如何使用flink計數窗口

Question

當我在 windowfunction 之前使用 countwindow 時，在運行代碼后，我收到了錯誤，例如State migration is currently not supported ，詳細信息為

org.apache.flink.util.FlinkRuntimeException: org.apache.flink.util.StateMigrationException: State migration is currently not supported.
    at org.apache.flink.runtime.state.heap.HeapKeyedStateBackend.create(HeapKeyedStateBackend.java:216)
    at org.apache.flink.streaming.api.operators.InternalTimeServiceManager.createTimerPriorityQueue(InternalTimeServiceManager.java:121)
    at org.apache.flink.streaming.api.operators.InternalTimeServiceManager.registerOrGetTimerService(InternalTimeServiceManager.java:106)
    at org.apache.flink.streaming.api.operators.InternalTimeServiceManager.getInternalTimerService(InternalTimeServiceManager.java:87)
    at org.apache.flink.streaming.api.operators.AbstractStreamOperator.getInternalTimerService(AbstractStreamOperator.java:741)
    at org.apache.flink.streaming.runtime.operators.windowing.WindowOperator.open(WindowOperator.java:225)
    at org.apache.flink.streaming.runtime.operators.windowing.EvictingWindowOperator.open(EvictingWindowOperator.java:430)
    at org.apache.flink.streaming.runtime.tasks.StreamTask.openAllOperators(StreamTask.java:424)
    at org.apache.flink.streaming.runtime.tasks.StreamTask.invoke(StreamTask.java:290)
    at org.apache.flink.runtime.taskmanager.Task.run(Task.java:711)
    at java.lang.Thread.run(Thread.java:748)
Caused by: org.apache.flink.util.StateMigrationException: State migration is currently not supported.
    at org.apache.flink.util.StateMigrationException.notSupported(StateMigrationException.java:42)
    ... 11 more

我的生成器代碼是

SingleOutputStreamOperator singleOutputStream = iotDataSource.flatMap(new String2MetricFlatMapFunctionWithOutTimeAlignment())
    .filter(new WaterMarkFilterFunction(Time.seconds(WATERMARK_SIZE)))
    .filter(new DataCleanFilterFunction())
    .map(new RelateParkMapFunction())
    .assignTimestampsAndWatermarks(new CommonBoundedOutOfOrdernessTimestampExtractor(Time.seconds(MAX_OUT_OF_ORDERNESS)))
    .keyBy((KeySelector<Metric, String>) metric -> metric.getPark())
    .countWindow(10,1)
    .apply(new RealTimeAlarmWindowApplyFunction(CACHE_SIZE))
    .map(new ResultMapFunction())
    .disableChaining();

也許RealTimeAlarmWindowApplyFunction ，

public class RealTimeAlarmWindowApplyFunction extends RichWindowFunction<Metric, ResultAction, String, GlobalWindow> {

    private Logger logger = LoggerFactory.getLogger(RealTimeAlarmWindowApplyFunction.class);

    private int cacheDataSize;

    public RealTimeAlarmWindowApplyFunction(int cacheDataSize) {
        this.cacheDataSize = cacheDataSize;
    }

    @Override
    public void apply(String key, GlobalWindow window, Iterable<Metric> input, Collector<ResultAction> out) throws Exception {
        //...
        // process code with no error 
        //...
    }
}

Answer 1

現在這個問題在運行 flink 作業時被取消的 savepoit 選項解決了：

$FLINK_HOME/bin/flink run -ynm jobname -yjm 1024 -ytm 4096 -yn 1 -ys 4 -qu root.users.root -flink_job_jar /root/flink_jar/test/flinkJobTest.jar -args 16

ps：我總是用這樣的savepoit選項運行這個工作

$FLINK_HOME/bin/flink run -ynm jobname -yjm 1024 -ytm 4096 -yn 1 -ys 4 -qu root.users.root -flink_job_jar /root/flink_jar/test/flinkJobTest.jar -savepoint_prefix hdfs://nameservice1/flink-checkpoint -args 16

但是在我更改窗口類型后，不允許恢復保存點