Deprecated API

Contents

• Deprecated Interfaces
• Deprecated Classes
• Deprecated Enums
• Deprecated Fields
• Deprecated Methods
• Deprecated Constructors

Deprecated Interfaces

Interface and Description
org.apache.flink.streaming.api.functions.AssignerWithPeriodicWatermarks
org.apache.flink.streaming.api.functions.AssignerWithPunctuatedWatermarks
org.apache.flink.streaming.util.serialization.DeserializationSchema Use DeserializationSchema instead.
org.apache.flink.streaming.api.checkpoint.ExternallyInducedSource This interface is based on the SourceFunction API, which is due to be removed. Use the new Source API instead.
org.apache.flink.streaming.api.checkpoint.ListCheckpointed If you need to do non-keyed state snapshots of your operator, use CheckpointedFunction. This should only be needed in rare cases, though.
org.apache.flink.streaming.api.functions.source.ParallelSourceFunction This interface is based on the SourceFunction API, which is due to be removed. Use the new Source API instead.
org.apache.flink.streaming.util.serialization.SerializationSchema Use SerializationSchema instead.
org.apache.flink.streaming.api.operators.SetupableStreamOperator This class is deprecated in favour of using StreamOperatorFactory and it's StreamOperatorFactory.createStreamOperator(org.apache.flink.streaming.api.operators.StreamOperatorParameters<OUT>) and passing the required parameters to the Operator's constructor in create method.
org.apache.flink.streaming.api.functions.source.SourceFunction This interface will be removed in future versions. Use the new Source interface instead. NOTE: All sub-tasks from FLINK-28045 must be closed before this API can be completely removed.
org.apache.flink.streaming.api.functions.TimestampAssigner use TimestampAssigner
org.apache.flink.streaming.api.connector.sink2.WithPostCommitTopology Please implement Sink, SupportsCommitter and SupportsPostCommitTopology instead.
org.apache.flink.streaming.api.connector.sink2.WithPreCommitTopology Please implement Sink, SupportsCommitter and SupportsPreCommitTopology instead.
org.apache.flink.streaming.api.connector.sink2.WithPreWriteTopology Please implement Sink and SupportsPreWriteTopology instead.

Deprecated Classes

Class and Description
org.apache.flink.streaming.util.serialization.AbstractDeserializationSchema Use AbstractDeserializationSchema instead.
org.apache.flink.streaming.api.functions.AscendingTimestampExtractor Extend AscendingTimestampExtractor instead.
org.apache.flink.streaming.api.functions.timestamps.AscendingTimestampExtractor
org.apache.flink.streaming.api.functions.source.ContinuousFileMonitoringFunction This class is based on the SourceFunction API, which is due to be removed. Use the new Source API instead.
org.apache.flink.streaming.api.functions.source.datagen.DataGeneratorSource Use org.apache.flink.connector.datagen.source.DataGeneratorSource instead.
org.apache.flink.streaming.api.functions.source.FileMonitoringFunction Internal class deprecated in favour of ContinuousFileMonitoringFunction.
org.apache.flink.streaming.api.functions.source.FileReadFunction Internal class deprecated in favour of ContinuousFileMonitoringFunction.
org.apache.flink.streaming.api.functions.source.FromElementsFunction This class is based on the SourceFunction API, which is due to be removed. Use the new Source API instead.
org.apache.flink.streaming.api.functions.source.FromIteratorFunction This class is based on the SourceFunction API, which is due to be removed. Use the new Source API instead.
org.apache.flink.streaming.api.functions.source.FromSplittableIteratorFunction This class is based on the SourceFunction API, which is due to be removed. Use the new Source API instead.
org.apache.flink.streaming.api.functions.IngestionTimeExtractor
org.apache.flink.streaming.api.functions.source.InputFormatSourceFunction This class is based on the SourceFunction API, which is due to be removed. Use the new Source API instead.
org.apache.flink.streaming.api.datastream.IterativeStream This method is deprecated since Flink 1.19. The only known use case of this Iteration API comes from Flink ML, which already has its own implementation of iteration and no longer uses this API. If there's any use cases other than Flink ML that needs iteration support, please reach out to dev@flink.apache.org and we can consider making the Flink ML iteration implementation a separate common library.
org.apache.flink.streaming.api.datastream.IterativeStream.ConnectedIterativeStreams This method is deprecated since Flink 1.19. The only known use case of this Iteration API comes from Flink ML, which already has its own implementation of iteration and no longer uses this API. If there's any use cases other than Flink ML that needs iteration support, please reach out to dev@flink.apache.org and we can consider making the Flink ML iteration implementation a separate common library.
org.apache.flink.streaming.api.operators.co.LegacyKeyedCoProcessOperator Replaced by KeyedCoProcessOperator which takes KeyedCoProcessFunction
org.apache.flink.streaming.api.operators.LegacyKeyedProcessOperator Replaced by KeyedProcessOperator which takes KeyedProcessFunction
org.apache.flink.streaming.api.functions.source.MessageAcknowledgingSourceBase This class is based on the SourceFunction API, which is due to be removed. Use the new Source API instead.
org.apache.flink.streaming.api.functions.source.MultipleIdsMessageAcknowledgingSourceBase This class is based on the SourceFunction API, which is due to be removed. Use the new Source API instead.
org.apache.flink.streaming.api.functions.sink.OutputFormatSinkFunction Please use the StreamingFileSink for writing to files from a streaming program.
org.apache.flink.streaming.api.datastream.QueryableStateStream The Queryable State feature is deprecated since Flink 1.18, and will be removed in a future Flink major version.
org.apache.flink.streaming.api.functions.source.RichParallelSourceFunction This class is based on the SourceFunction API, which is due to be removed. Use the new Source API instead.
org.apache.flink.streaming.api.functions.windowing.RichProcessAllWindowFunction use ProcessAllWindowFunction instead
org.apache.flink.streaming.api.functions.windowing.RichProcessWindowFunction use ProcessWindowFunction instead
org.apache.flink.streaming.api.functions.source.RichSourceFunction This class is based on the SourceFunction API, which is due to be removed. Use the new Source API instead.
org.apache.flink.streaming.util.serialization.SimpleStringSchema Use SimpleStringSchema instead.
org.apache.flink.streaming.api.functions.source.SocketTextStreamFunction This class is based on the SourceFunction API, which is due to be removed. Use the new Source API instead.
org.apache.flink.streaming.runtime.tasks.SourceStreamTask This class is based on the SourceFunction API, which is due to be removed. Use the new Source API instead.
org.apache.flink.streaming.api.functions.source.StatefulSequenceSource This class is based on the SourceFunction API, which is due to be removed. Use the new Source API instead.
org.apache.flink.streaming.api.functions.sink.filesystem.StreamingFileSink Use org.apache.flink.connector.file.sink.FileSink instead.
org.apache.flink.streaming.api.operators.StreamSource This class is based on the SourceFunction API, which is due to be removed. Use the new Source API instead.
org.apache.flink.streaming.api.operators.StreamSourceContexts This class is based on the SourceFunction API, which is due to be removed. Use the new Source API instead.
org.apache.flink.streaming.runtime.tasks.StreamTaskActionExecutor.SynchronizedStreamTaskActionExecutor this class should only be used in SourceStreamTask which exposes the checkpoint lock as part of Public API.
org.apache.flink.streaming.api.windowing.time.Time Use Duration
org.apache.flink.streaming.util.serialization.TypeInformationSerializationSchema Use TypeInformationSerializationSchema instead.
org.apache.flink.streaming.api.functions.sink.WriteFormat Please use the StreamingFileSink for writing to files from a streaming program.
org.apache.flink.streaming.api.functions.sink.WriteFormatAsCsv Please use the StreamingFileSink for writing to files from a streaming program.
org.apache.flink.streaming.api.functions.sink.WriteFormatAsText Please use the StreamingFileSink for writing to files from a streaming program.
org.apache.flink.streaming.api.functions.sink.WriteSinkFunction Please use the StreamingFileSink for writing to files from a streaming program.
org.apache.flink.streaming.api.functions.sink.WriteSinkFunctionByMillis Please use the StreamingFileSink for writing to files from a streaming program.

Deprecated Enums

Enum and Description

org.apache.flink.streaming.api.TimeCharacteristic In Flink 1.12 the default stream time characteristic has been changed to TimeCharacteristic.EventTime, thus you don't need to call this method for enabling event-time support anymore. Explicitly using processing-time windows and timers works in event-time mode. If you need to disable watermarks, please use ExecutionConfig.setAutoWatermarkInterval(long). If you are using TimeCharacteristic.IngestionTime, please manually set an appropriate WatermarkStrategy. If you are using generic "time window" operations (for example KeyedStream.timeWindow(org.apache.flink.streaming.api.windowing.time.Time) that change behaviour based on the time characteristic, please use equivalent operations that explicitly specify processing time or event time.

Deprecated Fields

Field and Description
org.apache.flink.streaming.api.environment.ExecutionCheckpointingOptions.ALIGNMENT_TIMEOUT Use ExecutionCheckpointingOptions.ALIGNED_CHECKPOINT_TIMEOUT instead.
org.apache.flink.streaming.api.environment.CheckpointConfig.DEFAULT_CHECKPOINT_ID_OF_IGNORED_IN_FLIGHT_DATA This field is no longer used. Please use ExecutionCheckpointingOptions.CHECKPOINT_ID_OF_IGNORED_IN_FLIGHT_DATA instead.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.DEFAULT_JOB_NAME This constant does not fit well to batch runtime mode.
org.apache.flink.streaming.api.environment.CheckpointConfig.DEFAULT_MAX_CONCURRENT_CHECKPOINTS This field is no longer used. Please use ExecutionCheckpointingOptions.MAX_CONCURRENT_CHECKPOINTS instead.
org.apache.flink.streaming.api.environment.CheckpointConfig.DEFAULT_MIN_PAUSE_BETWEEN_CHECKPOINTS This field is no longer used. Please use ExecutionCheckpointingOptions.MIN_PAUSE_BETWEEN_CHECKPOINTS instead.
org.apache.flink.streaming.api.environment.CheckpointConfig.DEFAULT_MODE
org.apache.flink.streaming.api.environment.CheckpointConfig.DEFAULT_TIMEOUT This field is no longer used. Please use ExecutionCheckpointingOptions.CHECKPOINTING_TIMEOUT instead.
org.apache.flink.streaming.api.environment.ExecutionCheckpointingOptions.FORCE_CHECKPOINTING This will be removed once iterations properly participate in checkpointing.
org.apache.flink.streaming.api.environment.StreamPipelineOptions.TIME_CHARACTERISTIC In Flink 1.12 the default stream time characteristic has been changed to TimeCharacteristic.EventTime, thus you don't need to set this option for enabling event-time support anymore. Explicitly using processing-time windows and timers works in event-time mode. If you need to disable watermarks, please set PipelineOptions.AUTO_WATERMARK_INTERVAL to 0. If you are using TimeCharacteristic.IngestionTime, please manually set an appropriate WatermarkStrategy. If you are using generic "time window" operations (for example KeyedStream.timeWindow(org.apache.flink.streaming.api.windowing.time.Time) that change behaviour based on the time characteristic, please use equivalent operations that explicitly specify processing time or event time.
org.apache.flink.streaming.api.environment.CheckpointConfig.UNDEFINED_TOLERABLE_CHECKPOINT_NUMBER This field is no longer used.

Deprecated Methods

Method and Description
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.addDefaultKryoSerializer(Class<?>, Class<? extends Serializer<?>>) Register data types and serializers through hard codes is deprecated, because you need to modify the codes when upgrading job version. You should configure this by config option PipelineOptions.SERIALIZATION_CONFIG.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.addDefaultKryoSerializer(Class<?>, T) Register data types and serializers through hard codes is deprecated, because you need to modify the codes when upgrading job version. Instance-type serializer definition where serializers are serialized and written into the snapshot and deserialized for use is deprecated as well. Use class-type serializer definition by PipelineOptions.SERIALIZATION_CONFIG instead, where only the class name is written into the snapshot and new instance of the serializer is created for use. This is a breaking change, and it will be removed in Flink 2.0.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.addSource(SourceFunction<OUT>) This method relies on the SourceFunction API, which is due to be removed. Use the StreamExecutionEnvironment.fromSource(Source, WatermarkStrategy, String) method based on the new Source API instead.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.addSource(SourceFunction<OUT>, String) This method relies on the SourceFunction API, which is due to be removed. Use the StreamExecutionEnvironment.fromSource(Source, WatermarkStrategy, String) method based on the new Source API instead.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.addSource(SourceFunction<OUT>, String, TypeInformation<OUT>) This method relies on the SourceFunction API, which is due to be removed. Use the StreamExecutionEnvironment.fromSource(Source, WatermarkStrategy, String, TypeInformation) method based on the new Source API instead.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.addSource(SourceFunction<OUT>, TypeInformation<OUT>) This method relies on the SourceFunction API, which is due to be removed. Use the StreamExecutionEnvironment.fromSource(Source, WatermarkStrategy, String, TypeInformation) method based on the new Source API instead.
org.apache.flink.streaming.api.datastream.JoinedStreams.WithWindow.allowedLateness(Time) Use JoinedStreams.WithWindow.allowedLateness(Duration).
org.apache.flink.streaming.api.datastream.CoGroupedStreams.WithWindow.allowedLateness(Time) Use CoGroupedStreams.WithWindow.allowedLateness(Duration)
org.apache.flink.streaming.api.datastream.AllWindowedStream.allowedLateness(Time) Use AllWindowedStream.allowedLateness(Duration), instead.
org.apache.flink.streaming.api.datastream.WindowedStream.allowedLateness(Time) Use WindowedStream.allowedLateness(Duration)
org.apache.flink.streaming.runtime.operators.windowing.WindowOperatorBuilder.allowedLateness(Time) Use WindowOperatorBuilder.allowedLateness(Duration).
org.apache.flink.streaming.api.datastream.AllWindowedStream.apply(ReduceFunction<T>, AllWindowFunction<T, R, W>) Use AllWindowedStream.reduce(ReduceFunction, AllWindowFunction) instead.
org.apache.flink.streaming.api.datastream.AllWindowedStream.apply(ReduceFunction<T>, AllWindowFunction<T, R, W>, TypeInformation<R>) Use AllWindowedStream.reduce(ReduceFunction, AllWindowFunction, TypeInformation) instead.
org.apache.flink.streaming.api.datastream.WindowedStream.apply(ReduceFunction<T>, WindowFunction<T, R, K, W>) Use WindowedStream.reduce(ReduceFunction, WindowFunction) instead.
org.apache.flink.streaming.api.datastream.WindowedStream.apply(ReduceFunction<T>, WindowFunction<T, R, K, W>, TypeInformation<R>) Use WindowedStream.reduce(ReduceFunction, WindowFunction, TypeInformation) instead.
org.apache.flink.streaming.api.datastream.KeyedStream.asQueryableState(String) The Queryable State feature is deprecated since Flink 1.18, and will be removed in a future Flink major version.
org.apache.flink.streaming.api.datastream.KeyedStream.asQueryableState(String, ReducingStateDescriptor<T>) The Queryable State feature is deprecated since Flink 1.18, and will be removed in a future Flink major version.
org.apache.flink.streaming.api.datastream.KeyedStream.asQueryableState(String, ValueStateDescriptor<T>) The Queryable State feature is deprecated since Flink 1.18, and will be removed in a future Flink major version.
org.apache.flink.streaming.api.datastream.DataStream.assignTimestampsAndWatermarks(AssignerWithPeriodicWatermarks<T>) Please use DataStream.assignTimestampsAndWatermarks(WatermarkStrategy) instead.
org.apache.flink.streaming.api.datastream.DataStream.assignTimestampsAndWatermarks(AssignerWithPunctuatedWatermarks<T>) Please use DataStream.assignTimestampsAndWatermarks(WatermarkStrategy) instead.
org.apache.flink.streaming.api.datastream.KeyedStream.IntervalJoin.between(Time, Time) Use KeyedStream.IntervalJoin.between(Duration, Duration)
org.apache.flink.streaming.api.datastream.DataStreamUtils.collect(DataStream<OUT>) Please use DataStream.executeAndCollect().
org.apache.flink.streaming.api.datastream.DataStreamUtils.collect(DataStream<OUT>, String) Please use DataStream.executeAndCollect().
org.apache.flink.streaming.api.datastream.DataStreamUtils.collectBoundedStream(DataStream<E>, String) Please use DataStream.executeAndCollect().
org.apache.flink.streaming.api.datastream.DataStreamUtils.collectRecordsFromUnboundedStream(ClientAndIterator<E>, int) Please use DataStream.executeAndCollect().
org.apache.flink.streaming.api.datastream.DataStreamUtils.collectUnboundedStream(DataStream<E>, int, String) Please use DataStream.executeAndCollect().
org.apache.flink.streaming.api.datastream.DataStreamUtils.collectWithClient(DataStream<OUT>, String) Please use DataStream.executeAndCollect().
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.enableCheckpointing() Use StreamExecutionEnvironment.enableCheckpointing(long) instead.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.enableCheckpointing(long, CheckpointingMode, boolean) Use StreamExecutionEnvironment.enableCheckpointing(long, CheckpointingMode) instead. Forcing checkpoints will be removed in the future.
org.apache.flink.streaming.api.environment.CheckpointConfig.enableExternalizedCheckpoints(CheckpointConfig.ExternalizedCheckpointCleanup) use CheckpointConfig.setExternalizedCheckpointCleanup(ExternalizedCheckpointCleanup) instead.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.fromCollection(Collection<OUT>) This method will be removed a future release, possibly as early as version 2.0. Use StreamExecutionEnvironment.fromData(Collection) instead.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.fromCollection(Collection<OUT>, TypeInformation<OUT>) This method will be removed a future release, possibly as early as version 2.0. Use StreamExecutionEnvironment.fromData(Collection, TypeInformation) instead.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.fromCollection(Iterator<OUT>, Class<OUT>) This method will be removed a future release, possibly as early as version 2.0. Use StreamExecutionEnvironment.fromData(Collection, TypeInformation) instead. For rate-limited data generation, use DataGeneratorSource with RateLimiterStrategy. If you need to use a fixed set of elements in such scenario, combine it with FromElementsGeneratorFunction.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.fromCollection(Iterator<OUT>, TypeInformation<OUT>) This method will be removed a future release, possibly as early as version 2.0. Use StreamExecutionEnvironment.fromData(Collection, TypeInformation) instead. For rate-limited data generation, use DataGeneratorSource with RateLimiterStrategy. If you need to use a fixed set of elements in such scenario, combine it with FromElementsGeneratorFunction.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.fromElements(Class<OUT>, OUT...) This method will be removed a future release, possibly as early as version 2.0. Use #fromData(OUT...) instead.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.fromElements(OUT...) This method will be removed a future release, possibly as early as version 2.0. Use #fromData(OUT...) instead.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.generateSequence(long, long) Use StreamExecutionEnvironment.fromSequence(long, long) instead to create a new data stream that contains NumberSequenceSource.
org.apache.flink.streaming.api.environment.CheckpointConfig.getAlignmentTimeout() User CheckpointConfig.getAlignedCheckpointTimeout() instead.
org.apache.flink.streaming.api.environment.CheckpointConfig.getCheckpointStorage() The method is marked as deprecated because starting from Flink 1.19, the usage of all complex Java objects related to configuration, including their getter and setter methods, should be replaced by ConfigOption. In a future major version of Flink, this method will be removed entirely. It is recommended to find which checkpoint storage is used by checkpoint storage ConfigOption. For more details on using ConfigOption for checkpoint storage configuration, please refer to the Flink documentation: Checkpoints
org.apache.flink.streaming.api.environment.RemoteStreamEnvironment.getClientConfiguration() This method is going to be removed in the next releases.
org.apache.flink.streaming.api.windowing.assigners.WindowAssigner.getDefaultTrigger(StreamExecutionEnvironment) the method is deprecated since Flink 1.19 because StreamExecutionEnvironment is unused. Please use getDefaultTrigger and override this method with an empty body instead. 1. If you override getDefaultTrigger(), the getDefaultTrigger() will be invoked and the getDefaultTrigger(StreamExecutionEnvironment env) won't be invoked. 2. If you don't override getDefaultTrigger(), the getDefaultTrigger(StreamExecutionEnvironment env) will be invoked in the default implementation of the getDefaultTrigger().
org.apache.flink.streaming.api.windowing.triggers.Trigger.TriggerContext.getKeyValueState(String, Class<S>, S) Use Trigger.TriggerContext.getPartitionedState(StateDescriptor).
org.apache.flink.streaming.api.windowing.triggers.Trigger.TriggerContext.getKeyValueState(String, TypeInformation<S>, S) Use Trigger.TriggerContext.getPartitionedState(StateDescriptor).
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.getNumberOfExecutionRetries() This method will be replaced by StreamExecutionEnvironment.getRestartStrategy().
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.getRestartStrategy() The method is marked as deprecated because starting from Flink 1.19, the usage of all complex Java objects related to configuration, including their getter and setter methods, should be replaced by ConfigOption. In a future major version of Flink, this method will be removed entirely. It is recommended to switch to using the ConfigOptions provided by RestartStrategyOptions for configuring restart strategies.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.getStateBackend() The method is marked as deprecated because starting from Flink 1.19, the usage of all complex Java objects related to configuration, including their getter and setter methods, should be replaced by ConfigOption. In a future major version of Flink, this method will be removed entirely. It is recommended to find which state backend is used by state backend ConfigOption. For more details on using ConfigOption for state backend configuration, please refer to the Flink documentation: state-backends
org.apache.flink.streaming.api.graph.StreamGraph.getStreamEdgesOrThrow(int, int)
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.getStreamTimeCharacteristic() See StreamExecutionEnvironment.setStreamTimeCharacteristic(TimeCharacteristic) for deprecation notice.
org.apache.flink.streaming.api.graph.StreamConfig.getTypeSerializerIn1(ClassLoader)
org.apache.flink.streaming.api.graph.StreamConfig.getTypeSerializerIn2(ClassLoader)
org.apache.flink.streaming.api.functions.sink.SinkFunction.invoke(IN) Use SinkFunction.invoke(Object, Context).
org.apache.flink.streaming.api.environment.CheckpointConfig.isFailOnCheckpointingErrors() Use CheckpointConfig.getTolerableCheckpointFailureNumber().
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.isForceCheckpointing() Forcing checkpoints will be removed in future version.
org.apache.flink.streaming.api.environment.CheckpointConfig.isForceCheckpointing() This will be removed once iterations properly participate in checkpointing.
org.apache.flink.streaming.api.datastream.DataStream.iterate() This method is deprecated since Flink 1.19. The only known use case of this Iteration API comes from Flink ML, which already has its own implementation of iteration and no longer uses this API. If there's any use cases other than Flink ML that needs iteration support, please reach out to dev@flink.apache.org and we can consider making the Flink ML iteration implementation a separate common library.
org.apache.flink.streaming.api.datastream.DataStream.iterate(long) This method is deprecated since Flink 1.19. The only known use case of this Iteration API comes from Flink ML, which already has its own implementation of iteration and no longer uses this API. If there's any use cases other than Flink ML that needs iteration support, please reach out to dev@flink.apache.org and we can consider making the Flink ML iteration implementation a separate common library.
org.apache.flink.streaming.api.datastream.DataStream.keyBy(int...) Use DataStream.keyBy(KeySelector).
org.apache.flink.streaming.api.datastream.DataStream.keyBy(String...) Use DataStream.keyBy(KeySelector).
org.apache.flink.streaming.api.windowing.assigners.TumblingEventTimeWindows.of(Time) Use TumblingEventTimeWindows.of(Duration)
org.apache.flink.streaming.api.windowing.assigners.TumblingProcessingTimeWindows.of(Time) Use TumblingProcessingTimeWindows.of(Duration)
org.apache.flink.streaming.api.windowing.triggers.ContinuousProcessingTimeTrigger.of(Time) Use ContinuousProcessingTimeTrigger.of(Duration)
org.apache.flink.streaming.api.windowing.triggers.ContinuousEventTimeTrigger.of(Time) Use ContinuousEventTimeTrigger.of(Duration)
org.apache.flink.streaming.api.windowing.evictors.TimeEvictor.of(Time) Use TimeEvictor.of(Duration)
org.apache.flink.streaming.api.windowing.evictors.TimeEvictor.of(Time, boolean) Use TimeEvictor.of(Duration, boolean)
org.apache.flink.streaming.api.windowing.assigners.SlidingProcessingTimeWindows.of(Time, Time) SlidingProcessingTimeWindows.of(Duration, Duration)
org.apache.flink.streaming.api.windowing.assigners.SlidingEventTimeWindows.of(Time, Time) Use SlidingEventTimeWindows.of(Duration, Duration)
org.apache.flink.streaming.api.windowing.assigners.TumblingEventTimeWindows.of(Time, Time) Use TumblingEventTimeWindows.of(Duration, Duration)
org.apache.flink.streaming.api.windowing.assigners.TumblingProcessingTimeWindows.of(Time, Time) Use TumblingProcessingTimeWindows.of(Duration, Duration)
org.apache.flink.streaming.api.windowing.assigners.SlidingProcessingTimeWindows.of(Time, Time, Time) Use SlidingProcessingTimeWindows.of(Duration, Duration, Duration)
org.apache.flink.streaming.api.windowing.assigners.SlidingEventTimeWindows.of(Time, Time, Time) Use SlidingEventTimeWindows.of(Duration, Duration, Duration)
org.apache.flink.streaming.api.windowing.assigners.TumblingEventTimeWindows.of(Time, Time, WindowStagger) Use TumblingEventTimeWindows.of(Duration, Duration, WindowStagger)
org.apache.flink.streaming.api.windowing.assigners.TumblingProcessingTimeWindows.of(Time, Time, WindowStagger) Use TumblingProcessingTimeWindows.of(Duration, Duration, WindowStagger)
org.apache.flink.streaming.api.functions.source.FromSplittableIteratorFunction.open(Configuration) This method is deprecated since Flink 1.19. The users are recommended to implement open(OpenContext openContext) and override open(Configuration parameters) with an empty body instead. 1. If you implement open(OpenContext openContext), the open(OpenContext openContext) will be invoked and the open(Configuration parameters) won't be invoked. 2. If you don't implement open(OpenContext openContext), the open(Configuration parameters) will be invoked in the default implementation of the open(OpenContext openContext).
org.apache.flink.streaming.api.functions.source.MultipleIdsMessageAcknowledgingSourceBase.open(Configuration) This method is deprecated since Flink 1.19. The users are recommended to implement open(OpenContext openContext) and override open(Configuration parameters) with an empty body instead. 1. If you implement open(OpenContext openContext), the open(OpenContext openContext) will be invoked and the open(Configuration parameters) won't be invoked. 2. If you don't implement open(OpenContext openContext), the open(Configuration parameters) will be invoked in the default implementation of the open(OpenContext openContext).
org.apache.flink.streaming.api.functions.sink.PrintSinkFunction.open(Configuration) This method is deprecated since Flink 1.19. The users are recommended to implement open(OpenContext openContext) and override open(Configuration parameters) with an empty body instead. 1. If you implement open(OpenContext openContext), the open(OpenContext openContext) will be invoked and the open(Configuration parameters) won't be invoked. 2. If you don't implement open(OpenContext openContext), the open(Configuration parameters) will be invoked in the default implementation of the open(OpenContext openContext).
org.apache.flink.streaming.api.functions.sink.OutputFormatSinkFunction.open(Configuration) This method is deprecated since Flink 1.19. The users are recommended to implement open(OpenContext openContext) and override open(Configuration parameters) with an empty body instead. 1. If you implement open(OpenContext openContext), the open(OpenContext openContext) will be invoked and the open(Configuration parameters) won't be invoked. 2. If you don't implement open(OpenContext openContext), the open(Configuration parameters) will be invoked in the default implementation of the open(OpenContext openContext).
org.apache.flink.streaming.api.functions.sink.SocketClientSink.open(Configuration) This method is deprecated since Flink 1.19. The users are recommended to implement open(OpenContext openContext) and override open(Configuration parameters) with an empty body instead. 1. If you implement open(OpenContext openContext), the open(OpenContext openContext) will be invoked and the open(Configuration parameters) won't be invoked. 2. If you don't implement open(OpenContext openContext), the open(Configuration parameters) will be invoked in the default implementation of the open(OpenContext openContext).
org.apache.flink.streaming.api.datastream.DataStream.partitionCustom(Partitioner<K>, int) use DataStream.partitionCustom(Partitioner, KeySelector).
org.apache.flink.streaming.api.datastream.DataStream.partitionCustom(Partitioner<K>, String) use DataStream.partitionCustom(Partitioner, KeySelector).
org.apache.flink.streaming.api.datastream.KeyedStream.process(ProcessFunction<T, R>) Use KeyedStream.process(KeyedProcessFunction)
org.apache.flink.streaming.api.datastream.KeyedStream.process(ProcessFunction<T, R>, TypeInformation<R>) Use KeyedStream.process(KeyedProcessFunction, TypeInformation)
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.readFile(FileInputFormat<OUT>, String) Use FileSource#forRecordStreamFormat()/forBulkFileFormat()/forRecordFileFormat() instead. An example of reading a file using a simple TextLineInputFormat: FileSource<String> source = FileSource.forRecordStreamFormat( new TextLineInputFormat(), new Path("/foo/bar")) .build();
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.readFile(FileInputFormat<OUT>, String, FileProcessingMode, long) Use FileSource#forRecordStreamFormat()/forBulkFileFormat()/forRecordFileFormat() instead. An example of reading a file using a simple TextLineInputFormat: FileSource<String> source = FileSource.forRecordStreamFormat( new TextLineInputFormat(), new Path("/foo/bar")) .monitorContinuously(Duration.of(10, SECONDS)) .build();
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.readFile(FileInputFormat<OUT>, String, FileProcessingMode, long, FilePathFilter) Use FileInputFormat.setFilesFilter(FilePathFilter) to set a filter and StreamExecutionEnvironment.readFile(FileInputFormat, String, FileProcessingMode, long)
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.readFile(FileInputFormat<OUT>, String, FileProcessingMode, long, TypeInformation<OUT>) Use FileSource#forRecordStreamFormat()/forBulkFileFormat()/forRecordFileFormat() instead. An example of reading a file using a simple TextLineInputFormat: FileSource<String> source = FileSource.forRecordStreamFormat( new TextLineInputFormat(), new Path("/foo/bar")) .monitorContinuously(Duration.of(10, SECONDS)) .build();
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.readFileStream(String, long, FileMonitoringFunction.WatchType) Use StreamExecutionEnvironment.readFile(FileInputFormat, String, FileProcessingMode, long) instead.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.readTextFile(String) Use FileSource#forRecordStreamFormat()/forBulkFileFormat()/forRecordFileFormat() instead. An example of reading a file using a simple TextLineInputFormat: FileSource<String> source = FileSource.forRecordStreamFormat( new TextLineInputFormat(), new Path("/foo/bar")) .build();
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.readTextFile(String, String) Use FileSource#forRecordStreamFormat()/forBulkFileFormat()/forRecordFileFormat() instead. An example of reading a file using a simple TextLineInputFormat: FileSource<String> source = FileSource.forRecordStreamFormat( new TextLineInputFormat("UTF-8"), new Path("/foo/bar")) .build();
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.registerType(Class<?>) Register data types and serializers through hard codes is deprecated, because you need to modify the codes when upgrading job version. You should configure this by config option PipelineOptions.SERIALIZATION_CONFIG.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.registerTypeWithKryoSerializer(Class<?>, Class<? extends Serializer>) Register data types and serializers through hard codes is deprecated, because you need to modify the codes when upgrading job version. You should configure this by config option PipelineOptions.SERIALIZATION_CONFIG.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.registerTypeWithKryoSerializer(Class<?>, T) Register data types and serializers through hard codes is deprecated, because you need to modify the codes when upgrading job version. Instance-type serializer definition where serializers are serialized and written into the snapshot and deserialized for use is deprecated as well. Use class-type serializer definition by PipelineOptions.SERIALIZATION_CONFIG instead, where only the class name is written into the snapshot and new instance of the serializer is created for use. This is a breaking change, and it will be removed in Flink 2.0.
org.apache.flink.streaming.api.environment.CheckpointConfig.setAlignmentTimeout(Duration) Use CheckpointConfig.setAlignedCheckpointTimeout(Duration) instead.
org.apache.flink.streaming.api.environment.CheckpointConfig.setCheckpointStorage(CheckpointStorage) The method is marked as deprecated because starting from Flink 1.19, the usage of all complex Java objects related to configuration, including their getter and setter methods, should be replaced by ConfigOption. In a future major version of Flink, this method will be removed entirely. It is recommended to switch to using the ConfigOptions provided for configuring checkpoint storage like the following code snippet: Configuration config = new Configuration(); config.set(CheckpointingOptions.CHECKPOINT_STORAGE, "filesystem"); config.set(CheckpointingOptions.CHECKPOINTS_DIRECTORY, "file:///flink/checkpoints"); StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment(config); For more details on using ConfigOption for checkpoint storage configuration, please refer to the Flink documentation: Checkpoints
org.apache.flink.streaming.api.environment.CheckpointConfig.setCheckpointStorage(Path) The method is marked as deprecated because starting from Flink 1.19, the usage of all complex Java objects related to configuration, including their getter and setter methods, should be replaced by ConfigOption. In a future major version of Flink, this method will be removed entirely. It is recommended to switch to using the ConfigOptions provided for configuring checkpoint storage like the following code snippet: Configuration config = new Configuration(); config.set(CheckpointingOptions.CHECKPOINT_STORAGE, "filesystem"); config.set(CheckpointingOptions.CHECKPOINTS_DIRECTORY, "file:///flink/checkpoints"); StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment(config); For more details on using ConfigOption for checkpoint storage configuration, please refer to the Flink documentation: Checkpoints
org.apache.flink.streaming.api.environment.CheckpointConfig.setCheckpointStorage(String) The method is marked as deprecated because starting from Flink 1.19, the usage of all complex Java objects related to configuration, including their getter and setter methods, should be replaced by ConfigOption. In a future major version of Flink, this method will be removed entirely. It is recommended to switch to using the ConfigOptions provided for configuring checkpoint storage like the following code snippet: Configuration config = new Configuration(); config.set(CheckpointingOptions.CHECKPOINT_STORAGE, "filesystem"); config.set(CheckpointingOptions.CHECKPOINTS_DIRECTORY, "file:///flink/checkpoints"); StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment(config); For more details on using ConfigOption for checkpoint storage configuration, please refer to the Flink documentation: Checkpoints
org.apache.flink.streaming.api.environment.CheckpointConfig.setCheckpointStorage(URI) The method is marked as deprecated because starting from Flink 1.19, the usage of all complex Java objects related to configuration, including their getter and setter methods, should be replaced by ConfigOption. In a future major version of Flink, this method will be removed entirely. It is recommended to switch to using the ConfigOptions provided for configuring checkpoint storage like the following code snippet: Configuration config = new Configuration(); config.set(CheckpointingOptions.CHECKPOINT_STORAGE, "filesystem"); config.set(CheckpointingOptions.CHECKPOINTS_DIRECTORY, "file:///flink/checkpoints"); StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment(config); For more details on using ConfigOption for checkpoint storage configuration, please refer to the Flink documentation: Checkpoints
org.apache.flink.streaming.api.environment.CheckpointConfig.setFailOnCheckpointingErrors(boolean) Use CheckpointConfig.setTolerableCheckpointFailureNumber(int).
org.apache.flink.streaming.api.environment.CheckpointConfig.setForceCheckpointing(boolean) This will be removed once iterations properly participate in checkpointing.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.setNumberOfExecutionRetries(int) This method will be replaced by StreamExecutionEnvironment.setRestartStrategy(org.apache.flink.api.common.restartstrategy.RestartStrategies.RestartStrategyConfiguration). The RestartStrategies.fixedDelayRestart(int, Duration) contains the number of execution retries.
org.apache.flink.streaming.api.operators.AbstractStreamOperator.setProcessingTimeService(ProcessingTimeService) The ProcessingTimeService instance should be passed by the operator constructor and this method will be removed along with SetupableStreamOperator.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.setRestartStrategy(RestartStrategies.RestartStrategyConfiguration) The method is marked as deprecated because starting from Flink 1.19, the usage of all complex Java objects related to configuration, including their getter and setter methods, should be replaced by ConfigOption. In a future major version of Flink, this method will be removed entirely. It is recommended to switch to using the ConfigOptions provided by RestartStrategyOptions for configuring restart strategies.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.setStateBackend(StateBackend) The method is marked as deprecated because starting from Flink 1.19, the usage of all complex Java objects related to configuration, including their getter and setter methods, should be replaced by ConfigOption. In a future major version of Flink, this method will be removed entirely. It is recommended to switch to using the ConfigOptions provided for configuring state backend like the following code snippet: Configuration config = new Configuration(); config.set(StateBackendOptions.STATE_BACKEND, "hashmap"); StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment(config); For more details on using ConfigOption for state backend configuration, please refer to the Flink documentation: state-backends
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.setStreamTimeCharacteristic(TimeCharacteristic) In Flink 1.12 the default stream time characteristic has been changed to TimeCharacteristic.EventTime, thus you don't need to call this method for enabling event-time support anymore. Explicitly using processing-time windows and timers works in event-time mode. If you need to disable watermarks, please use ExecutionConfig.setAutoWatermarkInterval(long). If you are using TimeCharacteristic.IngestionTime, please manually set an appropriate WatermarkStrategy. If you are using generic "time window" operations (for example through KeyedStream.window(WindowAssigner) that change behaviour based on the time characteristic, please use equivalent operations that explicitly specify processing time or event time.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.socketTextStream(String, int, char) Use StreamExecutionEnvironment.socketTextStream(String, int, String) instead.
org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.socketTextStream(String, int, char, long) Use StreamExecutionEnvironment.socketTextStream(String, int, String, long) instead.
org.apache.flink.streaming.api.datastream.KeyedStream.timeWindow(Time) Please use KeyedStream.window(WindowAssigner) with either TumblingEventTimeWindows or TumblingProcessingTimeWindows. For more information, see the deprecation notice on TimeCharacteristic
org.apache.flink.streaming.api.datastream.KeyedStream.timeWindow(Time, Time) Please use KeyedStream.window(WindowAssigner) with either SlidingEventTimeWindows or SlidingProcessingTimeWindows. For more information, see the deprecation notice on TimeCharacteristic
org.apache.flink.streaming.api.datastream.DataStream.timeWindowAll(Time) Please use DataStream.windowAll(WindowAssigner) with either TumblingEventTimeWindows or TumblingProcessingTimeWindows. For more information, see the deprecation notice on TimeCharacteristic
org.apache.flink.streaming.api.datastream.DataStream.timeWindowAll(Time, Time) Please use DataStream.windowAll(WindowAssigner) with either SlidingEventTimeWindows or SlidingProcessingTimeWindows. For more information, see the deprecation notice on TimeCharacteristic
org.apache.flink.streaming.api.datastream.CoGroupedStreams.WithWindow.with(CoGroupFunction<T1, T2, T>) This method will be removed once the CoGroupedStreams.WithWindow.apply(CoGroupFunction) method is fixed in the next major version of Flink (2.0).
org.apache.flink.streaming.api.datastream.CoGroupedStreams.WithWindow.with(CoGroupFunction<T1, T2, T>, TypeInformation<T>) This method will be removed once the CoGroupedStreams.WithWindow.apply(CoGroupFunction, TypeInformation) method is fixed in the next major version of Flink (2.0).
org.apache.flink.streaming.api.datastream.JoinedStreams.WithWindow.with(FlatJoinFunction<T1, T2, T>) This method will be removed once the JoinedStreams.WithWindow.apply(FlatJoinFunction) method is fixed in the next major version of Flink (2.0).
org.apache.flink.streaming.api.datastream.JoinedStreams.WithWindow.with(FlatJoinFunction<T1, T2, T>, TypeInformation<T>) This method will be replaced by JoinedStreams.WithWindow.apply(FlatJoinFunction, TypeInformation) in Flink 2.0. So use the JoinedStreams.WithWindow.apply(FlatJoinFunction, TypeInformation) in the future.
org.apache.flink.streaming.api.datastream.JoinedStreams.WithWindow.with(JoinFunction<T1, T2, T>) This method will be removed once the JoinedStreams.WithWindow.apply(JoinFunction) method is fixed in the next major version of Flink (2.0).
org.apache.flink.streaming.api.datastream.JoinedStreams.WithWindow.with(JoinFunction<T1, T2, T>, TypeInformation<T>) This method will be removed once the JoinedStreams.WithWindow.apply(JoinFunction, TypeInformation) method is fixed in the next major version of Flink (2.0).
org.apache.flink.streaming.api.windowing.assigners.ProcessingTimeSessionWindows.withGap(Time) Use ProcessingTimeSessionWindows.withGap(Duration)
org.apache.flink.streaming.api.windowing.assigners.EventTimeSessionWindows.withGap(Time) Use EventTimeSessionWindows.withGap(Duration)
org.apache.flink.streaming.api.datastream.DataStream.writeAsCsv(String) Please use the StreamingFileSink explicitly using the DataStream.addSink(SinkFunction) method.
org.apache.flink.streaming.api.datastream.DataStream.writeAsCsv(String, FileSystem.WriteMode) Please use the StreamingFileSink explicitly using the DataStream.addSink(SinkFunction) method.
org.apache.flink.streaming.api.datastream.DataStream.writeAsCsv(String, FileSystem.WriteMode, String, String) Please use the StreamingFileSink explicitly using the DataStream.addSink(SinkFunction) method.
org.apache.flink.streaming.api.datastream.DataStream.writeAsText(String) Please use the StreamingFileSink explicitly using the DataStream.addSink(SinkFunction) method.
org.apache.flink.streaming.api.datastream.DataStream.writeAsText(String, FileSystem.WriteMode) Please use the StreamingFileSink explicitly using the DataStream.addSink(SinkFunction) method.
org.apache.flink.streaming.api.datastream.DataStream.writeUsingOutputFormat(OutputFormat<T>) Please use the StreamingFileSink explicitly using the DataStream.addSink(SinkFunction) method.

Deprecated Constructors

Constructor and Description
org.apache.flink.streaming.api.functions.timestamps.BoundedOutOfOrdernessTimestampExtractor(Time) Use BoundedOutOfOrdernessTimestampExtractor(Duration)
org.apache.flink.streaming.api.datastream.JoinedStreams.WithWindow(DataStream<T1>, DataStream<T2>, KeySelector<T1, KEY>, KeySelector<T2, KEY>, TypeInformation<KEY>, WindowAssigner<? super CoGroupedStreams.TaggedUnion<T1, T2>, W>, Trigger<? super CoGroupedStreams.TaggedUnion<T1, T2>, ? super W>, Evictor<? super CoGroupedStreams.TaggedUnion<T1, T2>, ? super W>, Time) Use WithWindow(DataStream, DataStream, KeySelector, KeySelector, TypeInformation, WindowAssigner, Trigger, Evictor, Duration).
org.apache.flink.streaming.api.datastream.CoGroupedStreams.WithWindow(DataStream<T1>, DataStream<T2>, KeySelector<T1, KEY>, KeySelector<T2, KEY>, TypeInformation<KEY>, WindowAssigner<? super CoGroupedStreams.TaggedUnion<T1, T2>, W>, Trigger<? super CoGroupedStreams.TaggedUnion<T1, T2>, ? super W>, Evictor<? super CoGroupedStreams.TaggedUnion<T1, T2>, ? super W>, Time) Use WithWindow(DataStream, DataStream, KeySelector, KeySelector, TypeInformation, WindowAssigner, Trigger, Evictor, Duration)