前言
随着业务的发展单库中的分表的数量越来越多, 使用在单库上存放过多的表这样是不合理的。因此，我们就需要考虑将数据根据数据库进行拆分。
一般mysql不建议表的数量超过1000个。当然，这不能一概而论，还需要根据你的数据量，和硬件来确定然后根据自己的服务器调整几个mysql '%open%' 参数，从而来确定你的库应该不超过几张表性能能在可接受范围内。文章源自运维生存时间-https://www.ttlsa.com/mysql/mysql-distributed-database-and-table-prepare/
分库思路
在分库前我们需要确定一下我们应该如何去分库：文章源自运维生存时间-https://www.ttlsa.com/mysql/mysql-distributed-database-and-table-prepare/
1、我们是根据用户ID来进行分库，和分表的思路一样。文章源自运维生存时间-https://www.ttlsa.com/mysql/mysql-distributed-database-and-table-prepare/
2、我们需要在用户表中标记一下用户的数据是在哪个库。文章源自运维生存时间-https://www.ttlsa.com/mysql/mysql-distributed-database-and-table-prepare/
3、在系统设置表中应该记录下当前最大分库数量。文章源自运维生存时间-https://www.ttlsa.com/mysql/mysql-distributed-database-and-table-prepare/
4、在系统设置表中应该记录现在所有分库的库名。文章源自运维生存时间-https://www.ttlsa.com/mysql/mysql-distributed-database-and-table-prepare/
5、在系统设置表中应该记录每个分库的数据库连接描述符信息。文章源自运维生存时间-https://www.ttlsa.com/mysql/mysql-distributed-database-and-table-prepare/
分库规则
我们以 '数字' 为分库标识最终分库的名称如：test_1、test_2、test_3 ...文章源自运维生存时间-https://www.ttlsa.com/mysql/mysql-distributed-database-and-table-prepare/
在新增加库的时候，我们在新库中创建的表的数量是在系统设置表中的最大分表数。如在系统设置表中 name='max_sharding_table_num' 的 value='10'，这时我们会初始化每个分表的个数为10个。文章源自运维生存时间-https://www.ttlsa.com/mysql/mysql-distributed-database-and-table-prepare/
数据迁移
和分表一样我们应该很清楚哪些表是需要进行分库，我们需要分库的表有 buy_order_n、goods_n、sell_order_n、order_goods_n。文章源自运维生存时间-https://www.ttlsa.com/mysql/mysql-distributed-database-and-table-prepare/
我们应该将之前的数据的库名进行统一。如之前test库的数据要先迁移到 test_1 上文章源自运维生存时间-https://www.ttlsa.com/mysql/mysql-distributed-database-and-table-prepare/
提醒：数据迁移慎重，不是说迁移就迁移的。其实也可以不用迁移的，如果不迁移之后的自动分库的代码就需要做多一点的判断。这为了统一我就做了迁移。文章源自运维生存时间-https://www.ttlsa.com/mysql/mysql-distributed-database-and-table-prepare/
数据迁移SQL
-- 创建新库
CREATE DATABASE test_1;
use test;
-- 拼出需要创建的表
SELECT CONCAT('CREATE TABLE test_1.',
TABLE_NAME,
' LIKE ',
TABLE_SCHEMA, '.', TABLE_NAME, ';'
)
FROM information_schema.tables
WHERE TABLE_SCHEMA = 'test';
-- 创建表这边我们不迁移公用的表:user、store、user_guide、system_setting
CREATE TABLE test_1.buy_order_1 LIKE test.buy_order_1;
CREATE TABLE test_1.buy_order_10 LIKE test.buy_order_10;
CREATE TABLE test_1.buy_order_2 LIKE test.buy_order_2;
CREATE TABLE test_1.buy_order_3 LIKE test.buy_order_3;
CREATE TABLE test_1.buy_order_4 LIKE test.buy_order_4;
CREATE TABLE test_1.buy_order_5 LIKE test.buy_order_5;
CREATE TABLE test_1.buy_order_6 LIKE test.buy_order_6;
CREATE TABLE test_1.buy_order_7 LIKE test.buy_order_7;
CREATE TABLE test_1.buy_order_8 LIKE test.buy_order_8;
CREATE TABLE test_1.buy_order_9 LIKE test.buy_order_9;
CREATE TABLE test_1.goods_1 LIKE test.goods_1;
CREATE TABLE test_1.goods_10 LIKE test.goods_10;
CREATE TABLE test_1.goods_2 LIKE test.goods_2;
CREATE TABLE test_1.goods_3 LIKE test.goods_3;
CREATE TABLE test_1.goods_4 LIKE test.goods_4;
CREATE TABLE test_1.goods_5 LIKE test.goods_5;
CREATE TABLE test_1.goods_6 LIKE test.goods_6;
CREATE TABLE test_1.goods_7 LIKE test.goods_7;
CREATE TABLE test_1.goods_8 LIKE test.goods_8;
CREATE TABLE test_1.goods_9 LIKE test.goods_9;
CREATE TABLE test_1.order_goods_1 LIKE test.order_goods_1;
CREATE TABLE test_1.order_goods_10 LIKE test.order_goods_10;
CREATE TABLE test_1.order_goods_2 LIKE test.order_goods_2;
CREATE TABLE test_1.order_goods_3 LIKE test.order_goods_3;
CREATE TABLE test_1.order_goods_4 LIKE test.order_goods_4;
CREATE TABLE test_1.order_goods_5 LIKE test.order_goods_5;
CREATE TABLE test_1.order_goods_6 LIKE test.order_goods_6;
CREATE TABLE test_1.order_goods_7 LIKE test.order_goods_7;
CREATE TABLE test_1.order_goods_8 LIKE test.order_goods_8;
CREATE TABLE test_1.order_goods_9 LIKE test.order_goods_9;
CREATE TABLE test_1.sell_order_1 LIKE test.sell_order_1;
CREATE TABLE test_1.sell_order_10 LIKE test.sell_order_10;
CREATE TABLE test_1.sell_order_2 LIKE test.sell_order_2;
CREATE TABLE test_1.sell_order_3 LIKE test.sell_order_3;
CREATE TABLE test_1.sell_order_4 LIKE test.sell_order_4;
CREATE TABLE test_1.sell_order_5 LIKE test.sell_order_5;
CREATE TABLE test_1.sell_order_6 LIKE test.sell_order_6;
CREATE TABLE test_1.sell_order_7 LIKE test.sell_order_7;
CREATE TABLE test_1.sell_order_8 LIKE test.sell_order_8;
CREATE TABLE test_1.sell_order_9 LIKE test.sell_order_9;
-- 生成插入表的数据
SELECT CONCAT('INSERT INTO ',
TABLE_SCHEMA, '.', TABLE_NAME,
' SELECT * FROM test', '.', TABLE_NAME, ';'
)
FROM information_schema.tables
WHERE TABLE_SCHEMA = 'test_1';
-- 插入数据
INSERT INTO test_1.buy_order_1 SELECT * FROM test.buy_order_1;
INSERT INTO test_1.buy_order_10 SELECT * FROM test.buy_order_10;
INSERT INTO test_1.buy_order_2 SELECT * FROM test.buy_order_2;
INSERT INTO test_1.buy_order_3 SELECT * FROM test.buy_order_3;
INSERT INTO test_1.buy_order_4 SELECT * FROM test.buy_order_4;
INSERT INTO test_1.buy_order_5 SELECT * FROM test.buy_order_5;
INSERT INTO test_1.buy_order_6 SELECT * FROM test.buy_order_6;
INSERT INTO test_1.buy_order_7 SELECT * FROM test.buy_order_7;
INSERT INTO test_1.buy_order_8 SELECT * FROM test.buy_order_8;
INSERT INTO test_1.buy_order_9 SELECT * FROM test.buy_order_9;
INSERT INTO test_1.goods_1 SELECT * FROM test.goods_1;
INSERT INTO test_1.goods_10 SELECT * FROM test.goods_10;
INSERT INTO test_1.goods_2 SELECT * FROM test.goods_2;
INSERT INTO test_1.goods_3 SELECT * FROM test.goods_3;
INSERT INTO test_1.goods_4 SELECT * FROM test.goods_4;
INSERT INTO test_1.goods_5 SELECT * FROM test.goods_5;
INSERT INTO test_1.goods_6 SELECT * FROM test.goods_6;
INSERT INTO test_1.goods_7 SELECT * FROM test.goods_7;
INSERT INTO test_1.goods_8 SELECT * FROM test.goods_8;
INSERT INTO test_1.goods_9 SELECT * FROM test.goods_9;
INSERT INTO test_1.order_goods_1 SELECT * FROM test.order_goods_1;
INSERT INTO test_1.order_goods_10 SELECT * FROM test.order_goods_10;
INSERT INTO test_1.order_goods_2 SELECT * FROM test.order_goods_2;
INSERT INTO test_1.order_goods_3 SELECT * FROM test.order_goods_3;
INSERT INTO test_1.order_goods_4 SELECT * FROM test.order_goods_4;
INSERT INTO test_1.order_goods_5 SELECT * FROM test.order_goods_5;
INSERT INTO test_1.order_goods_6 SELECT * FROM test.order_goods_6;
INSERT INTO test_1.order_goods_7 SELECT * FROM test.order_goods_7;
INSERT INTO test_1.order_goods_8 SELECT * FROM test.order_goods_8;
INSERT INTO test_1.order_goods_9 SELECT * FROM test.order_goods_9;
INSERT INTO test_1.sell_order_1 SELECT * FROM test.sell_order_1;
INSERT INTO test_1.sell_order_10 SELECT * FROM test.sell_order_10;
INSERT INTO test_1.sell_order_2 SELECT * FROM test.sell_order_2;
INSERT INTO test_1.sell_order_3 SELECT * FROM test.sell_order_3;
INSERT INTO test_1.sell_order_4 SELECT * FROM test.sell_order_4;
INSERT INTO test_1.sell_order_5 SELECT * FROM test.sell_order_5;
INSERT INTO test_1.sell_order_6 SELECT * FROM test.sell_order_6;
INSERT INTO test_1.sell_order_7 SELECT * FROM test.sell_order_7;
INSERT INTO test_1.sell_order_8 SELECT * FROM test.sell_order_8;
INSERT INTO test_1.sell_order_9 SELECT * FROM test.sell_order_9;
-- 向系统表中添加当前最大分库数量
INSERT INTO test.system_setting
VALUES(NULL, 'max_sharding_database_num', 1);
-- 向系统表中添加分库名前缀
INSERT INTO test.system_setting
VALUES(NULL, 'sharding_database_prefix', 'test');
-- 向系统表中添加当前有哪些分库
INSERT INTO test.system_setting
VALUES(NULL, 'sharding_database', 'test_1');
-- 修改系统表字段类value型为varchar(120)
ALTER TABLE test.system_setting
MODIFY `value` varchar(120) NOT NULL COMMENT '系统设置值';
-- 向系统表添加响应数据库链接描述符
INSERT INTO test.system_setting
VALUES(NULL, 'test_1', '{"user":"root","password":"root","host":"127.0.0.1","port":3306,"database":"test_1"}');
-- 初始化用户所在库为test_1
ALTER TABLE user
ADD db_name VARCHAR(45) NOT NULL DEFAULT 'test_1'
COMMENT '用户数据所在数据库名';
文章源自运维生存时间-https://www.ttlsa.com/mysql/mysql-distributed-database-and-table-prepare/文章源自运维生存时间-https://www.ttlsa.com/mysql/mysql-distributed-database-and-table-prepare/ 
评论