springboot集成starrocks、以及采用flink实现mysql与starrocks亚秒级同步

一、application.yml

（因采用dynamic-datasource-spring-boot-starter动态数据源，所以才是以下配置文件的样式，像redis，druid根据自己情况导入依赖）

这个配置文件的场景是把starrocks当成slave库在用。某些大数据慢查询就走starrocks

就这样配置好后就可把starrocks当mysql用了

# spring配置
spring:
  redis:
    host: localhost
    port: 6379
    password:
  datasource:
    druid:
      stat-view-servlet:
        enabled: true
        loginUsername: admin
        loginPassword: 123456
    dynamic:
      druid:
        initial-size: 5
        min-idle: 5
        maxActive: 20
        maxWait: 60000
        timeBetweenEvictionRunsMillis: 60000
        minEvictableIdleTimeMillis: 300000
        validationQuery: SELECT 1
        testWhileIdle: true
        testOnBorrow: false
        testOnReturn: false
        poolPreparedStatements: true
        maxPoolPreparedStatementPerConnectionSize: 20
        filters: stat,slf4j
        connectionProperties: druid.stat.mergeSql\=true;druid.stat.slowSqlMillis\=5000
      datasource:
        # 主库数据源
        master:
          driver-class-name: com.mysql.cj.jdbc.Driver
          url: jdbc:mysql://xx.xx.xx.xx:3306/inst_ops?useUnicode=true&characterEncoding=utf8&zeroDateTimeBehavior=convertToNull&useSSL=true&serverTimezone=GMT%2B8&rewriteBatchedStatements=true
          username: root
          password: 123456
          # 从库数据源
        slave:
          driver-class-name: com.mysql.cj.jdbc.Driver
          username: root
          password: 
          url: jdbc:mysql://xx.xx.xx.xx:9030/inst_ops?useUnicode=true&characterEncoding=utf8&zeroDateTimeBehavior=convertToNull&useSSL=false&serverTimezone=GMT%2B8&rewriteBatchedStatements=true&autoReconnect=true&failOverReadOnly=false
      seata: true

# mybatis配置
# mybatis:
#     # 搜索指定包别名
#     typeAliasesPackage: com.ruoyi.stream
#     # 配置mapper的扫描，找到所有的mapper.xml映射文件
#     mapperLocations: classpath:mapper/**/*.xml

# swagger配置
swagger:
  title: mongodb接口文档
  license: Powered By ruoyi
  licenseUrl: http://localhost:8080/doc.html

二、实验场景：查询mysql数据，使用insert into批量往starrocks插入数据

重点：采用这种方式有限制，插入几千条starrocks就不让你插入了，只用于demo测试

package com.ruoyi.starrocks.test;

import com.baomidou.dynamic.datasource.toolkit.DynamicDataSourceContextHolder;
import com.baomidou.mybatisplus.core.conditions.query.LambdaQueryWrapper;
import com.baomidou.mybatisplus.core.toolkit.Wrappers;
import com.ruoyi.starrocks.domain.DevMc;
import com.ruoyi.starrocks.mapper.DevMcMapper;
import com.ruoyi.starrocks.service.IDevMcService;
import lombok.extern.slf4j.Slf4j;
import org.junit.jupiter.api.Test;
import org.springframework.boot.test.context.SpringBootTest;

import javax.annotation.Resource;
import java.util.List;

/**
 *
 * @author yh
 * @date 2023-03-24 10:50
 */
@SpringBootTest
@Slf4j
public class StarrocksTest {

    public static final String NAME = "INSERT INTO `inst_ops`.dev_mc (id,sign_id,`year`,oid,ono,prov,city,area,park,tmpl," +
            "dtm,over_dtm,over_state,mc_tag,mc_cid,mc_uid,mc_man,mc_day,mc_sta,mc_end,audt_cid,audt_tag_chk,audt_tag_ok," +
            "audt_dtm,audt_uid,audt_man,audt_inf,submit,version,c_id,to_upd) VALUES " ;

    @Resource
    private IDevMcService devMcService;
    @Resource
    private DevMcMapper mapper;

    @Test
    public void test1() {
        LambdaQueryWrapper<DevMc> last = Wrappers.lambdaQuery(DevMc.class);
        List<DevMc> list = devMcService.list(last);
        System.err.println("查询完毕");
        DynamicDataSourceContextHolder.push("slave"); // 手动切换
        System.err.println("切换数据源成功---------------");
        StringBuilder sb = new StringBuilder();
        int count = 0;
        for (DevMc devMc : list) {
            count ++;
            sb.append("(");
            sb.append(devMc.getId()).append(",");
            sb.append(devMc.getSignId()).append(",");
            sb.append(devMc.getYear()).append(",");
            sb.append(devMc.getOid()).append(",");
            sb.append("'").append(devMc.getOno()).append("',");
            sb.append(devMc.getProv()).append(",");
            sb.append(devMc.getCity()).append(",");
            sb.append(devMc.getArea()).append(",");
            sb.append(devMc.getPark()).append(",");
            sb.append(devMc.getTmpl()).append(",");
            sb.append("'").append(devMc.getDtm()).append("',");
            sb.append("'").append(devMc.getOverDtm()).append("',");
            sb.append(devMc.getOverState()).append(",");
            sb.append(devMc.getMcTag()).append(",");
            sb.append(devMc.getMcCid()).append(",");
            sb.append("'").append(devMc.getMcUid()).append("',");
            sb.append("'").append(devMc.getMcMan()).append("',");
            sb.append("'").append(devMc.getMcDay()).append("',");
            sb.append("'").append(devMc.getMcSta()).append("',");
            sb.append("'").append(devMc.getMcEnd()).append("',");
            sb.append(devMc.getAudtCid()).append(",");
            sb.append(devMc.getAudtTagChk()).append(",");
            sb.append(devMc.getAudtTagOk()).append(",");
            sb.append("'").append(devMc.getAudtDtm()).append("',");
            sb.append(devMc.getAudtUid()).append(",");
            sb.append("'").append(devMc.getAudtMan()).append("',");
            sb.append("'").append(devMc.getAudtInf()).append("',");
            sb.append(devMc.getSubmit()).append(",");
            sb.append(devMc.getVersion()).append(",");
            sb.append(devMc.getCId()).append(",");
            sb.append(devMc.getToUpd());
            sb.append(")");
            if (count == 5000) {
                sb.append(";");
                StringBuilder insert = new StringBuilder(NAME);
                insert.append(sb.toString());
                mapper.insertBatch(insert.toString());
                count = 0;
                sb.setLength(0);
            } else {
                sb.append(",");
            }
        }

        if (sb.length() > 0) {
            StringBuilder insert = new StringBuilder(NAME);
            insert.append(sb.toString().substring(0, sb.length()-1));
            insert.append(";");
            mapper.insertBatch(insert.toString());
            count = 0;
            sb.setLength(0);
        }

    }

    @Test
    public void test2() {
        DevMc devMc = devMcService.getById(1558);
        System.err.println("查询完毕");
        DynamicDataSourceContextHolder.push("slave"); // 手动切换
        System.err.println("切换数据源成功---------------");
        devMcService.save(devMc);
    }
}

三、使用flink往starrocks导入数据，并实现亚秒级同步

第3.4章：StarRocks数据导入--Flink Connector与CDC秒级数据同步_flink-connector-starrocks_流木随风的博客-CSDN博客

所需文件：注意！有些jar包在外网没得，只有阿里云仓库有。而且搜索的出来的jar包与博客展示的名称有细微差别。按博客的名称搜根本搜不出来，所以我觉得是他文档好久没更新了

flink版本：

flink-1.13.5-bin-scala_2.11.tgz、

需要jar包：

flink-connector-starrocks-1.1.14_flink-1.13_2.11.jar

flink-sql-connector-mysql-cdc-2.0.2.jar

flink-connector-jdbc_2.11-1.13.5.jar

jar包阿里网盘下载连接：阿里云盘分享

步骤就不用说了，按那个博客进行操作

四、相关建表语句(私人保留记录用，跟文章无关)

1、starrocks建库语句

create database inst_ops;

2、starrocks建表语句

CREATE TABLE `dev_mc`  (
  `id` bigint NOT NULL COMMENT '维保流水号',
  `sign_id` bigint NOT NULL  COMMENT '签到ID',
  `year` int NOT NULL  COMMENT '维保年份',
  `oid` int NOT NULL  COMMENT '设备编号',
  `ono` varchar(90)  NOT NULL DEFAULT '' COMMENT '设备号',
  `prov` int NOT NULL  COMMENT '省份',
  `city` int NOT NULL  COMMENT '城市',
  `area` int NOT NULL,
  `park` int NOT NULL,
  `tmpl` int NOT NULL COMMENT '模板编号',
  `dtm` varchar(57)  NOT NULL DEFAULT '' COMMENT '时间-维保最迟时间',
  `over_dtm` varchar(57)  NOT NULL DEFAULT '' COMMENT '维保超期时间',
  `over_state` tinyint(1) NOT NULL  COMMENT '超期是否计算',
  `mc_tag` tinyint(1) NOT NULL  COMMENT '是否维保',
  `mc_cid` int NOT NULL  COMMENT '维保单位',
  `mc_uid` varchar(90)  NOT NULL DEFAULT '0' COMMENT '维保人编号',
  `mc_man` varchar(90)  NOT NULL DEFAULT '' COMMENT '维保人姓名',
  `mc_day` varchar(33)  NOT NULL DEFAULT '' COMMENT '保养日期(xxxx年xx月xx日)',
  `mc_sta` varchar(57)  NOT NULL DEFAULT '' COMMENT '维保开始时间(xx时xx分)',
  `mc_end` varchar(57)  NOT NULL DEFAULT '' COMMENT '维保结束时间(xx时xx分)',
  `audt_cid` int NOT NULL  COMMENT '审核公司ID',
  `audt_tag_chk` tinyint(1) NOT NULL  COMMENT '使用单位是否审核',
  `audt_tag_ok` tinyint(1) NOT NULL  COMMENT '使用单位审核结果：通过/不通过',
  `audt_dtm` varchar(57)  NOT NULL DEFAULT '' COMMENT '审核日期',
  `audt_uid` int NOT NULL  COMMENT '使用单位审核人',
  `audt_man` varchar(90)  NOT NULL  COMMENT '使用单位审核人',
  `audt_inf` varchar(150)  NOT NULL  COMMENT '使用单位审核结果',
  `submit` tinyint(1) NOT NULL  COMMENT '是否提交物业审核',
  `version` int NOT NULL  COMMENT '版本号(1判断审核后是否更新)',
  `c_id` bigint NOT NULL  COMMENT '96333返回ID',
  `to_upd` tinyint NULL COMMENT '是否要同步(0否1是)'
) PRIMARY KEY (id)
DISTRIBUTED BY HASH(id) BUCKETS 4
PROPERTIES("replication_num" = "1",
"enable_persistent_index" = "true");

3、flink sql——mysql、starrocks的建表映射语句

mysql：

CREATE TABLE source_mysql_devmc (
  id BIGINT,
  sign_id BIGINT,
  `year` int ,
  `oid` int ,
  `ono` STRING  ,
  `prov` int ,
  `city` int,
  `area` int,
  `park` int ,
  `tmpl` int ,
  `dtm` STRING  ,
  `over_dtm` STRING  ,
  `over_state` BOOLEAN ,
  `mc_tag` BOOLEAN ,
  `mc_cid` int ,
  `mc_uid` STRING  ,
  `mc_man` STRING ,
  `mc_day` STRING ,
  `mc_sta` STRING  ,
  `mc_end` STRING ,
  `audt_cid` int ,
  `audt_tag_chk` BOOLEAN ,
  `audt_tag_ok` BOOLEAN ,
  `audt_dtm` STRING  ,
  `audt_uid` int ,
  `audt_man` STRING  ,
  `audt_inf` STRING , 
  `submit` BOOLEAN ,
  `version` int ,
  `c_id` bigint ,
  `to_upd` tinyint ,
   PRIMARY KEY (id) NOT ENFORCED
)WITH (
   'connector' = 'jdbc',
   'url' = 'jdbc:mysql://localhost:3306/inst_ops',
   'table-name' = 'dev_mc',
   'username' = 'root',
   'password' = '123456'
);

starrocks：

CREATE TABLE sink_starrocks_devmc (
  id BIGINT,
  sign_id BIGINT,
  `year` int ,
  `oid` int ,
  `ono` STRING  ,
  `prov` int ,
  `city` int,
  `area` int,
  `park` int ,
  `tmpl` int ,
  `dtm` STRING  ,
  `over_dtm` STRING  ,
  `over_state` BOOLEAN ,
  `mc_tag` BOOLEAN ,
  `mc_cid` int ,
  `mc_uid` STRING  ,
  `mc_man` STRING ,
  `mc_day` STRING ,
  `mc_sta` STRING  ,
  `mc_end` STRING ,
  `audt_cid` int ,
  `audt_tag_chk` BOOLEAN ,
  `audt_tag_ok` BOOLEAN ,
  `audt_dtm` STRING  ,
  `audt_uid` int ,
  `audt_man` STRING  ,
  `audt_inf` STRING , 
  `submit` BOOLEAN ,
  `version` int ,
  `c_id` bigint ,
  `to_upd` tinyint ,
   PRIMARY KEY (id) NOT ENFORCED
)WITH (
   'connector' = 'starrocks',
   'jdbc-url'='jdbc:mysql://localhost:9030',
   'load-url'='localhost:8030',
   'database-name' = 'inst_ops',
   'table-name' = 'dev_mc',
   'username' = 'root',
   'password' = '',
   'sink.buffer-flush.interval-ms' = '5000',
   'sink.properties.column_separator' = '\x01',
   'sink.properties.row_delimiter' = '\x02'
);

需要注意的点：因为我mysql与starrocks都是用docker启动的，所以ip是localhost。用公网ip就会连不上数据库，这是因为docker自身的问题，他们说用红帽版的docker就没有这个问题

4、同步数据语句

insert into sink_starrocks_devmc select * from source_mysql_devmc limit 300000;

5、亚秒级同步（只需要修改mysql映射）

CREATE TABLE source_mysql_devmc (
  id BIGINT,
  sign_id BIGINT,
  `year` int ,
  `oid` int ,
  `ono` STRING  ,
  `prov` int ,
  `city` int,
  `area` int,
  `park` int ,
  `tmpl` int ,
  `dtm` STRING  ,
  `over_dtm` STRING  ,
  `over_state` BOOLEAN ,
  `mc_tag` BOOLEAN ,
  `mc_cid` int ,
  `mc_uid` STRING  ,
  `mc_man` STRING ,
  `mc_day` STRING ,
  `mc_sta` STRING  ,
  `mc_end` STRING ,
  `audt_cid` int ,
  `audt_tag_chk` BOOLEAN ,
  `audt_tag_ok` BOOLEAN ,
  `audt_dtm` STRING  ,
  `audt_uid` int ,
  `audt_man` STRING  ,
  `audt_inf` STRING , 
  `submit` BOOLEAN ,
  `version` int ,
  `c_id` bigint ,
  `to_upd` tinyint ,
   PRIMARY KEY (id) NOT ENFORCED
)WITH (
   'connector' = 'mysql-cdc',
   'hostname' = 'localhost',
   'port' = '3306',
   'username' = 'root',
   'password' = '123456',
   'database-name' = 'inst_ops',
   'scan.incremental.snapshot.enabled'='false',
   'table-name' = 'dev_mc'
);