系列文章目录

物流实时数仓：采集通道搭建
物流实时数仓：数仓搭建

---

---

前言

现在我们开始进行数仓的搭建，我们用Kafka来代替数仓的ods层。
基本流程为使用Flink从MySQL读取数据然后写入Kafka中

---

一、IDEA环境准备

1.pom.xml

写入项目需要的配置

<properties><maven.compiler.source>8</maven.compiler.source><maven.compiler.target>8</maven.compiler.target><project.build.sourceEncoding>UTF-8</project.build.sourceEncoding><java.version>1.8</java.version><flink.version>1.17.0</flink.version><hadoop.version>3.2.3</hadoop.version><flink-cdc.version>2.3.0</flink-cdc.version></properties><dependencies><dependency><groupId>org.apache.flink</groupId><artifactId>flink-java</artifactId><version>${flink.version}</version></dependency><dependency><groupId>org.apache.flink</groupId><artifactId>flink-streaming-java</artifactId><version>${flink.version}</version></dependency><dependency><groupId>org.apache.flink</groupId><artifactId>flink-connector-kafka</artifactId><version>${flink.version}</version></dependency><dependency><groupId>com.alibaba</groupId><artifactId>fastjson</artifactId><version>1.2.68</version></dependency><dependency><groupId>org.apache.hadoop</groupId><artifactId>hadoop-client</artifactId><version>${hadoop.version}</version><exclusions><exclusion><groupId>org.slf4j</groupId><artifactId>slf4j-reload4j</artifactId></exclusion></exclusions></dependency><dependency><groupId>org.apache.flink</groupId><artifactId>flink-clients</artifactId><version>${flink.version}</version></dependency><dependency><groupId>org.slf4j</groupId><artifactId>slf4j-api</artifactId><version>1.7.25</version></dependency><dependency><groupId>org.slf4j</groupId><artifactId>slf4j-log4j12</artifactId><version>1.7.25</version></dependency><dependency><groupId>org.apache.logging.log4j</groupId><artifactId>log4j-to-slf4j</artifactId><version>2.14.0</version></dependency><dependency><groupId>com.ververica</groupId><artifactId>flink-connector-mysql-cdc</artifactId><version>${flink-cdc.version}</version></dependency><dependency><groupId>org.apache.flink</groupId><artifactId>flink-table-runtime</artifactId><version>${flink.version}</version></dependency><dependency><groupId>org.apache.flink</groupId><artifactId>flink-table-planner-loader</artifactId><version>${flink.version}</version></dependency></dependencies>

基本上项目需要的所有jar包都有了，不够以后在加。

2.目录创建

按照以上目录结构进行目录创建

二、代码编写

1.log4j.properties

log4j.rootLogger=error,stdout
log4j.appender.stdout=org.apache.log4j.ConsoleAppender
log4j.appender.stdout.target=System.out
log4j.appender.stdout.layout=org.apache.log4j.PatternLayout
log4j.appender.stdout.layout.ConversionPattern=%d %p [%c] - %m%n

2.CreateEnvUtil.java

这个文件中有两个方法
创建初始化Flink的env
Flink连接mysql的MySqlSource

package com.atguigu.tms.realtime.utils;import com.esotericsoftware.minlog.Log;
import com.ververica.cdc.connectors.mysql.source.MySqlSource;
import com.ververica.cdc.connectors.mysql.source.MySqlSourceBuilder;
import com.ververica.cdc.connectors.mysql.table.StartupOptions;
import com.ververica.cdc.debezium.JsonDebeziumDeserializationSchema;
import org.apache.flink.api.common.restartstrategy.RestartStrategies;
import org.apache.flink.api.common.time.Time;
import org.apache.flink.api.java.utils.ParameterTool;
import org.apache.flink.runtime.state.hashmap.HashMapStateBackend;
import org.apache.flink.streaming.api.CheckpointingMode;
import org.apache.flink.streaming.api.environment.CheckpointConfig;
import org.apache.flink.streaming.api.environment.StreamExecutionEnvironment;
import org.apache.kafka.connect.json.DecimalFormat;
import org.apache.kafka.connect.json.JsonConverterConfig;import java.util.HashMap;public class CreateEnvUtil {public static StreamExecutionEnvironment getStreamEnv(String[] args) {// 1.1 指定流处理环境StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();// 2.检查点相关设置// 2.1 开启检查点env.enableCheckpointing(6000L, CheckpointingMode.EXACTLY_ONCE);// 2.2 设置检查点的超时时间env.getCheckpointConfig().setCheckpointTimeout(120000L);// 2.3 设置job取消之后 检查点是否保留env.getCheckpointConfig().setExternalizedCheckpointCleanup(CheckpointConfig.ExternalizedCheckpointCleanup.RETAIN_ON_CANCELLATION);// 2.4 设置两个检查点之间的最小时间间隔env.getCheckpointConfig().setMinPauseBetweenCheckpoints(30000L);// 2.5 设置重启策略env.setRestartStrategy(RestartStrategies.failureRateRestart(3, Time.days(1), Time.seconds(3)));// 2.6 设置状态后端env.setStateBackend(new HashMapStateBackend());env.getCheckpointConfig().setCheckpointStorage("hdfs://hadoop102:8020/tms/ck");// 2.7 设置操作hdfs用户// 获取命令行参数ParameterTool parameterTool = ParameterTool.fromArgs(args);String hdfsUserName = parameterTool.get("hadoop-user-name", "atguigu");System.setProperty("HADOOP_USER_NAME", hdfsUserName);return env;}public static MySqlSource<String> getMysqlSource(String option, String serverId, String[] args) {ParameterTool parameterTool = ParameterTool.fromArgs(args);String mysqlHostname = parameterTool.get("hadoop-user-name", "hadoop102");int mysqlPort = Integer.parseInt(parameterTool.get("mysql-port", "3306"));String mysqlUsername = parameterTool.get("mysql-username", "root");String mysqlPasswd = parameterTool.get("mysql-passwd", "000000");option = parameterTool.get("start-up-option", option);serverId = parameterTool.get("server-id", serverId);// 创建配置信息 Map 集合，将 Decimal 数据类型的解析格式配置 k-v 置于其中HashMap config = new HashMap<>();config.put(JsonConverterConfig.DECIMAL_FORMAT_CONFIG, DecimalFormat.NUMERIC.name());// 将前述 Map 集合中的配置信息传递给 JSON 解析 Schema，该 Schema 将用于 MysqlSource 的初始化JsonDebeziumDeserializationSchema jsonDebeziumDeserializationSchema =new JsonDebeziumDeserializationSchema(false, config);MySqlSourceBuilder<String> builder = MySqlSource.<String>builder().hostname(mysqlHostname).port(mysqlPort).username(mysqlUsername).password(mysqlPasswd).deserializer(jsonDebeziumDeserializationSchema);switch (option) {// 读取实时数据case "dwd":String[] dwdTables = new String[]{"tms.order_info","tms.order_cargo","tms.transport_task","tms.order_org_bound"};return builder.databaseList("tms").tableList(dwdTables).startupOptions(StartupOptions.latest()).serverId(serverId).build();// 读取维度数据case "realtime_dim":String[] realtimeDimTables = new String[]{"tms.user_info","tms.user_address","tms.base_complex","tms.base_dic","tms.base_region_info","tms.base_organ","tms.express_courier","tms.express_courier_complex","tms.employee_info","tms.line_base_shift","tms.line_base_info","tms.truck_driver","tms.truck_info","tms.truck_model","tms.truck_team"};return builder.databaseList("tms").tableList(realtimeDimTables).startupOptions(StartupOptions.initial()).serverId(serverId).build();}Log.error("不支持操作类型");return null;}
}

3.KafkaUtil.java

该文件中有一个方法，创建Flink连接Kafka需要的Sink

package com.atguigu.tms.realtime.utils;import org.apache.flink.api.common.serialization.SimpleStringSchema;
import org.apache.flink.api.java.utils.ParameterTool;
import org.apache.flink.connector.base.DeliveryGuarantee;
import org.apache.flink.connector.kafka.sink.KafkaRecordSerializationSchema;
import org.apache.flink.connector.kafka.sink.KafkaSink;
import org.apache.kafka.clients.producer.ProducerConfig;public class KafkaUtil {private static final String KAFKA_SERVER = "hadoop102:9092,hadoop103:9092,hadoop104:9092";public static KafkaSink<String> getKafkaSink(String topic, String transIdPrefix, String[] args) {// 将命令行参数对象封装为 ParameterTool 类对象ParameterTool parameterTool = ParameterTool.fromArgs(args);// 提取命令行传入的 key 为 topic 的配置信息，并将默认值指定为方法参数 topic// 当命令行没有指定 topic 时，会采用默认值topic = parameterTool.get("topic", topic);// 如果命令行没有指定主题名称且默认值为 null 则抛出异常if (topic == null) {throw new IllegalArgumentException("主题名不可为空：命令行传参为空且没有默认值!");}// 获取命令行传入的 key 为 bootstrap-servers 的配置信息，并指定默认值String bootstrapServers = parameterTool.get("bootstrap-severs", KAFKA_SERVER);// 获取命令行传入的 key 为 transaction-timeout 的配置信息，并指定默认值String transactionTimeout = parameterTool.get("transaction-timeout", 15 * 60 * 1000 + "");return KafkaSink.<String>builder().setBootstrapServers(bootstrapServers).setRecordSerializer(KafkaRecordSerializationSchema.builder().setTopic(topic).setValueSerializationSchema(new SimpleStringSchema()).build()).setDeliveryGuarantee(DeliveryGuarantee.AT_LEAST_ONCE).setTransactionalIdPrefix(transIdPrefix).setProperty(ProducerConfig.TRANSACTION_TIMEOUT_CONFIG, transactionTimeout).build();}public static KafkaSink<String> getKafkaSink(String topic, String[] args) {return getKafkaSink(topic, topic + "_trans", args);}
}

4.OdsApp.java

Ods层的app创建，负责读取和写入数据

package com.atguigu.tms.realtime.app.ods;import com.alibaba.fastjson.JSON;
import com.alibaba.fastjson.JSONObject;
import com.atguigu.tms.realtime.utils.CreateEnvUtil;
import com.atguigu.tms.realtime.utils.KafkaUtil;
import com.esotericsoftware.minlog.Log;
import com.ververica.cdc.connectors.mysql.source.MySqlSource;
import org.apache.flink.api.common.eventtime.WatermarkStrategy;
import org.apache.flink.api.java.functions.KeySelector;
import org.apache.flink.streaming.api.datastream.KeyedStream;
import org.apache.flink.streaming.api.datastream.SingleOutputStreamOperator;
import org.apache.flink.streaming.api.environment.StreamExecutionEnvironment;
import org.apache.flink.streaming.api.functions.ProcessFunction;
import org.apache.flink.util.Collector;public class OdsApp {public static void main(String[] args) throws Exception {// 1.获取流处理环境并指定检查点StreamExecutionEnvironment env = CreateEnvUtil.getStreamEnv(args);env.setParallelism(4);// 2 使用FlinkCDC从MySQL中读取数据-事实数据String dwdOption = "dwd";String dwdServerId = "6030";String dwdsourceName = "ods_app_dwd_source";mysqlToKafka(dwdOption, dwdServerId, dwdsourceName, env, args);// 3 使用FlinkCDC从MySQL中读取数据-维度数据String realtimeDimOption = "realtime_dim";String realtimeDimServerId = "6040";String realtimeDimsourceName = "ods_app_realtimeDim_source";mysqlToKafka(realtimeDimOption, realtimeDimServerId, realtimeDimsourceName, env, args);env.execute();}public static void mysqlToKafka(String option, String serverId, String sourceName, StreamExecutionEnvironment env, String[] args) {MySqlSource<String> MySqlSource = CreateEnvUtil.getMysqlSource(option, serverId, args);SingleOutputStreamOperator<String> dwdStrDS = env.fromSource(MySqlSource, WatermarkStrategy.noWatermarks(), sourceName).setParallelism(1).uid(option + sourceName);// 3 简单ETLSingleOutputStreamOperator<String> processDS = dwdStrDS.process(new ProcessFunction<String, String>() {@Overridepublic void processElement(String jsonStr, ProcessFunction<String, String>.Context ctx, Collector<String> out) {try {JSONObject jsonObj = JSONObject.parseObject(jsonStr);if (jsonObj.getJSONObject("after") != null && !"d".equals(jsonObj.getString("op"))) {
//                                System.out.println(jsonObj);Long tsMs = jsonObj.getLong("ts_ms");jsonObj.put("ts", tsMs);jsonObj.remove("ts_ms");String jsonString = jsonObj.toJSONString();out.collect(jsonString);}} catch (Exception e) {Log.error("从Flink-CDC得到的数据不是一个标准的json格式",e);}}}).setParallelism(1);// 4 按照主键进行分组，避免出现乱序KeyedStream<String, String> keyedDS = processDS.keyBy((KeySelector<String, String>) jsonStr -> {JSONObject jsonObj = JSON.parseObject(jsonStr);return jsonObj.getJSONObject("after").getString("id");});//将数据写入KafkakeyedDS.sinkTo(KafkaUtil.getKafkaSink("tms_ods", sourceName + "_transPre", args)).uid(option + "_ods_app_sink");}
}

三、代码测试

在虚拟机启动我们需要的组件，目前需要hadoop、zk、kafka和MySQL。

先开一个消费者进行消费。

bin/kafka-console-consumer.sh --bootstrap-server hadoop102:9092 --topic tms_ods

然后运行OdsApp.java
他会先读取维度数据，因为维度数据需要全量更新之前的数据。

当他消费结束后，我们运行jar包，获取事实数据。

java -jar tms-mock-2023-01-06.jar 

如果能消费到新数据，代表通道没问题，ODS层创建完成。

---

总结

至此ODS搭建完成。