public abstract class UnawareBucketSink<T> extends FlinkWriteSink<T>
Note: in unaware-bucket mode, we don't shuffle by bucket in inserting. We can assign compaction to the inserting jobs aside.
Modifier and Type | Field and Description |
---|---|
protected LogSinkFunction |
logSinkFunction |
protected Integer |
parallelism |
protected FileStoreTable |
table |
Constructor and Description |
---|
UnawareBucketSink(FileStoreTable table,
Map<String,String> overwritePartitions,
LogSinkFunction logSinkFunction,
Integer parallelism) |
Modifier and Type | Method and Description |
---|---|
org.apache.flink.streaming.api.datastream.DataStream<Committable> |
doWrite(org.apache.flink.streaming.api.datastream.DataStream<T> input,
String initialCommitUser,
Integer parallelism) |
createCommittableStateManager, createCommitterFactory
assertBatchAdaptiveParallelism, assertBatchAdaptiveParallelism, assertBatchAdaptiveParallelism, assertStreamingConfiguration, configureGlobalCommitter, createWriteOperator, doCommit, isStreaming, isStreaming, sinkFrom, sinkFrom
protected final FileStoreTable table
protected final LogSinkFunction logSinkFunction
public UnawareBucketSink(FileStoreTable table, @Nullable Map<String,String> overwritePartitions, LogSinkFunction logSinkFunction, @Nullable Integer parallelism)
Copyright © 2023–2024 The Apache Software Foundation. All rights reserved.