mysql实现分区

懒惰の天真热2023-12-30 19:35

文章目录

- 一、根据时间创建分区
- 二、根据业务主键id创建分区

一、根据时间创建分区

创建时间分区表

java 复制代码

CREATE TABLE `table_time`  (
  `ID` int(11) NOT NULL AUTO_INCREMENT COMMENT '主键ID',
  `STA_ID` int(11) NULL DEFAULT NULL COMMENT '业务ID',
  `STA_NAME` varchar(100) CHARACTER SET utf8 COLLATE utf8_general_ci NULL DEFAULT NULL COMMENT '业务名称',
  `MONITOR_TIME` datetime(0) NOT NULL   COMMENT '监测时间',
  PRIMARY KEY (`ID`,`MONITOR_TIME`) USING BTREE
) ENGINE = InnoDB AUTO_INCREMENT = 2111262 CHARACTER SET = utf8 COLLATE = utf8_general_ci COMMENT = '测试表' ROW_FORMAT = Dynamic
partition by range (to_days(MONITOR_TIME)) (
 	PARTITION p0 VALUES LESS THAN (TO_DAYS('2023-05-01')),
    PARTITION p1 VALUES LESS THAN (TO_DAYS('2023-06-01')) 	
    );

查看分区是否生效

java 复制代码

select partition_name part, partition_expression expr, partition_description descr, table_rows
 from  INFORMATION_SCHEMA.partitions  where TABLE_SCHEMA="库名" AND TABLE_NAME="表名";

时间格式化

ps：由于我们是以时间作为分区，时间会被转成特殊的数字，所以如果想要查看分区属于哪个时间，还需要格式化下时间数据才可以查看
java 复制代码
```
 SELECT DATE_SUB(DATE_ADD('0000-01-01', INTERVAL 740543 DAY), INTERVAL 1 DAY);
```

二、根据业务主键id创建分区

创建业务主键分区表

java 复制代码

CREATE TABLE `table_id`  (
  `id` varchar(32) CHARACTER SET utf8 COLLATE utf8_general_ci NOT NULL COMMENT '主键  字符串',
  `car_id` int(11) NOT NULL  COMMENT '车辆ID',
  `car_no` varchar(32) CHARACTER SET utf8 COLLATE utf8_general_ci NULL DEFAULT NULL COMMENT '车牌号'
) ENGINE = InnoDB CHARACTER SET = utf8 COLLATE = utf8_general_ci COMMENT = '车辆行驶轨迹' ROW_FORMAT = Dynamic
PARTITION BY LIST COLUMNS(car_id)
(PARTITION p0 VALUES IN (0) ENGINE = InnoDB);

根据业务主键添加分区

java 复制代码

ALTER TABLE table_id ADD PARTITION (PARTITION p_197401 VALUES IN (197401) ENGINE = InnoDB);

插入数据

ps：插入数据的时候，可以指定分区。如果不指定分区，也会自行根据分区入库。
java 复制代码
```
insert into table_id partition(p_197401) (car_id,car_no)  values XXXXX;
```