Skip to content

Z
ZHHT-IRN-BD-ANALYSIS
Commit 93f800de by 芮蒙
Options

2022-12-22 切换使用旅行信息进行计算行车情况相关指标

parent ea80470e
Showing with 32 additions and 51 deletions
realtime/hologram-streaming/src/main/java/com/zhht/irn/functions/TravelCarInfoCoProcessFunction.java
......@@ -41,6 +41,8 @@ public class TravelCarInfoCoProcessFunction extends CoProcessFunction<TravelInfo
private transient DruidPooledConnection connection;
private static Map<String, Map<String, String>> dim_cnt_cross_lane_position;
SimpleDateFormat sdf = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss");
@Override
public void open(Configuration parameters) throws Exception {
......@@ -59,22 +61,9 @@ public class TravelCarInfoCoProcessFunction extends CoProcessFunction<TravelInfo
ListStateDescriptor cycleSectionStateDescriptor = new ListStateDescriptor("cycleSectionState", TypeInformation.of(Long.class));
StateTtlConfig stateTtlConfig4TravelInfo = StateTtlConfig
// 状态有效时间 12小时
.newBuilder(Time.minutes(15))
// 设置状态的更新类型
.setUpdateType(StateTtlConfig.UpdateType.OnCreateAndWrite)
// 已过期还未被清理掉的状态数据不返回给用户
.setStateVisibility(StateTtlConfig.StateVisibility.NeverReturnExpired)
// 过期对象的清理策略 全量清理
.cleanupFullSnapshot()
.build();
travelInfoStateDescriptor.enableTimeToLive(stateTtlConfig4TravelInfo);
StateTtlConfig stateTtlConfig4CycleData = StateTtlConfig
// 状态有效时间 3小时
.newBuilder(Time.minutes(3))
.newBuilder(Time.hours(3))
// 设置状态的更新类型
.setUpdateType(StateTtlConfig.UpdateType.OnCreateAndWrite)
// 已过期还未被清理掉的状态数据不返回给用户
......@@ -103,15 +92,19 @@ public class TravelCarInfoCoProcessFunction extends CoProcessFunction<TravelInfo
@Override
public void close() throws Exception {
super.close();
if(connection!=null){
connection.close();
}
}
//把每个路口的数据 车辆旅行信息数据，存储在状态中，当周期数据进来的时候，再进行触发计算
// 所有旅行信息存储在状态中，当周期数据过来的时候，去状态中挑选属于这个周期内的旅行信息的数据即可
@Override
public void processElement1(TravelInfo value, CoProcessFunction<TravelInfo, CycleSignalData, List<TravelLineSinkInfo>>.Context ctx,
Collector<List<TravelLineSinkInfo>> out) throws Exception {
System.out.println("路口id"+value.getCrossId() +"车辆id"+value.getCarId() + "进入路口时间："+sdf.format(value.getInCrossTime()));
String crossId = value.getCrossId();
Iterator<Map<String, List<TravelInfo>>> iterator = travelInfoState.values().iterator();
//如果状态为空，则进行初始化
......@@ -143,7 +136,7 @@ public class TravelCarInfoCoProcessFunction extends CoProcessFunction<TravelInfo
if(crossListMap != null && crossListMap.containsKey(value.getInCrossLineId())) {
List<TravelInfo> lineTravelInfos = crossListMap.get(value.getInCrossLineId());
//如果，当前车道的数据，积压超过2000条，则进行清理操作
if(lineTravelInfos.size()>2000){
if(lineTravelInfos.size()>10000){
System.out.println("移除积压的数据。。。。");
List<TravelInfo> travelInfoList = orderByInCrossTime(lineTravelInfos);
List<TravelInfo> travelInfoListNew = travelInfoList.subList(500, travelInfoList.size());
......@@ -166,6 +159,7 @@ public class TravelCarInfoCoProcessFunction extends CoProcessFunction<TravelInfo
Collector<List<TravelLineSinkInfo>> out) {
try {
System.out.println("路口id"+value.getCrossCode() +"周期id"+value.getCycleOrder() + "周期时间："+sdf.format(value.getBeginDateTime())+"~"+sdf.format(value.getEndDateTime()));
//判断当前周期是否执行过了,存在周期数据重复下发的情况，只触发一次计算
if (alreadyHandleCycleSignalDataState.contains(value.getCycleOrder())) {
System.out.println("周期数据重复----当前路口id是" + value.getCrossCode() + "当前周期已经计算完成了" + value.getCycleOrder());
......@@ -234,12 +228,6 @@ public class TravelCarInfoCoProcessFunction extends CoProcessFunction<TravelInfo
public void onTimer(long timestamp, CoProcessFunction<TravelInfo, CycleSignalData, List<TravelLineSinkInfo>>.OnTimerContext ctx, Collector<List<TravelLineSinkInfo>> out) throws Exception {
super.onTimer(timestamp, ctx, out);
//更新维表数据 只在每天12点的时候更新
SimpleDateFormat sdf = new SimpleDateFormat("HH");
if(sdf.format(timestamp).equals("12")){
dim_cnt_cross_lane_position = getLaneDimData(connection);
}
dim_cnt_cross_lane_position = getLaneDimData(connection);
}
......@@ -388,9 +376,12 @@ public class TravelCarInfoCoProcessFunction extends CoProcessFunction<TravelInfo
//当前周期用到的数据
thisData.add(t);
//如果，进入路口时距小于周期开始时间，且已经超过30min，则清理这个数据
} else if(t.getInCrossTime()<beginDateTime && beginDateTime - t.getInCrossTime() > 1000*60*30 ) {
}
else if(t.getInCrossTime()<beginDateTime && beginDateTime - t.getInCrossTime() > 1000*60*60 ) {
System.out.println("移除当前旅行信息-超过30分钟都没有触发计算"+t);
} else {
}
else {
//当前周期没有用的的数据
leftData.add(t);
}
......
realtime/hologram-streaming/src/main/java/com/zhht/irn/job/TravelSituationAnalysisJob.java
......@@ -44,7 +44,7 @@ public class TravelSituationAnalysisJob {
// checkpoint配置
CheckpointConfig checkpointConfig = env.getCheckpointConfig();
// checkpoint时间间隔3分钟
checkpointConfig.setCheckpointInterval(1 * 60 * 1000);
checkpointConfig.setCheckpointInterval(3 * 60 * 1000);
// 两次checkpoint中最短时间间隔1分钟
checkpointConfig.setMinPauseBetweenCheckpoints(60 * 1000);
......@@ -54,17 +54,13 @@ public class TravelSituationAnalysisJob {
env.getCheckpointConfig().setCheckpointingMode(CheckpointingMode.EXACTLY_ONCE);
//设置重启策略
env.setRestartStrategy(RestartStrategies.fixedDelayRestart(1, Time.seconds(10)));
env.setRestartStrategy(RestartStrategies.fixedDelayRestart(5, Time.seconds(10)));
/*
关于flink checkpoint存储的设置统一在集群配置文件中设置
checkpointConfig.enableExternalizedCheckpoints(CheckpointConfig.ExternalizedCheckpointCleanup.RETAIN_ON_CANCELLATION);
env.setStateBackend(new RocksDBStateBackend("file:///opt/rocksDb/", true));*/
Properties kafkaProperties = new Properties();
// kafkaProperties.setProperty("bootstrap.servers", "dn3.zhht:9092");
kafkaProperties.setProperty("bootstrap.servers", "139.9.157.176:9092");
kafkaProperties.setProperty("bootstrap.servers", "dn3.zhht:9092");
// kafkaProperties.setProperty("bootstrap.servers", "139.9.157.176:9092");
// kafkaProperties.setProperty("group.id", "travel_event");
kafkaProperties.setProperty("group.id", "trips_info2");
kafkaProperties.setProperty(ConsumerConfig.ENABLE_AUTO_COMMIT_CONFIG, "true");
......@@ -72,36 +68,23 @@ public class TravelSituationAnalysisJob {
Properties kafkaProperties2 = new Properties();
// kafkaProperties2.setProperty("bootstrap.servers", "dn3.zhht:9092");
kafkaProperties2.setProperty("bootstrap.servers", "139.9.157.176:9092");
kafkaProperties2.setProperty("bootstrap.servers", "dn3.zhht:9092");
// kafkaProperties2.setProperty("bootstrap.servers", "139.9.157.176:9092");
kafkaProperties2.setProperty("group.id", "cycle_signal2");
kafkaProperties2.setProperty(ConsumerConfig.ENABLE_AUTO_COMMIT_CONFIG, "true");
kafkaProperties2.setProperty(ConsumerConfig.AUTO_OFFSET_RESET_CONFIG, "latest");
//旅行事件
/* FlinkKafkaConsumer<TravelEvent> travelEvent =
new FlinkKafkaConsumer<TravelEvent>("trips_event_info",
new TravelEventKafkaSchema(),
kafkaProperties) {
};
DataStreamSource<TravelEvent> travelEventStream = env.addSource(travelEvent).setParallelism(3);*/
// FlinkKafkaConsumer<TravelInfo> travelInfo = new FlinkKafkaConsumer<>("trips_info", new TravelInfoKafkaSchema(), kafkaProperties);
// FlinkKafkaConsumer<String> travelInfo = new FlinkKafkaConsumer<>("t_info2", new SimpleStringSchema(), kafkaProperties);
//旅行信息数据
// FlinkKafkaConsumer<TravelInfo> travelInfo = new FlinkKafkaConsumer<>("t_info2", new TravelInfoKafkaSchema(), kafkaProperties);
FlinkKafkaConsumer<TravelInfo> travelInfo = new FlinkKafkaConsumer<>("trips_info", new TravelInfoKafkaSchema(), kafkaProperties);
// travelInfo.setStartFromEarliest();
DataStreamSource<TravelInfo> travelInfoStream = env.addSource(travelInfo).setParallelism(3);
// travelInfoStream.print();
//周期信号
// FlinkKafkaConsumer<CycleSignalData> cycleSignalData = new FlinkKafkaConsumer<>("signal_cycle_data", new CycleSignalKafkaSchema(), kafkaProperties2);
// FlinkKafkaConsumer<CycleSignalData> cycleSignalData = new FlinkKafkaConsumer<>("s_data2", new CycleSignalKafkaSchema(), kafkaProperties2);
FlinkKafkaConsumer<CycleSignalData> cycleSignalData = new FlinkKafkaConsumer<>("signal_cycle_data", new CycleSignalKafkaSchema(), kafkaProperties2);
// cycleSignalData.setStartFromEarliest();
DataStreamSource<CycleSignalData> cycleSignalDataStream = env.addSource(cycleSignalData);
// cycleSignalDataStream.print();
//2、使用connect，把2个流联合处理
ConnectedStreams<TravelInfo, CycleSignalData> connect = travelInfoStream
......
realtime/hologram-streaming/src/main/java/com/zhht/irn/sink/TravelLaneInfoSink.java
......@@ -6,6 +6,7 @@ import com.zhht.irn.utils.DruidConnectPoolUtils;
import org.apache.flink.configuration.Configuration;
import org.apache.flink.streaming.api.functions.sink.RichSinkFunction;
import java.sql.Connection;
import java.sql.PreparedStatement;
import java.text.SimpleDateFormat;
import java.util.Date;
......@@ -20,22 +21,26 @@ import java.util.List;
*/
public class TravelLaneInfoSink extends RichSinkFunction<List<TravelLineSinkInfo>> {
private transient DruidPooledConnection connection;
Connection connection;
@Override
public void open(Configuration parameters) throws Exception {
super.open(parameters);
connection = DruidConnectPoolUtils.getConnection();
System.out.println("建立数据库连接成功。。。。。。。");
}
@Override
public void close() throws Exception {
super.close();
if(connection!=null){
connection.close();
}
}
@Override
public void invoke(List<TravelLineSinkInfo> value, Context context) throws Exception {
super.invoke(value, context);
......@@ -79,7 +84,9 @@ public class TravelLaneInfoSink extends RichSinkFunction<List<TravelLineSinkInfo
preparedStatement.setDouble(17, t.getAverage_car_head_time_gap());
preparedStatement.addBatch();
}
preparedStatement.executeBatch();
int[] ints = preparedStatement.executeBatch();
System.out.println("写入数据成功。。。。。"+ints.length+"...");
}
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment