Flink-join的三种方式
Join/**** 将两个数据流,进行join** 如果让两个流能够join上,必须满足以下两个条件* 1.由于数据是分散在多台机器上,必须将join条件相同的数据通过网络传输到同一台机器的同一个分区中(按照条件进行KeyBy)* 2.让每个流中的数据都放慢,等等对方(划分相同类型,长度一样的窗口)**/public class EventTumblingWindowJoin {public st
·
Join
/**
*
* 将两个数据流,进行join
*
* 如果让两个流能够join上,必须满足以下两个条件
* 1.由于数据是分散在多台机器上,必须将join条件相同的数据通过网络传输到同一台机器的同一个分区中(按照条件进行KeyBy)
* 2.让每个流中的数据都放慢,等等对方(划分相同类型,长度一样的窗口)
*
*/
public class EventTumblingWindowJoin {
public static void main(String[] args) throws Exception {
StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
//1000,o001,c001
DataStreamSource<String> lines1 = env.socketTextStream("linux01", 7777);
//1200,c001,图书
DataStreamSource<String> lines2 = env.socketTextStream("linux01", 8888);
//按照EventTime进行join，窗口长度为5000秒，使用新的提取EventTime生成WaterMark的API
//提取两个流的Watermark
SingleOutputStreamOperator<String> lines1WithWatermark
= lines1.assignTimestampsAndWatermarks(WatermarkStrategy.<String>forBoundedOutOfOrderness(Duration.ofSeconds(0)).withTimestampAssigner(new SerializableTimestampAssigner<String>() {
@Override
public long extractTimestamp(String element, long recordTimestamp) {
return Long.parseLong(element.split(",")[0]);
}
}));
SingleOutputStreamOperator<String> lines2WithWatermark
= lines2.assignTimestampsAndWatermarks(WatermarkStrategy.<String>forBoundedOutOfOrderness(Duration.ofSeconds(0)).withTimestampAssigner(new SerializableTimestampAssigner<String>() {
@Override
public long extractTimestamp(String element, long recordTimestamp) {
return Long.parseLong(element.split(",")[0]);
}
}));
//对两个流进行处理
SingleOutputStreamOperator<Tuple3<Long, String, String>> tpStream1
= lines1WithWatermark.map(new MapFunction<String, Tuple3<Long, String, String>>() {
@Override
public Tuple3<Long, String, String> map(String input) throws Exception {
String[] fields = input.split(",");
return Tuple3.of(Long.parseLong(fields[0]), fields[1], fields[2]);
}
});
SingleOutputStreamOperator<Tuple3<Long, String, String>> tpStream2
= lines2WithWatermark.map(new MapFunction<String, Tuple3<Long, String, String>>() {
@Override
public Tuple3<Long, String, String> map(String input) throws Exception {
String[] fields = input.split(",");
return Tuple3.of(Long.parseLong(fields[0]), fields[1], fields[2]);
}
});
//将两个流join
DataStream<Tuple5<Long, String, String, Long, String>> result = tpStream1.join(tpStream2)
.where(tp1 -> tp1.f2) //第一个流keyBY的字段
.equalTo(tp2 -> tp2.f1) //第二个流keyBy的字段
.window(TumblingEventTimeWindows.of(Time.seconds(5))) //划分窗口
//全量聚合的处理逻辑
.apply(new JoinFunction<Tuple3<Long, String, String>, Tuple3<Long, String, String>, Tuple5<Long, String, String, Long, String>>() {
//窗口触发后，条件相同的，并且在同一个窗口内的数据，会传入到join方法中
@Override
public Tuple5<Long, String, String, Long, String> join(Tuple3<Long, String, String> first, Tuple3<Long, String, String> second) throws Exception {
return Tuple5.of(first.f0,first.f1,first.f2,second.f0,second.f2);
}
});
result.print();
env.execute();
}
}LeftOuterJoin
/**
* 将两个数据流,实现LeftOuterJoin
*
* 如果让两个流能够join上,必须满足以下两个条件
* 1.由于数据是分散在多台机器上,必须将join条件相同的数据通过网络传输到同一台机器的同一个分区中(按照条件进行KeyBy)
* 2.让每个流中的数据都放慢,等等对方(划分相同类型,长度一样的窗口)
*
*/
public class EventTumblingWindowLeftOuterJoin {
public static void main(String[] args) throws Exception {
StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
//1000,o001,c001
DataStreamSource<String> lines1 = env.socketTextStream("linux01", 7777);
//1200,c001,图书
DataStreamSource<String> lines2 = env.socketTextStream("linux01", 8888);
//按照EventTime进行join，窗口长度为5000秒，使用新的提取EventTime生成WaterMark的API
//提取两个流的Watermark
SingleOutputStreamOperator<String> lines1WithWatermark
= lines1.assignTimestampsAndWatermarks(WatermarkStrategy.<String>forBoundedOutOfOrderness(Duration.ofSeconds(0)).withTimestampAssigner(new SerializableTimestampAssigner<String>() {
@Override
public long extractTimestamp(String element, long recordTimestamp) {
return Long.parseLong(element.split(",")[0]);
}
}));
SingleOutputStreamOperator<String> lines2WithWatermark
= lines2.assignTimestampsAndWatermarks(WatermarkStrategy.<String>forBoundedOutOfOrderness(Duration.ofSeconds(0)).withTimestampAssigner(new SerializableTimestampAssigner<String>() {
@Override
public long extractTimestamp(String element, long recordTimestamp) {
return Long.parseLong(element.split(",")[0]);
}
}));
//对两个流进行处理
SingleOutputStreamOperator<Tuple3<Long, String, String>> tpStream1
= lines1WithWatermark.map(new MapFunction<String, Tuple3<Long, String, String>>() {
@Override
public Tuple3<Long, String, String> map(String input) throws Exception {
String[] fields = input.split(",");
return Tuple3.of(Long.parseLong(fields[0]), fields[1], fields[2]);
}
});
SingleOutputStreamOperator<Tuple3<Long, String, String>> tpStream2
= lines2WithWatermark.map(new MapFunction<String, Tuple3<Long, String, String>>() {
@Override
public Tuple3<Long, String, String> map(String input) throws Exception {
String[] fields = input.split(",");
return Tuple3.of(Long.parseLong(fields[0]), fields[1], fields[2]);
}
});
//将两个流leftOuterJoin
DataStream<Tuple5<Long, String, String, Long, String>> result = tpStream1.coGroup(tpStream2)
.where(tp1 -> tp1.f2) //第一个流keyBy的字段
.equalTo(tp2 -> tp2.f1)//第二个流keyBy的字段
.window(TumblingEventTimeWindows.of(Time.seconds(5)))//划分窗口
.apply(new CoGroupFunction<Tuple3<Long, String, String>, Tuple3<Long, String, String>, Tuple5<Long, String, String, Long, String>>() {
/**
* coGroup当窗口触发后，每个key会调用一次coGroup
* 三种情况会调用coGroup方法
* 1.第一个流和第二个流中，都有key相同的数据数据，并且在同一个窗口呢，那么coGroup方法中的两个Iterable都不为empty
* 2.第一个流中出现了同一个key的数据，.第二个流中没有出现相同key的数据，那么coGroup方法中的第一个Iterable不为empty，第二个为empty
* 3.第二个流中出现了同一个key的数据，.第一个流中没有出现相同key的数据，那么coGroup方法中的第二个Iterable不为empty，第一个为empty
* @param first
* @param second
* @param out
* @throws Exception
*/
@Override
public void coGroup(Iterable<Tuple3<Long, String, String>> first, Iterable<Tuple3<Long, String, String>> second, Collector<Tuple5<Long, String, String, Long, String>> out) throws Exception {
for (Tuple3<Long, String, String> left : first) {
//实现左外连接
//先循环左流的数据
boolean isEmpty = false;
for (Tuple3<Long, String, String> right : second) {
isEmpty = true;
out.collect(Tuple5.of(left.f0, left.f1, left.f2, right.f0, right.f2));
}
if (!isEmpty) {
out.collect(Tuple5.of(left.f0, left.f1, left.f2, null, null));
}
}
}
});
result.print();
env.execute();
}
}intervalJoin
/**
* 将两个数据流不划分窗口，按照时间范围进行join，即intervalJoin
*
* 以第一个流中的数据为标准进行比较时间
*
* 实现步骤：
* 1.分别将两个流按照相同的条件进行KeyBy（可以保证key等值的数据一定进入到同一台机器的同一个分区中）
* 2.将两个数据流的数据缓存到KeyedState，然后将两个流Connected到一起（可以共享状态）
*
*/
public class EventTumblingWindowIntervalJoin {
public static void main(String[] args) throws Exception {
StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
//1000,o001,c001
DataStreamSource<String> lines1 = env.socketTextStream("linux01", 7777);
//1200,c001,图书
DataStreamSource<String> lines2 = env.socketTextStream("linux01", 8888);
//按照EventTime进行join，窗口长度为5000秒，使用新的提取EventTime生成WaterMark的API
//提取两个流的Watermark
SingleOutputStreamOperator<String> lines1WithWatermark
= lines1.assignTimestampsAndWatermarks(WatermarkStrategy.<String>forBoundedOutOfOrderness(Duration.ofSeconds(0)).withTimestampAssigner(new SerializableTimestampAssigner<String>() {
@Override
public long extractTimestamp(String element, long recordTimestamp) {
return Long.parseLong(element.split(",")[0]);
}
}));
SingleOutputStreamOperator<String> lines2WithWatermark
= lines2.assignTimestampsAndWatermarks(WatermarkStrategy.<String>forBoundedOutOfOrderness(Duration.ofSeconds(0)).withTimestampAssigner(new SerializableTimestampAssigner<String>() {
@Override
public long extractTimestamp(String element, long recordTimestamp) {
return Long.parseLong(element.split(",")[0]);
}
}));
//对两个流进行处理
SingleOutputStreamOperator<Tuple3<Long, String, String>> tpStream1
= lines1WithWatermark.map(new MapFunction<String, Tuple3<Long, String, String>>() {
@Override
public Tuple3<Long, String, String> map(String input) throws Exception {
String[] fields = input.split(",");
return Tuple3.of(Long.parseLong(fields[0]), fields[1], fields[2]);
}
});
SingleOutputStreamOperator<Tuple3<Long, String, String>> tpStream2
= lines2WithWatermark.map(new MapFunction<String, Tuple3<Long, String, String>>() {
@Override
public Tuple3<Long, String, String> map(String input) throws Exception {
String[] fields = input.split(",");
return Tuple3.of(Long.parseLong(fields[0]), fields[1], fields[2]);
}
});
//将两个流join
KeyedStream<Tuple3<Long, String, String>, String> keyedStream1 = tpStream1.keyBy(tp -> tp.f2);
KeyedStream<Tuple3<Long, String, String>, String> keyedStream2 = tpStream2.keyBy(tp -> tp.f1);
SingleOutputStreamOperator<Tuple5<Long, String, String, Long, String>> result = keyedStream1.intervalJoin(keyedStream2)
.between(Time.seconds(-1), Time.seconds(1)) //指定的时间范围
.upperBoundExclusive() //不包括上界
.process(new ProcessJoinFunction<Tuple3<Long, String, String>, Tuple3<Long, String, String>, Tuple5<Long, String, String, Long, String>>() {
@Override
public void processElement(Tuple3<Long, String, String> left, Tuple3<Long, String, String> right, Context ctx, Collector<Tuple5<Long, String, String, Long, String>> out) throws Exception {
out.collect(Tuple5.of(left.f0,left.f1,left.f2,right.f0,right.f2));
}
});
result.print();
env.execute();
}
}
为开发者提供学习成长、分享交流、生态实践、资源工具等服务，帮助开发者快速成长。
更多推荐
1
0- 0
已为社区贡献4条内容
所有评论(0)