sharding-jdbc水平分库分表实战

• 资料
  分库分表的概念及应用场景详解
  分库分表带来的一些问题
  sharding-jdbc水平垂直分库分表环境搭建
  sharding-jdbc水平分库分表实战
  sharding-jdbc垂直分库分表实战

• 水平分库分表目的
  1.水平分表,缓解单表数据量大的性能问题.
  2.水平分库,缓解单库高并发请求的性能问题

• 水平分表实现

  实现方案:将一张表的数据,根据某种算法,分摊到多表中
  public class springJdbcTest {
        
        
      public static void main(String[] args) throws SQLException {
        
        
  
          //获取数据源
          DataSource dataSource = getDataSource();
  
          //获取jdbctemplate
          JdbcTemplate jdbcTemplate = getJdbcTemplate(dataSource);
  
  
          //执行语句
          for(int i=1;i<=10;i++) {
        
        
              jdbcTemplate.update("insert into t_user(user_name,user_age,user_type) values (?,?,?)",i,i,i);
          }
  
      }
  
  	//创建JdbcTemplate
      public static JdbcTemplate getJdbcTemplate( DataSource dataSource){
        
        
          JdbcTemplate jdbcTemplate = new JdbcTemplate();
          jdbcTemplate.setDataSource(dataSource);
          return jdbcTemplate;
      }
  
  	//创建数据源,这个数据源是sharding-jdbc级别的数据源
      public static DataSource getDataSource() throws SQLException {
        
        
          // 配置真实数据源
          Map<String, DataSource> dataSourceMap = new HashMap<>();
  
          // 配置第 1 个数据源
          DruidDataSource dataSource1 = new DruidDataSource();
          dataSource1.setDriverClassName("com.mysql.jdbc.Driver");
          dataSource1.setUrl("jdbc:mysql://localhost:3306/ds0?characterEncoding=utf-8&useSSL=false");
          dataSource1.setUsername("root");
          dataSource1.setPassword("123456");
          dataSourceMap.put("ds0", dataSource1);
  
          // 配置表规则bean
          ShardingRuleConfiguration shardingRuleConfiguration=new ShardingRuleConfiguration();
  
          //配置逻辑表和实际表分布情况,$是一个变量,取值范围是[1,2],表都在一个数据源ds0
          TableRuleConfiguration tableRuleConfiguration = new TableRuleConfiguration("t_user","ds0.t_user_$->{1..2}");
  
          //实际表生成主键策略,分表用数据库自增会重复,必须用第三方生成的id
          KeyGeneratorConfiguration keyGeneratorConfiguration = new KeyGeneratorConfiguration("snowflake","user_id");
  
          //配置选择表的路由配置,当user_id是偶数数据存t_user_1中
          //配置选择表的路由配置,当user_id是奇数数据存t_user_2中
          ShardingStrategyConfiguration shardingStrategyConfiguration = new InlineShardingStrategyConfiguration("user_id","t_user_$->{user_id%2+1}");
          //配置选择库的路由配置,因为是分表,所以库是固定的
          ShardingStrategyConfiguration shardingStrategyConfiguration1 = new InlineShardingStrategyConfiguration("user_id","ds0");
  
          shardingRuleConfiguration.setTableRuleConfigs(Arrays.asList(tableRuleConfiguration));
          shardingRuleConfiguration.setDefaultKeyGeneratorConfig(keyGeneratorConfiguration);
          shardingRuleConfiguration.setDefaultTableShardingStrategyConfig(shardingStrategyConfiguration);
          shardingRuleConfiguration.setDefaultDatabaseShardingStrategyConfig(shardingStrategyConfiguration1);
  
          return ShardingDataSourceFactory.createDataSource(dataSourceMap, shardingRuleConfiguration, new Properties());
      }
  }
  

• 水平分表效果
  
  

• 水平分库实现

  public class springJdbcTest {
        
        
      public static void main(String[] args) throws SQLException {
        
        
  
          //获取数据源
          DataSource dataSource = getDataSource();
  
          //获取jdbctemplate
          JdbcTemplate jdbcTemplate = getJdbcTemplate(dataSource);
  
  
          //执行语句
          for(int i=1;i<=10;i++) {
        
        
              jdbcTemplate.update("insert into t_user(user_name,user_age,user_type) values (?,?,?)",i,i,i);
          }
  
      }
  
  
      public static JdbcTemplate getJdbcTemplate( DataSource dataSource){
        
        
          JdbcTemplate jdbcTemplate = new JdbcTemplate();
          jdbcTemplate.setDataSource(dataSource);
          return jdbcTemplate;
      }
  
      public static DataSource getDataSource() throws SQLException {
        
        
          // 配置真实数据源
          Map<String, DataSource> dataSourceMap = new HashMap<>();
  
          // 配置第 1 个数据源
          DruidDataSource dataSource1 = new DruidDataSource();
          dataSource1.setDriverClassName("com.mysql.jdbc.Driver");
          dataSource1.setUrl("jdbc:mysql://localhost:3306/ds0");
          dataSource1.setUsername("root");
          dataSource1.setPassword("123456");
          dataSourceMap.put("ds0", dataSource1);
  
          // 配置第 2 个数据源
          DruidDataSource dataSource2 = new DruidDataSource();
          dataSource2.setDriverClassName("com.mysql.jdbc.Driver");
          dataSource2.setUrl("jdbc:mysql://localhost:3307/ds0");
          dataSource2.setUsername("root");
          dataSource2.setPassword("123456");
          dataSourceMap.put("ds1", dataSource2);
  
          // 配置表规则bean
          ShardingRuleConfiguration shardingRuleConfiguration=new ShardingRuleConfiguration();
  
          //配置逻辑表和实际表分布情况,两个t_user表分布在不同的数据库中
          TableRuleConfiguration tableRuleConfiguration = new TableRuleConfiguration("t_user","ds$->{0..1}.t_user");
  
          //实际表生成主键策略,分表用数据库自增会重复,必须用第三方生成的id
          KeyGeneratorConfiguration keyGeneratorConfiguration = new KeyGeneratorConfiguration("snowflake","user_id");
  
          //配置选择表的路由配置,这里我们分库了,表名称一样了
          ShardingStrategyConfiguration shardingStrategyConfiguration = new InlineShardingStrategyConfiguration("user_id","t_user");
  
  		//配置选择库的路由配置,当user_id是奇数,数据存在ds0数据库中
  		//配置选择库的路由配置,当user_id是偶数,数据存在ds1数据库中
          ShardingStrategyConfiguration shardingStrategyConfiguration1 = new InlineShardingStrategyConfiguration("user_id","ds$->{user_id%2}");
  
          shardingRuleConfiguration.setTableRuleConfigs(Arrays.asList(tableRuleConfiguration));
          shardingRuleConfiguration.setDefaultKeyGeneratorConfig(keyGeneratorConfiguration);
          shardingRuleConfiguration.setDefaultTableShardingStrategyConfig(shardingStrategyConfiguration);
          shardingRuleConfiguration.setDefaultDatabaseShardingStrategyConfig(shardingStrategyConfiguration1);
  
          return ShardingDataSourceFactory.createDataSource(dataSourceMap, shardingRuleConfiguration, new Properties());
      }
  }
  
  

• 水平分库效果