flink 状态参数设置

本文主要是介绍flink 状态参数设置，希望对大家解决编程问题提供一定的参考价值，需要的开发者们随着小编来一起学习吧！

前提

代码示例，通过flink消费kafka，查看list状态中的数据，确定参数的具体含义
kafka的代码：发送两个key值，一秒发送一次

	for(int i = 0; i< 100; i++){JSONObject object = new JSONObject();object.put("id", 1);object.put("value", i);String s = object.toJSONString();kafkaProducer.send(new ProducerRecord("test_topic_partition_one", s.getBytes(StandardCharsets.UTF_8))).get();object = new JSONObject();object.put("id", 2);object.put("value", 100 + i);s = object.toJSONString();kafkaProducer.send(new ProducerRecord("test_topic_partition_one", s.getBytes(StandardCharsets.UTF_8))).get();Thread.sleep(1000);}

flink消费kafka示例：

	final StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();env.enableCheckpointing(10 * 1000);KafkaSource<String> source = KafkaSource.<String>builder().setBootstrapServers("broker:9092").setProperties(properties).setTopics("test_topic_partition_one").setGroupId("my-group").setStartingOffsets(OffsetsInitializer.latest()).setValueOnlyDeserializer(new SimpleStringSchema()).build();DataStreamSource<String> kafkaSource = env.fromSource(source, WatermarkStrategy.noWatermarks(), "Kafka Source").setParallelism(2);DataStream<Tuple2<String, Integer>> dataStream = kafkaSource.map(new MapFunction<String, Tuple2<String, Integer>>() {@Overridepublic Tuple2<String, Integer> map(String value) throws Exception {JSONObject object = JSONObject.parseObject(value);return new Tuple2<String, Integer>(object.getString("id"), object.getInteger("value"));}});DataStream<String> resultStream = dataStream.keyBy(value -> value.f0) // 根据第一个字段（键）进行分组.process(new ListValueProcess());// 打印结果resultStream.print();

ListValueProcess状态函数

 	@Overridepublic void processElement(Tuple2<String, Integer> value, KeyedProcessFunction<String, Tuple2<String, Integer>, String>.Context ctx, Collector<String> out) throws Exception {// 添加元素到 ListStatelistState.add(value.f1);// 获取 ListState 中的所有元素，并输出它们String key = value.f0;List<Integer> list = new ArrayList<>();for (Integer integer : listState.get()) {list.add(integer);}String result = "key:" + key + ", value:" +list;// 输出结果out.collect(result);}@Overridepublic void open(Configuration parameters) throws Exception {super.open(parameters);StateTtlConfig ttlConfig = StateTtlConfig.newBuilder(Time.seconds(10)).setUpdateType(StateTtlConfig.UpdateType.OnCreateAndWrite).setStateVisibility(StateTtlConfig.StateVisibility.ReturnExpiredIfNotCleanedUp).build();// 初始化 ListState// 不同的key 具有不用的listState// 用于存储一个key多个值ListStateDescriptor<Integer> integerListStateDescriptor = new ListStateDescriptor<>("my-list-state", Integer.class);integerListStateDescriptor.enableTimeToLive(ttlConfig);listState = getRuntimeContext().getListState(integerListStateDescriptor);}

可以看到StateTtlConfig大部份有三个参数

• 指定状态保存时间
• setUpdateType 设置状态更新策略：OnCreateAndWrite和 OnReadAndWrite
• setStateVisibility 设置状态可见行 ：ReturnExpiredIfNotCleanedUp和 NeverReturnExpired

这里我们保存状态时间是10s
OnCreateAndWrite: 表示当状态被创建与更新的时候，表示更新了状态
OnReadAndWrite：表示状态被创建与更新和读取的时候，表示更新了状态
ReturnExpiredIfNotCleanedUp：表示状态过期了但没有删除，也可以读取到状态
NeverReturnExpired：表示状态过期就读取不到

结果示例：
当：OnCreateAndWrite和ReturnExpiredIfNotCleanedUp时

1> key:1, value:[6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24]
1> key:2, value:[113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124]
1> key:1, value:[6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25]
1> key:2, value:[113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125]
1> key:1, value:[6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26]
1> key:2, value:[113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126]
1> key:1, value:[18, 19, 20, 21, 22, 23, 24, 25, 26, 27]
1> key:2, value:[113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127]
1> key:1, value:[18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28]
1> key:2, value:[113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128]
1> key:1, value:[18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29]
1> key:2, value:[113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129]
1> key:1, value:[18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30]
1> key:2, value:[113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130]

可以看到，状态会定期删除过期的数据，而且数据可见可能大于10s的范围。

当OnCreateAndWrite和NeverReturnExpired时

1> key:2, value:[109, 110, 111, 112, 113, 114, 115, 116, 117]
1> key:1, value:[10, 11, 12, 13, 14, 15, 16, 17, 18]
1> key:2, value:[110, 111, 112, 113, 114, 115, 116, 117, 118]
1> key:1, value:[11, 12, 13, 14, 15, 16, 17, 18, 19]
1> key:2, value:[111, 112, 113, 114, 115, 116, 117, 118, 119]
1> key:1, value:[12, 13, 14, 15, 16, 17, 18, 19, 20]
1> key:2, value:[112, 113, 114, 115, 116, 117, 118, 119, 120]
1> key:1, value:[13, 14, 15, 16, 17, 18, 19, 20, 21]
1> key:2, value:[113, 114, 115, 116, 117, 118, 119, 120, 121]
1> key:1, value:[14, 15, 16, 17, 18, 19, 20, 21, 22]
1> key:2, value:[114, 115, 116, 117, 118, 119, 120, 121, 122]

可以看到，状态的数据只保留最近10s内的值

当OnReadAndWrite和ReturnExpiredIfNotCleanedUp时

1> key:1, value:[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25]
1> key:2, value:[101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125]
1> key:1, value:[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26]
1> key:2, value:[101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126]
1> key:1, value:[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27]
1> key:2, value:[101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127]
1> key:1, value:[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28]
1> key:2, value:[101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128]
1> key:1, value:[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29]
1> key:2, value:[101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129]
1> key:1, value:[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30]
1> key:2, value:[101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130]
1> key:1, value:[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31]
1> key:2, value:[101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131]
1> key:1, value:[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32]
1> key:2, value:[101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132]

可以看到，状态保留了所有的数据，因为每次都会读取了数据，所以不会过期

当OnReadAndWrite和NeverReturnExpired时

1> key:1, value:[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26]
1> key:2, value:[101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126]
1> key:1, value:[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27]
1> key:2, value:[101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127]
1> key:1, value:[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28]
1> key:2, value:[101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128]
1> key:1, value:[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29]
1> key:2, value:[101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129]
1> key:1, value:[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30]
1> key:2, value:[101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130]
1> key:1, value:[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31]
1> key:2, value:[101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131]

可以看到，状态保留了所有的数据，因为每次都会读取了数据，所以不会过期

这篇关于flink 状态参数设置的文章就介绍到这儿，希望我们推荐的文章对编程师们有所帮助！

---