parallel. A simple example of such a query is "SELECT salesperson_id, COUNT (orders) as order_total FROM sales GROUP BY salesperson_id;". By "parallel", it means that the query can be performed on each partition concurrently, and the final result is simply obtained by summing up all the partitions.
5) Obtain greater query throughput by dispersing data queries across multiple disks.
3. Partition Type
· RANGE Partition: Assign multiple rows to a partition based on column values belonging to a given continuous interval.
· LIST partitioning: Similar to partitioning by RANGE, the difference is that LIST partitioning is based on the column value matching a value in a discrete value set to select.
HASH partitioning: Partitioning that is selected based on the return value of a user-defined expression computed using the column values of the rows to be inserted into the table. This function can contain any expression valid in MySQL that yields a non-negative integer value.
· KEY partition: similar to partition by HASH, the difference is that KEY partition only supports calculation of one or more columns, and the MySQL server provides its own hash function. One or more columns must contain integer values.
1. RANGE partition
Assigns multiple rows to partitions based on column values that belong to a given contiguous interval. These intervals must be contiguous and cannot overlap each other, and are defined using the VALUES LESS THAN operator. The following are examples.
CREATE TABLE employees (
id INT NOT NULL,
fname VARCHAR(30),
lname VARCHAR(30),
hired DATE NOT NULL DEFAULT '1970-01-01',
separated DATE NOT NULL DEFAULT '9999-12-31',
job_code INT NOT NULL,
store_id INT NOT NULL
)
partition BY RANGE (store_id) (
partition p0 VALUES LESS THAN (6),
partition p1 VALUES LESS THAN (11),
partition p2 VALUES LESS THAN (16),
partition p3 VALUES LESS THAN (21)
);According to this partitioning scheme, all rows corresponding to employees working in stores 1 to 5 are stored in partition P0, employees of stores 6 to 10 are stored in P1, and so on. Note that each partition is defined in order, from lowest to highest. This is a requirement of the PARTITION BY RANGE syntax; in this respect it is similar to a "switch ... case" statement in C or Java.
For a new row containing data (72, 'Michael', 'Widenius', '1998-06-25', NULL, 13), it can be easily determined that it will be inserted into the p2 partition, but if a number is added For the 21st store, what will happen? Under this scheme, since there is no rule to include stores with store_id greater than 20, the server will not know where to save the row and will cause an error. This error can be avoided by using a "catchall" VALUES LESS THAN clause in the CREATE TABLE statement that provides all values greater than an explicitly specified highest value:
CREATE TABLE employees (
id INT NOT NULL,
fname VARCHAR(30),
lname VARCHAR(30),
hired DATE NOT NULL DEFAULT '1970-01-01',
separated DATE NOT NULL DEFAULT '9999-12-31',
job_code INT NOT NULL,
store_id INT NOT NULL
)
PARTITION BY RANGE (store_id) (
PARTITION p0 VALUES LESS THAN (6),
PARTITION p1 VALUES LESS THAN (11),
PARTITION p2 VALUES LESS THAN (16),
PARTITION p3 VALUES LESS THAN MAXVALUE
)；MAXVALUE represents the largest possible integer value. Now, all rows with a store_id column value greater than or equal to 16 (the highest value defined) will be stored in partition p3. At some point in the future, when the number of stores has grown to 25, 30, or more, an ALTER TABLE statement can be used to add new partitions for stores 21-25, 26-30, and so on.
In almost the same structure, you can also split the table based on the employee's job code, that is, on a continuous interval based on the value of the job_code column. For example - assuming a 2-digit work code is used for general (in-store) workers, a three-digit code for office and support staff, and a four-digit code for management, you can create this partition table using the following statement:
CREATE TABLE employees (
id INT NOT NULL,
fname VARCHAR(30),
lname VARCHAR(30),
hired DATE NOT NULL DEFAULT '1970-01-01',
separated DATE NOT NULL DEFAULT '9999-12-31',
job_code INT NOT NULL,
store_id INT NOT NULL
)
PARTITION BY RANGE (job_code) (
PARTITION p0 VALUES LESS THAN (100),
PARTITION p1 VALUES LESS THAN (1000),
PARTITION p2 VALUES LESS THAN (10000)
)；在这个例子中, 店内工人相关的所有行将保存在分区p0中，办公室和支持人员相关的所有行保存在分区p1中，管理层相关的所有行保存在分区p2中。
在VALUES LESS THAN 子句中使用一个表达式也是可能的。这里最值得注意的限制是MySQL 必须能够计算表达式的返回值作为LESS THAN (<)比较的一部分；因此，表达式的值不能为NULL 。由于这个原因，雇员表的hired, separated, job_code,和store_id列已经被定义为非空（NOT NULL）。
除了可以根据商店编号分割表数据外，你还可以使用一个基于两个DATE （日期）中的一个的表达式来分割表数据。例如，假定你想基于每个雇员离开公司的年份来分割表，也就是说，YEAR(separated)的值。实现这种分区模式的CREATE TABLE 语句的一个例子如下所示：
CREATE TABLE employees (
id INT NOT NULL,
fname VARCHAR(30),
lname VARCHAR(30),
hired DATE NOT NULL DEFAULT '1970-01-01',
separated DATE NOT NULL DEFAULT '9999-12-31',
job_code INT,
store_id INT
)
PARTITION BY RANGE (YEAR(separated)) (
PARTITION p0 VALUES LESS THAN (1991),
PARTITION p1 VALUES LESS THAN (1996),
PARTITION p2 VALUES LESS THAN (2001),
PARTITION p3 VALUES LESS THAN MAXVALUE
)；在这个方案中，在1991年前雇佣的所有雇员的记录保存在分区p0中，1991年到1995年期间雇佣的所有雇员的记录保存在分区p1中， 1996年到2000年期间雇佣的所有雇员的记录保存在分区p2中，2000年后雇佣的所有工人的信息保存在p3中。
RANGE分区在如下场合特别有用：
1）、 当需要删除一个分区上的“旧的”数据时,只删除分区即可。如果你使用上面最近的那个例子给出的分区方案，你只需简单地使用 “ALTER TABLE employees DROP PARTITION p0；”来删除所有在1991年前就已经停止工作的雇员相对应的所有行。对于有大量行的表，这比运行一个如“DELETE FROM employees WHERE YEAR (separated) <= 1990；”这样的一个DELETE查询要有效得多。
2）、想要使用一个包含有日期或时间值，或包含有从一些其他级数开始增长的值的列。
3）、经常运行直接依赖于用于分割表的列的查询。例如，当执行一个如“SELECT COUNT(*) FROM employees WHERE YEAR(separated) = 2000 GROUP BY store_id；”这样的查询时，MySQL可以很迅速地确定只有分区p2需要扫描，这是因为余下的分区不可能包含有符合该WHERE子句的任何记录。
注释：这种优化还没有在MySQL 5.1源程序中启用，但是，有关工作正在进行中。
2.LIST分区
类似于按RANGE分区，区别在于LIST分区是基于列值匹配一个离散值集合中的某个值来进行选择。
LIST分区通过使用“PARTITION BY LIST(expr)”来实现，其中“expr” 是某列值或一个基于某个列值、并返回一个整数值的表达式，然后通过“VALUES IN (value_list)”的方式来定义每个分区，其中“value_list”是一个通过逗号分隔的整数列表。
注释：在MySQL 5.1中，当使用LIST分区时，有可能只能匹配整数列表。
CREATE TABLE employees (
id INT NOT NULL,
fname VARCHAR(30),
lname VARCHAR(30),
hired DATE NOT NULL DEFAULT '1970-01-01',
separated DATE NOT NULL DEFAULT '9999-12-31',
job_code INT,
store_id INT
)；假定有20个音像店，分布在4个有经销权的地区，如下表所示：
====================
地区 商店ID 号
------------------------------------
北区 3, 5, 6, 9, 17
东区 1, 2, 10, 11, 19, 20
西区 4, 12, 13, 14, 18
中心区 7, 8, 15, 16
====================
要按照属于同一个地区商店的行保存在同一个分区中的方式来分割表，可以使用下面的“CREATE TABLE”语句：
CREATE TABLE employees (
id INT NOT NULL,
fname VARCHAR(30),
lname VARCHAR(30),
hired DATE NOT NULL DEFAULT '1970-01-01',
separated DATE NOT NULL DEFAULT '9999-12-31',
job_code INT,
store_id INT
)
PARTITION BY LIST(store_id)
PARTITION pNorth VALUES IN (3,5,6,9,17),
PARTITION pEast VALUES IN (1,2,10,11,19,20),
PARTITION pWest VALUES IN (4,12,13,14,18),
PARTITION pCentral VALUES IN (7,8,15,16)
)；这使得在表中增加或删除指定地区的雇员记录变得容易起来。例如，假定西区的所有音像店都卖给了其他公司。那么与在西区音像店工作雇员相关的所有记录（行）可以使用查询“ALTER TABLE employees DROP PARTITION pWest；”来进行删除，它与具有同样作用的DELETE （删除）查询“DELETE query DELETE FROM employees WHERE store_id IN (4,12,13,14,18)；”比起来，要有效得多。
【要点】：如果试图插入列值（或分区表达式的返回值）不在分区值列表中的一行时，那么“INSERT”查询将失败并报错。例如，假定LIST分区的采用上面的方案，下面的查询将失败：
INSERT INTO employees VALUES(224, 'Linus', 'Torvalds', '2002-05-01', '2004-10-12', 42, 21);这是因为“store_id”列值21不能在用于定义分区pNorth, pEast, pWest,或pCentral的值列表中找到。要重点注意的是，LIST分区没有类似如“VALUES LESS THAN MAXVALUE”这样的包含其他值在内的定义。将要匹配的任何值都必须在值列表中找到。
LIST分区除了能和RANGE分区结合起来生成一个复合的子分区，与HASH和KEY分区结合起来生成复合的子分区也是可能的。
3.HASH分区
基于用户定义的表达式的返回值来进行选择的分区，该表达式使用将要插入到表中的这些行的列值进行计算。这个函数可以包含MySQL 中有效的、产生非负整数值的任何表达式。
要使用HASH分区来分割一个表，要在CREATE TABLE 语句上添加一个“PARTITION BY HASH (expr)”子句，其中“expr”是一个返回一个整数的表达式。它可以仅仅是字段类型为MySQL 整型的一列的名字。此外，你很可能需要在后面再添加一个“PARTITIONS num”子句，其中num 是一个非负的整数，它表示表将要被分割成分区的数量。
CREATE TABLE employees (
id INT NOT NULL,
fname VARCHAR(30),
lname VARCHAR(30),
hired DATE NOT NULL DEFAULT '1970-01-01',
separated DATE NOT NULL DEFAULT '9999-12-31',
job_code INT,
store_id INT
)
PARTITION BY HASH(store_id)
PARTITIONS 4；如果没有包括一个PARTITIONS子句，那么分区的数量将默认为1。 例外： 对于NDB Cluster（簇）表，默认的分区数量将与簇数据节点的数量相同，
这种修正可能是考虑任何MAX_ROWS 设置，以便确保所有的行都能合适地插入到分区中。
1.)LINER HASH
MySQL还支持线性哈希功能，它与常规哈希的区别在于，线性哈希功能使用的一个线性的2的幂（powers-of-two）运算法则，而常规 哈希使用的是求哈希函数值的模数。
线性哈希分区和常规哈希分区在语法上的唯一区别在于，在“PARTITION BY” 子句中添加“LINEAR”关键字。
CREATE TABLE employees (
id INT NOT NULL,
fname VARCHAR(30),
lname VARCHAR(30),
hired DATE NOT NULL DEFAULT '1970-01-01',
separated DATE NOT NULL DEFAULT '9999-12-31',
job_code INT,
store_id INT
)
PARTITION BY LINEAR HASH(YEAR(hired))
PARTITIONS 4；假设一个表达式expr, 当使用线性哈希功能时，记录将要保存到的分区是num 个分区中的分区N，其中N是根据下面的算法得到：
1. 找到下一个大于num.的、2的幂，我们把这个值称为V ，它可以通过下面的公式得到：
2. V = POWER(2, CEILING(LOG(2, num)))
（例如，假定num是13。那么LOG(2,13)就是3.7004397181411。 CEILING(3.7004397181411)就是4，则V = POWER(2,4), 即等于16）。
3. 设置 N = F(column_list) & (V - 1).
4. 当 N >= num:
· 设置 V = CEIL(V / 2)
· 设置 N = N & (V - 1)
例如，假设表t1，使用线性哈希分区且有4个分区，是通过下面的语句创建的：
CREATE TABLE t1 (col1 INT, col2 CHAR(5), col3 DATE)
PARTITION BY LINEAR HASH( YEAR(col3) )
PARTITIONS 6;
现在假设要插入两行记录到表t1中，其中一条记录col3列值为'2003-04-14'，另一条记录col3列值为'1998-10-19'。第一条记录将要保存到的分区确定如下：
V = POWER(2, CEILING(LOG(2,7))) = 8
N = YEAR('2003-04-14') & (8 - 1)
= 2003 & 7
= 3(3 >= 6 为假（FALSE）: 记录将被保存到#3号分区中)
第二条记录将要保存到的分区序号计算如下：
V = 8
N = YEAR('1998-10-19') & (8-1)
= 1998 & 7 = 6(6 >= 4 为真（TRUE）: 还需要附加的步骤)
N = 6 & CEILING(5 / 2)
= 6 & 3 = 2(2 >= 4 为假（FALSE）: 记录将被保存到#2分区中)
按照线性哈希分区的优点在于增加、删除、合并和拆分分区将变得更加快捷，有利于处理含有极其大量（1000吉）数据的表。它的缺点在于，与使用
常规HASH分区得到的数据分布相比，各个分区间数据的分布不大可能均衡。
4.KSY分区
类似于按HASH分区，区别在于KEY分区只支持计算一列或多列，且MySQL 服务器提供其自身的哈希函数。必须有一列或多列包含整数值。
CREATE TABLE tk (
col1 INT NOT NULL,
col2 CHAR(5),
col3 DATE
)
PARTITION BY LINEAR KEY (col1)
PARTITIONS 3;在KEY分区中使用关键字LINEAR和在HASH分区中使用具有同样的作用，分区的编号是通过2的幂（powers-of-two）算法得到，而不是通过模数算法。