《微服务实战》 第三十一章 ShardingSphere - ShardingSphere-JDBC

前言

Apache ShardingSphere 是一款分布式的数据库生态系统， 可以将任意数据库转换为分布式数据库，并通过数据分片、弹性伸缩、加密等能力对原有数据库进行增强。

Apache ShardingSphere 设计哲学为 Database Plus，旨在构建异构数据库上层的标准和生态。 它关注如何充分合理地利用数据库的计算和存储能力，而并非实现一个全新的数据库。 它站在数据库的上层视角，关注它们之间的协作多于数据库自身。

1、ShardingSphere-JDBC

ShardingSphere-JDBC 定位为轻量级 Java 框架，在 Java 的 JDBC 层提供的额外服务。

1.1、应用场景

Apache ShardingSphere-JDBC 可以通过Java 和 YAML 这 2 种方式进行配置，开发者可根据场景选择适合的配置方式。

• 数据库读写分离
• 数据库分表分库

1.2、原理

• Sharding-JDBC中的路由结果是通过分片字段和分片方法来确定的,如果查询条件中有 id 字段的情况还好，查询将会落到某个具体的分片
• 如果查询没有分片的字段，会向所有的db或者是表都会查询一遍，让后封装结果集给客户端。
  

1.3、spring boot整合

1.3.1、添加依赖

<!-- 分表分库依赖 -->
<dependency><groupId>org.apache.shardingsphere</groupId><artifactId>sharding-jdbc-spring-boot-starter</artifactId><version>4.1.1</version>
</dependency>

1.3.2、添加配置

spring:main:# 一个实体类对应多张表，覆盖allow-bean-definition-overriding: trueshardingsphere:datasource:ds0:#配置数据源具体内容，包含连接池，驱动，地址，用户名和密码driver-class-name: com.mysql.cj.jdbc.Driverjdbc-url: jdbc:mysql://127.0.0.1:3306/account?autoReconnect=true&allowMultiQueries=truepassword: roottype: com.zaxxer.hikari.HikariDataSourceusername: rootds1:driver-class-name: com.mysql.cj.jdbc.Driverjdbc-url: jdbc:mysql://127.0.0.1:3306/account?autoReconnect=true&allowMultiQueries=truepassword: roottype: com.zaxxer.hikari.HikariDataSourceusername: root# 配置数据源，给数据源起名称names: ds0,ds1props:sql:show: truesharding:tables:user_info:#指定 user_info 表分布情况，配置表在哪个数据库里面，表名称都是什么actual-data-nodes: ds0.user_info_${0..9}database-strategy:standard:preciseAlgorithmClassName: com.xxxx.store.account.config.PreciseDBShardingAlgorithmrangeAlgorithmClassName: com.xxxx.store.account.config.RangeDBShardingAlgorithmsharding-column: idtable-strategy:standard:preciseAlgorithmClassName: com.xxxx.store.account.config.PreciseTablesShardingAlgorithmrangeAlgorithmClassName: com.xxxx.store.account.config.RangeTablesShardingAlgorithmsharding-column: id

1.3.3、制定分片算法

1.3.3.1、精确分库算法

/*** 精确分库算法*/
public class PreciseDBShardingAlgorithm implements PreciseShardingAlgorithm<Long> {/**** @param availableTargetNames 配置所有的列表* @param preciseShardingValue 分片值* @return*/@Overridepublic String doSharding(Collection<String> availableTargetNames, PreciseShardingValue<Long> preciseShardingValue) {Long value = preciseShardingValue.getValue();//后缀 0，1String postfix = String.valueOf(value % 2);for (String availableTargetName : availableTargetNames) {if(availableTargetName.endsWith(postfix)){return availableTargetName;}}throw new UnsupportedOperationException();}}

1.3.3.2、范围分库算法

/*** 范围分库算法*/
public class RangeDBShardingAlgorithm implements RangeShardingAlgorithm<Long> {@Overridepublic Collection<String> doSharding(Collection<String> collection, RangeShardingValue<Long> rangeShardingValue) {return collection;}
}

1.3.3.3、精确分表算法

/*** 精确分表算法*/
public class PreciseTablesShardingAlgorithm implements PreciseShardingAlgorithm<Long> {/**** @param availableTargetNames 配置所有的列表* @param preciseShardingValue 分片值* @return*/@Overridepublic String doSharding(Collection<String> availableTargetNames, PreciseShardingValue<Long> preciseShardingValue) {Long value = preciseShardingValue.getValue();//后缀String postfix = String.valueOf(value % 10);for (String availableTargetName : availableTargetNames) {if(availableTargetName.endsWith(postfix)){return availableTargetName;}}throw new UnsupportedOperationException();}}

1.3.3.4、范围分表算法

/*** 范围分表算法*/
public class RangeTablesShardingAlgorithm implements RangeShardingAlgorithm<Long> {@Overridepublic Collection<String> doSharding(Collection<String> collection, RangeShardingValue<Long> rangeShardingValue) {Collection<String> result = new ArrayList<>();Range<Long> valueRange = rangeShardingValue.getValueRange();Long start = valueRange.lowerEndpoint();Long end = valueRange.upperEndpoint();Long min = start % 10;Long max = end % 10;for (Long i = min; i < max +1; i++) {Long finalI = i;collection.forEach(e -> {if(e.endsWith(String.valueOf(finalI))){result.add(e);}});}return result;}}

1.3.4、数据库建表

DROP TABLE IF EXISTS `user_info_0`;
CREATE TABLE `user_info_0` (`id` bigint(20) NOT NULL,`account` varchar(255) DEFAULT NULL,`user_name` varchar(255) DEFAULT NULL,`pwd` varchar(255) DEFAULT NULL,PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8;

1.3.5、业务应用

1.3.5.1、定义实体类

@Data
@TableName(value = "user_info")
public class UserInfo {/*** 主键*/private Long id;/*** 账号*/private String account;/*** 用户名*/private String userName;/*** 密码*/private String pwd;}

1.3.5.2、定义接口

public interface UserInfoService{/*** 保存* @param userInfo* @return*/public UserInfo saveUserInfo(UserInfo userInfo);public UserInfo getUserInfoById(Long id);public List<UserInfo> listUserInfo();
}

1.3.5.3、实现类

@Service
public class UserInfoServiceImpl extends ServiceImpl<UserInfoMapper, UserInfo> implements UserInfoService {@Override@Transactionalpublic UserInfo saveUserInfo(UserInfo userInfo) {userInfo.setId(IdUtils.getId());this.save(userInfo);return userInfo;}@Overridepublic UserInfo getUserInfoById(Long id) {return this.getById(id);}@Overridepublic List<UserInfo> listUserInfo() {QueryWrapper<UserInfo> userInfoQueryWrapper = new QueryWrapper<>();userInfoQueryWrapper.between("id",1623695688380448768L,1623695688380448769L);return this.list(userInfoQueryWrapper);}
}

1.3.6、生成ID - 雪花算法

package com.xxxx.tore.common.utils;import cn.hutool.core.lang.Snowflake;
import cn.hutool.core.util.IdUtil;/*** 生成各种组件ID*/
public class IdUtils {/*** 雪花算法* @return*/public static long getId(){Snowflake snowflake = IdUtil.getSnowflake(0, 0);long id = snowflake.nextId();return id;}
}

1.4、seata与sharding-jdbc整合

https://github.com/seata/seata-samples/tree/master/springcloud-seata-sharding-jdbc-mybatis-plus-samples

1.4.1、common中添加依赖

<!--seata依赖-->
<dependency><groupId>com.alibaba.cloud</groupId><artifactId>spring-cloud-starter-alibaba-seata</artifactId><version>2021.0.4.0</version>
</dependency>
<!-- sharding-jdbc整合seata分布式事务-->
<dependency><groupId>org.apache.shardingsphere</groupId><artifactId>sharding-transaction-base-seata-at</artifactId><version>4.1.1</version>
</dependency><dependency><groupId>com.alibaba.cloud</groupId><artifactId>spring-cloud-starter-alibaba-nacos-discovery</artifactId><version>2021.0.4.0</version><exclusions><exclusion><groupId>com.alibaba.nacos</groupId><artifactId>nacos-client</artifactId></exclusion></exclusions>
</dependency>
<dependency><groupId>com.alibaba.nacos</groupId><artifactId>nacos-client</artifactId><version>1.4.2</version>
</dependency>

1.4.2、改造account-service服务

@Service
public class AccountServiceImpl extends ServiceImpl<AccountMapper, Account> implements AccountService {@Autowiredprivate OrderService orderService;@Autowiredprivate StorageService storageService;/*** 存放商品编码及其对应的价钱*/private static Map<String,Integer> map = new HashMap<>();static {map.put("c001",3);map.put("c002",5);map.put("c003",10);map.put("c004",6);}@Override@Transactional@ShardingTransactionType(TransactionType.BASE)public void debit(OrderDTO orderDTO) {//扣减账户余额int calculate = this.calculate(orderDTO.getCommodityCode(), orderDTO.getCount());AccountDTO accountDTO = new AccountDTO(orderDTO.getUserId(), calculate);QueryWrapper<Account> objectQueryWrapper = new QueryWrapper<>();objectQueryWrapper.eq("id",1);objectQueryWrapper.eq(accountDTO.getUserId() != null,"user_id",accountDTO.getUserId());Account account = this.getOne(objectQueryWrapper);account.setMoney(account.getMoney() - accountDTO.getMoney());this.saveOrUpdate(account);//扣减库存this.storageService.deduct(new StorageDTO(null,orderDTO.getCommodityCode(),orderDTO.getCount()));//生成订单this.orderService.create(orderDTO);      }/*** 计算购买商品的总价钱* @param commodityCode* @param orderCount* @return*/private int calculate(String commodityCode, int orderCount){//商品价钱Integer price = map.get(commodityCode) == null ? 0 : map.get(commodityCode);return price * orderCount;}
}

注意：调单生成调用的逻辑修改，减余额->减库存->生成订单。调用入口方法注解加上：@ShardingTransactionType(TransactionType.BASE)

1.4.3、修改business-service服务

@Service
public class BusinessServiceImpl implements BusinessService {@Autowiredprivate OrderService orderService;@Autowiredprivate StorageService storageService;@Autowiredprivate AccountService accountService;@Overridepublic void purchase(OrderDTO orderDTO) {//扣减账号中的钱accountService.debit(orderDTO);        }
}

1.4.4、修改order-service服务

@Service
public class OrderServiceImpl extends ServiceImpl<OrderMapper,Order> implements OrderService {/*** 存放商品编码及其对应的价钱*/private static Map<String,Integer> map = new HashMap<>();static {map.put("c001",3);map.put("c002",5);map.put("c003",10);map.put("c004",6);}@Override@Transactional@ShardingTransactionType(TransactionType.BASE)public Order create(String userId, String commodityCode, int orderCount) {int orderMoney = calculate(commodityCode, orderCount);Order order = new Order();order.setUserId(userId);order.setCommodityCode(commodityCode);order.setCount(orderCount);order.setMoney(orderMoney);//保存订单this.save(order);try {TimeUnit.SECONDS.sleep(30);} catch (InterruptedException e) {e.printStackTrace();}if(true){throw new RuntimeException("回滚测试");}return order;}/*** 计算购买商品的总价钱* @param commodityCode* @param orderCount* @return*/private int calculate(String commodityCode, int orderCount){//商品价钱Integer price = map.get(commodityCode) == null ? 0 : map.get(commodityCode);return price * orderCount;}
}

1.4.5、配置文件参考

server:port: 8090spring:main:# 一个实体类对应多张表，覆盖allow-bean-definition-overriding: trueshardingsphere:datasource:ds0:#配置数据源具体内容，包含连接池，驱动，地址，用户名和密码driver-class-name: com.mysql.cj.jdbc.Driverjdbc-url: jdbc:mysql://127.0.0.1:3306/account?autoReconnect=true&allowMultiQueries=truepassword: roottype: com.zaxxer.hikari.HikariDataSourceusername: rootds1:driver-class-name: com.mysql.cj.jdbc.Driverjdbc-url: jdbc:mysql://127.0.0.1:3306/account?autoReconnect=true&allowMultiQueries=truepassword: roottype: com.zaxxer.hikari.HikariDataSourceusername: root# 配置数据源，给数据源起名称names: ds0,ds1props:sql:show: truesharding:tables:account_tbl:actual-data-nodes: ds0.account_tbl_${0..1}database-strategy:standard:preciseAlgorithmClassName: com.xxxx.store.account.config.PreciseDBExtShardingAlgorithm#rangeAlgorithmClassName: com.xxxx.store.account.config.RangeDBShardingAlgorithmsharding-column: idtable-strategy:standard:preciseAlgorithmClassName: com.xxxx.store.account.config.PreciseTablesExtShardingAlgorithm#rangeAlgorithmClassName: com.xxxx.store.account.config.RangeTablesShardingAlgorithmsharding-column: iduser_info:#指定 user_info 表分布情况，配置表在哪个数据库里面，表名称都是什么actual-data-nodes: ds0.user_info_${0..9}database-strategy:standard:preciseAlgorithmClassName: com.xxxx.store.account.config.PreciseDBShardingAlgorithmrangeAlgorithmClassName: com.xxxx.store.account.config.RangeDBShardingAlgorithmsharding-column: idtable-strategy:standard:preciseAlgorithmClassName: com.xxxx.store.account.config.PreciseTablesShardingAlgorithmrangeAlgorithmClassName: com.xxxx.store.account.config.RangeTablesShardingAlgorithmsharding-column: id#以上是sharding-jdbc配置cloud:nacos:discovery:server-addr: localhost:8848namespace: 1ff3782d-b62d-402f-8bc4-ebcf40254d0aapplication:name: account-service  #微服务名称
#  datasource:
#    username: root
#    password: root
#    url: jdbc:mysql://127.0.0.1:3306/account
#    driver-class-name: com.mysql.cj.jdbc.Driverseata:enabled: trueenable-auto-data-source-proxy: falseapplication-id: account-servicetx-service-group: default_tx_groupservice:vgroup-mapping:default_tx_group: defaultdisable-global-transaction: falseregistry:type: nacosnacos:application: seata-serverserver-addr: 127.0.0.1:8848namespace: 1ff3782d-b62d-402f-8bc4-ebcf40254d0agroup: SEATA_GROUPusername: nacospassword: nacosconfig:nacos:server-addr: 127.0.0.1:8848namespace: 1ff3782d-b62d-402f-8bc4-ebcf40254d0agroup: SEATA_GROUPusername: nacospassword: nacos