一、目的
由于部分数据类型频率为1s，从而数据规模特别大，因此完整的JSON放在Hive中解析起来，尤其是在单机环境下，效率特别慢，无法满足业务需求。
而Flume的拦截器并不能很好的转换数据，因为只能采用Java方式，从Kafka的主题A中采集数据，并解析字段，然后写入到放在Kafka主题B中
二 、原始数据格式
JSON格式比较正常，对象中包含数组
{
"deviceNo": "39",
"sourceDeviceType": null,
"sn": null,
"model": null,
"createTime": "2024-09-03 14:10:00",
"data": {
"cycle": 300,
"evaluationList": [{
"laneNo": 1,
"laneType": null,
"volume": 3,
"queueLenMax": 11.43,
"sampleNum": 0,
"stopAvg": 0.54,
"delayAvg": 0.0,
"passRate": 0.0,
"travelDist": 140.0,
"travelTimeAvg": 0.0
},
{
"laneNo": 2,
"laneType": null,
"volume": 7,
"queueLenMax": 23.18,
"sampleNum": 0,
"stopAvg": 0.47,
"delayAvg": 10.57,
"passRate": 0.0,
"travelDist": 140.0,
"travelTimeAvg": 0.0
},
{
"laneNo": 3,
"laneType": null,
"volume": 9,
"queueLenMax": 11.54,
"sampleNum": 0,
"stopAvg": 0.18,
"delayAvg": 9.67,
"passRate": 0.0,
"travelDist": 140.0,
"travelTimeAvg": 0.0
},
{
"laneNo": 4,
"laneType": null,
"volume": 6,
"queueLenMax": 11.36,
"sampleNum": 0,
"stopAvg": 0.27,
"delayAvg": 6.83,
"passRate": 0.0,
"travelDist": 140.0,
"travelTimeAvg": 0.0
}]
}
}
三、Java代码
package com.kgc;
import com.fasterxml.jackson.databind.JsonNode;
import com.fasterxml.jackson.databind.ObjectMapper;
import org.apache.kafka.clients.consumer.ConsumerConfig;
import org.apache.kafka.clients.consumer.ConsumerRecord;
import org.apache.kafka.clients.consumer.ConsumerRecords;
import org.apache.kafka.clients.consumer.KafkaConsumer;
import org.apache.kafka.clients.producer.KafkaProducer;
import org.apache.kafka.clients.producer.ProducerConfig;
import org.apache.kafka.clients.producer.ProducerRecord;
import org.apache.kafka.clients.producer.RecordMetadata;
import org.apache.kafka.common.serialization.StringDeserializer;
import org.apache.kafka.common.serialization.StringSerializer;
import java.time.Duration;
import java.util.Collections;
import java.util.Properties;
public class KafkaKafkaEvaluation {
// 添加 Kafka Producer 配置
private static Properties producerProps() {
Properties props = new Properties();
props.put(ProducerConfig.BOOTSTRAP_SERVERS_CONFIG, "192.168.0.70:9092");
props.put(ProducerConfig.KEY_SERIALIZER_CLASS_CONFIG, StringSerializer.class);
props.put(ProducerConfig.VALUE_SERIALIZER_CLASS_CONFIG, StringSerializer.class);
props.put(ProducerConfig.ACKS_CONFIG, "-1");
props.put(ProducerConfig.RETRIES_CONFIG, "3");
props.put(ProducerConfig.BATCH_SIZE_CONFIG, "16384");
props.put(ProducerConfig.LINGER_MS_CONFIG, "1");
props.put(ProducerConfig.BUFFER_MEMORY_CONFIG, "33554432");
return props;
}
public static void main(String[] args) {
Properties prop = new Properties();
prop.put(ConsumerConfig.BOOTSTRAP_SERVERS_CONFIG, "192.168.0.70:9092");
prop.put(ConsumerConfig.KEY_DESERIALIZER_CLASS_CONFIG, StringDeserializer.class);
prop.put(ConsumerConfig.VALUE_DESERIALIZER_CLASS_CONFIG, StringDeserializer.class);
prop.put(ConsumerConfig.ENABLE_AUTO_COMMIT_CONFIG, "false");
prop.put(ConsumerConfig.AUTO_COMMIT_INTERVAL_MS_CONFIG, "1000");
prop.put(ConsumerConfig.AUTO_OFFSET_RESET_CONFIG, "earliest");
// 每一个消费，都要定义不同的Group_ID
prop.put(ConsumerConfig.GROUP_ID_CONFIG, "evaluation_group");
KafkaConsumer<String, String> consumer = new KafkaConsumer<>(prop);
consumer.subscribe(Collections.singleton("topic_internal_data_evaluation"));
ObjectMapper mapper = new ObjectMapper();
// 初始化 Kafka Producer
KafkaProducer<String, String> producer = new KafkaProducer<>(producerProps());
while (true) {
ConsumerRecords<String, String> records = consumer.poll(Duration.ofMillis(1000));
for (ConsumerRecord<String, String> record : records) {
try {
JsonNode rootNode = mapper.readTree(record.value());
System.out.println("原始数据"+rootNode);
String device_no = rootNode.get("deviceNo").asText();
String source_device_type = rootNode.get("sourceDeviceType").asText();
String sn = rootNode.get("sn").asText();
String model = rootNode.get("model").asText();
String create_time = rootNode.get("createTime").asText();
String cycle = rootNode.get("data").get("cycle").asText();
JsonNode evaluationList = rootNode.get("data").get("evaluationList");
for (JsonNode evaluationItem : evaluationList) {
String lane_no = evaluationItem.get("laneNo").asText();
String lane_type = evaluationItem.get("laneType").asText();
String volume = evaluationItem.get("volume").asText();
String queue_len_max = evaluationItem.get("queueLenMax").asText();
String sample_num = evaluationItem.get("sampleNum").asText();
String stop_avg = evaluationItem.get("stopAvg").asText();
String delay_avg = evaluationItem.get("delayAvg").asText();
String pass_rate = evaluationItem.get("passRate").asText();
String travel_dist = evaluationItem.get("travelDist").asText();
String travel_time_avg = evaluationItem.get("travelTimeAvg").asText();
String outputLine = String.format("%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s",
device_no, source_device_type, sn, model, create_time, cycle,lane_no, lane_type,
volume,queue_len_max,sample_num,stop_avg,delay_avg,pass_rate,travel_dist,travel_time_avg);
// 发送数据到 Kafka
ProducerRecord<String, String> producerRecord = new ProducerRecord<>("topic_db_data_evaluation", record.key(), outputLine);
producer.send(producerRecord, (RecordMetadata metadata, Exception e) -> {
if (e != null) {
e.printStackTrace();
} else {
System.out.println("The offset of the record we just sent is: " + metadata.offset());
}
});
}
} catch (Exception e) {
e.printStackTrace();
}
}
consumer.commitAsync();
}
}
}1、服务器IP都是 192.168.0.70
2、消费Kafka主题（数据源）：topic_internal_data_evaluation
3、生产Kafka主题（目标源）：topic_db_data_evaluation
4、注意：字段顺序与ODS层表结构字段顺序一致！！！
四、开启Kafka主题topic_db_data_evaluation消费者
root@localhost bin\]# ./kafka-console-consumer.sh --bootstrap-server 192.168.0.70:9092 --topic topic_db_data_evaluation --from-beginning ## 五、运行测试 ### 1、启动项目  ### 2、消费者输出数据  然后再用Flume采集写入HDFS就行了，不过ODS层表结构需要转变 ## 六、ODS层新表结构 ``` create external table if not exists hurys_dc_ods.ods_evaluation( device_no string COMMENT '设备编号', source_device_type string COMMENT '设备类型', sn string COMMENT '设备序列号 ', model string COMMENT '设备型号', create_time timestamp COMMENT '创建时间', cycle int COMMENT '评价数据周期', lane_no int COMMENT '车道编号', lane_type int COMMENT '车道类型 0:渠化1:来向2:出口3:去向4:左弯待转区5:直行待行区6:右转专用道99:未定义车道', volume int COMMENT '车道内过停止线流量（辆）', queue_len_max float COMMENT '车道内最大排队长度（m）', sample_num int COMMENT '评价数据计算样本量', stop_avg float COMMENT '车道内平均停车次数（次）', delay_avg float COMMENT '车道内平均延误时间（s）', pass_rate float COMMENT '车道内一次通过率', travel_dist float COMMENT '车道内检测行程距离（m）', travel_time_avg float COMMENT '车道内平均行程时间' ) comment '评价数据外部表——静态分区' partitioned by (day string) row format delimited fields terminated by ',' stored as SequenceFile ; ``` ## 七、Flume采集配置文件  ## 八、运行Flume任务，检查HDFS文件、以及ODS表数据 ``` --刷新表分区 msck repair table ods_evaluation; --查看表分区 show partitions hurys_dc_ods.ods_evaluation; --查看表数据 select * from hurys_dc_ods.ods_evaluation where day='2024-09-03'; ``` 搞定，这样就不需要在Hive中解析JSON数据了！！！