apache · nsivabalan · Feb 26, 2022 · Feb 21, 2022
diff --git a/hudi-client/hudi-client-common/src/main/java/org/apache/hudi/table/HoodieTable.java b/hudi-client/hudi-client-common/src/main/java/org/apache/hudi/table/HoodieTable.java
@@ -18,11 +18,6 @@
 
 package org.apache.hudi.table;
 
-import org.apache.avro.Schema;
-import org.apache.avro.specific.SpecificRecordBase;
-import org.apache.hadoop.conf.Configuration;
-import org.apache.hadoop.fs.FileSystem;
-import org.apache.hadoop.fs.Path;
 import org.apache.hudi.avro.HoodieAvroUtils;
 import org.apache.hudi.avro.model.HoodieCleanMetadata;
 import org.apache.hudi.avro.model.HoodieCleanerPlan;
@@ -59,6 +54,7 @@
 import org.apache.hudi.common.table.view.TableFileSystemView;
 import org.apache.hudi.common.table.view.TableFileSystemView.BaseFileOnlyView;
 import org.apache.hudi.common.table.view.TableFileSystemView.SliceView;
+import org.apache.hudi.common.util.Functions;
 import org.apache.hudi.common.util.Option;
 import org.apache.hudi.common.util.collection.Pair;
 import org.apache.hudi.config.HoodieWriteConfig;
@@ -75,10 +71,17 @@
 import org.apache.hudi.table.marker.WriteMarkersFactory;
 import org.apache.hudi.table.storage.HoodieLayoutFactory;
 import org.apache.hudi.table.storage.HoodieStorageLayout;
+
+import org.apache.avro.Schema;
+import org.apache.avro.specific.SpecificRecordBase;
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.fs.FileSystem;
+import org.apache.hadoop.fs.Path;
 import org.apache.log4j.LogManager;
 import org.apache.log4j.Logger;
 
 import javax.annotation.Nonnull;
+
 import java.io.IOException;
 import java.io.Serializable;
 import java.util.ArrayList;
@@ -776,8 +779,8 @@ public final Option<HoodieTableMetadataWriter> getMetadataWriter(String triggeri
    * @param triggeringInstantTimestamp - The instant that is triggering this metadata write
    * @return instance of {@link HoodieTableMetadataWriter}
    */
-  public <T extends SpecificRecordBase> Option<HoodieTableMetadataWriter> getMetadataWriter(String triggeringInstantTimestamp,
-                                                                                            Option<T> actionMetadata) {
+  public <R extends SpecificRecordBase> Option<HoodieTableMetadataWriter> getMetadataWriter(String triggeringInstantTimestamp,
+                                                                                            Option<R> actionMetadata) {
     // Each engine is expected to override this and
     // provide the actual metadata writer, if enabled.
     return Option.empty();
@@ -786,4 +789,8 @@ public <T extends SpecificRecordBase> Option<HoodieTableMetadataWriter> getMetad
   public HoodieTableMetadata getMetadataTable() {
     return this.metadata;
   }
+
+  public Runnable getPreExecuteRunnable() {
+    return Functions.noop();
+  }
 }
diff --git a/...hudi-flink-client/src/main/java/org/apache/hudi/table/action/commit/FlinkMergeHelper.java b/...hudi-flink-client/src/main/java/org/apache/hudi/table/action/commit/FlinkMergeHelper.java
@@ -91,7 +91,7 @@ public void runMerge(HoodieTable<T, List<HoodieRecord<T>>, List<HoodieKey>, List
 
       ThreadLocal<BinaryEncoder> encoderCache = new ThreadLocal<>();
       ThreadLocal<BinaryDecoder> decoderCache = new ThreadLocal<>();
-      wrapper = new BoundedInMemoryExecutor(table.getConfig().getWriteBufferLimitBytes(), new IteratorBasedQueueProducer<>(readerIterator),
+      wrapper = new BoundedInMemoryExecutor<>(table.getConfig().getWriteBufferLimitBytes(), new IteratorBasedQueueProducer<>(readerIterator),
           Option.of(new UpdateHandler(mergeHandle)), record -> {
         if (!externalSchemaTransformation) {
           return record;

diff --git a/...t/hudi-java-client/src/main/java/org/apache/hudi/table/action/commit/JavaMergeHelper.java b/...t/hudi-java-client/src/main/java/org/apache/hudi/table/action/commit/JavaMergeHelper.java
@@ -91,7 +91,7 @@ public void runMerge(HoodieTable<T, List<HoodieRecord<T>>, List<HoodieKey>, List
 
       ThreadLocal<BinaryEncoder> encoderCache = new ThreadLocal<>();
       ThreadLocal<BinaryDecoder> decoderCache = new ThreadLocal<>();
-      wrapper = new BoundedInMemoryExecutor(table.getConfig().getWriteBufferLimitBytes(), new IteratorBasedQueueProducer<>(readerIterator),
+      wrapper = new BoundedInMemoryExecutor<>(table.getConfig().getWriteBufferLimitBytes(), new IteratorBasedQueueProducer<>(readerIterator),
           Option.of(new UpdateHandler(mergeHandle)), record -> {
         if (!externalSchemaTransformation) {
           return record;

diff --git a/...di-spark-client/src/main/java/org/apache/hudi/execution/SparkBoundedInMemoryExecutor.java b/...di-spark-client/src/main/java/org/apache/hudi/execution/SparkBoundedInMemoryExecutor.java
diff --git a/...nt/hudi-spark-client/src/main/java/org/apache/hudi/execution/SparkLazyInsertIterable.java b/...nt/hudi-spark-client/src/main/java/org/apache/hudi/execution/SparkLazyInsertIterable.java
@@ -18,7 +18,6 @@
 
 package org.apache.hudi.execution;
 
-import org.apache.avro.Schema;
 import org.apache.hudi.avro.HoodieAvroUtils;
 import org.apache.hudi.client.WriteStatus;
 import org.apache.hudi.common.engine.TaskContextSupplier;
@@ -30,6 +29,8 @@
 import org.apache.hudi.io.WriteHandleFactory;
 import org.apache.hudi.table.HoodieTable;
 
+import org.apache.avro.Schema;
+
 import java.util.Iterator;
 import java.util.List;
 
@@ -84,8 +85,8 @@ protected List<WriteStatus> computeNext() {
         schema = HoodieAvroUtils.addMetadataFields(schema);
       }
       bufferedIteratorExecutor =
-          new SparkBoundedInMemoryExecutor<>(hoodieConfig, inputItr, getInsertHandler(),
-            getTransformFunction(schema, hoodieConfig));
+          new BoundedInMemoryExecutor<>(hoodieConfig.getWriteBufferLimitBytes(), inputItr, getInsertHandler(),
+              getTransformFunction(schema, hoodieConfig), hoodieTable.getPreExecuteRunnable());
       final List<WriteStatus> result = bufferedIteratorExecutor.execute();
       assert result != null && !result.isEmpty() && !bufferedIteratorExecutor.isRemaining();
       return result;

diff --git a/hudi-client/hudi-spark-client/src/main/java/org/apache/hudi/table/HoodieSparkTable.java b/hudi-client/hudi-spark-client/src/main/java/org/apache/hudi/table/HoodieSparkTable.java
@@ -18,8 +18,6 @@
 
 package org.apache.hudi.table;
 
-import org.apache.avro.specific.SpecificRecordBase;
-import org.apache.hadoop.fs.Path;
 import org.apache.hudi.client.WriteStatus;
 import org.apache.hudi.client.common.HoodieSparkEngineContext;
 import org.apache.hudi.common.data.HoodieData;
@@ -39,6 +37,11 @@
 import org.apache.hudi.metadata.HoodieTableMetadataWriter;
 import org.apache.hudi.metadata.SparkHoodieBackedTableMetadataWriter;
 import org.apache.hudi.table.action.HoodieWriteMetadata;
+
+import org.apache.avro.specific.SpecificRecordBase;
+import org.apache.hadoop.fs.Path;
+import org.apache.spark.TaskContext;
+import org.apache.spark.TaskContext$;
 import org.apache.spark.api.java.JavaRDD;
 
 import java.io.IOException;
@@ -111,8 +114,8 @@ protected HoodieIndex getIndex(HoodieWriteConfig config, HoodieEngineContext con
    * @return instance of {@link HoodieTableMetadataWriter}
    */
   @Override
-  public <T extends SpecificRecordBase> Option<HoodieTableMetadataWriter> getMetadataWriter(String triggeringInstantTimestamp,
-                                                                                            Option<T> actionMetadata) {
+  public <R extends SpecificRecordBase> Option<HoodieTableMetadataWriter> getMetadataWriter(String triggeringInstantTimestamp,
+                                                                                            Option<R> actionMetadata) {
     if (config.isMetadataTableEnabled()) {
       // Create the metadata table writer. First time after the upgrade this creation might trigger
       // metadata table bootstrapping. Bootstrapping process could fail and checking the table
@@ -132,4 +135,10 @@ public <T extends SpecificRecordBase> Option<HoodieTableMetadataWriter> getMetad
 
     return Option.empty();
   }
+
+  @Override
+  public Runnable getPreExecuteRunnable() {
+    final TaskContext taskContext = TaskContext.get();
+    return () -> TaskContext$.MODULE$.setTaskContext(taskContext);
+  }
 }
diff --git a/...ent/src/main/java/org/apache/hudi/table/action/bootstrap/OrcBootstrapMetadataHandler.java b/...ent/src/main/java/org/apache/hudi/table/action/bootstrap/OrcBootstrapMetadataHandler.java
@@ -29,7 +29,6 @@
 import org.apache.hudi.common.util.queue.BoundedInMemoryExecutor;
 import org.apache.hudi.config.HoodieWriteConfig;
 import org.apache.hudi.exception.HoodieException;
-import org.apache.hudi.execution.SparkBoundedInMemoryExecutor;
 import org.apache.hudi.io.HoodieBootstrapHandle;
 import org.apache.hudi.keygen.KeyGeneratorInterface;
 import org.apache.hudi.table.HoodieTable;
@@ -68,15 +67,15 @@ void executeBootstrap(HoodieBootstrapHandle<?, ?, ?, ?> bootstrapHandle, Path so
     Reader orcReader = OrcFile.createReader(sourceFilePath, OrcFile.readerOptions(table.getHadoopConf()));
     TypeDescription orcSchema = orcReader.getSchema();
     try (RecordReader reader = orcReader.rows(new Reader.Options(table.getHadoopConf()).schema(orcSchema))) {
-      wrapper = new SparkBoundedInMemoryExecutor<GenericRecord, HoodieRecord, Void>(config,
+      wrapper = new BoundedInMemoryExecutor<GenericRecord, HoodieRecord, Void>(config.getWriteBufferLimitBytes(),
           new OrcReaderIterator(reader, avroSchema, orcSchema), new BootstrapRecordConsumer(bootstrapHandle), inp -> {
         String recKey = keyGenerator.getKey(inp).getRecordKey();
         GenericRecord gr = new GenericData.Record(HoodieAvroUtils.RECORD_KEY_SCHEMA);
         gr.put(HoodieRecord.RECORD_KEY_METADATA_FIELD, recKey);
         BootstrapRecordPayload payload = new BootstrapRecordPayload(gr);
         HoodieRecord rec = new HoodieAvroRecord(new HoodieKey(recKey, partitionPath), payload);
         return rec;
-      });
+      }, table.getPreExecuteRunnable());
       wrapper.execute();
     } catch (Exception e) {
       throw new HoodieException(e);

diff --git a/...src/main/java/org/apache/hudi/table/action/bootstrap/ParquetBootstrapMetadataHandler.java b/...src/main/java/org/apache/hudi/table/action/bootstrap/ParquetBootstrapMetadataHandler.java
@@ -28,7 +28,6 @@
 import org.apache.hudi.common.util.queue.BoundedInMemoryExecutor;
 import org.apache.hudi.config.HoodieWriteConfig;
 import org.apache.hudi.exception.HoodieException;
-import org.apache.hudi.execution.SparkBoundedInMemoryExecutor;
 import org.apache.hudi.io.HoodieBootstrapHandle;
 import org.apache.hudi.keygen.KeyGeneratorInterface;
 import org.apache.hudi.table.HoodieTable;
@@ -72,15 +71,15 @@ void executeBootstrap(HoodieBootstrapHandle<?, ?, ?, ?> bootstrapHandle,
     try {
       ParquetReader<IndexedRecord> reader =
           AvroParquetReader.<IndexedRecord>builder(sourceFilePath).withConf(table.getHadoopConf()).build();
-      wrapper = new SparkBoundedInMemoryExecutor<GenericRecord, HoodieRecord, Void>(config,
+      wrapper = new BoundedInMemoryExecutor<GenericRecord, HoodieRecord, Void>(config.getWriteBufferLimitBytes(),
           new ParquetReaderIterator(reader), new BootstrapRecordConsumer(bootstrapHandle), inp -> {
         String recKey = keyGenerator.getKey(inp).getRecordKey();
         GenericRecord gr = new GenericData.Record(HoodieAvroUtils.RECORD_KEY_SCHEMA);
         gr.put(HoodieRecord.RECORD_KEY_METADATA_FIELD, recKey);
         BootstrapRecordPayload payload = new BootstrapRecordPayload(gr);
         HoodieRecord rec = new HoodieAvroRecord(new HoodieKey(recKey, partitionPath), payload);
         return rec;
-      });
+      }, table.getPreExecuteRunnable());
       wrapper.execute();
     } catch (Exception e) {
       throw new HoodieException(e);

diff --git a/...hudi-spark-client/src/main/java/org/apache/hudi/table/action/commit/SparkMergeHelper.java b/...hudi-spark-client/src/main/java/org/apache/hudi/table/action/commit/SparkMergeHelper.java
@@ -25,7 +25,6 @@
 import org.apache.hudi.common.model.HoodieRecordPayload;
 import org.apache.hudi.common.util.queue.BoundedInMemoryExecutor;
 import org.apache.hudi.exception.HoodieException;
-import org.apache.hudi.execution.SparkBoundedInMemoryExecutor;
 import org.apache.hudi.io.HoodieMergeHandle;
 import org.apache.hudi.io.storage.HoodieFileReader;
 import org.apache.hudi.io.storage.HoodieFileReaderFactory;
@@ -90,13 +89,13 @@ public void runMerge(HoodieTable<T, JavaRDD<HoodieRecord<T>>, JavaRDD<HoodieKey>
 
       ThreadLocal<BinaryEncoder> encoderCache = new ThreadLocal<>();
       ThreadLocal<BinaryDecoder> decoderCache = new ThreadLocal<>();
-      wrapper = new SparkBoundedInMemoryExecutor(table.getConfig(), readerIterator,
+      wrapper = new BoundedInMemoryExecutor(table.getConfig().getWriteBufferLimitBytes(), readerIterator,
           new UpdateHandler(mergeHandle), record -> {
         if (!externalSchemaTransformation) {
           return record;
         }
         return transformRecordBasedOnNewSchema(gReader, gWriter, encoderCache, decoderCache, (GenericRecord) record);
-      });
+      }, table.getPreExecuteRunnable());
       wrapper.execute();
     } catch (Exception e) {
       throw new HoodieException(e);

diff --git a/...ion/TestSparkBoundedInMemoryExecutor.java → ...n/TestBoundedInMemoryExecutorInSpark.java b/...ion/TestSparkBoundedInMemoryExecutor.java → ...n/TestBoundedInMemoryExecutorInSpark.java
@@ -22,12 +22,15 @@
 import org.apache.hudi.common.table.timeline.HoodieActiveTimeline;
 import org.apache.hudi.common.testutils.HoodieTestDataGenerator;
 import org.apache.hudi.common.util.Option;
+import org.apache.hudi.common.util.queue.BoundedInMemoryExecutor;
 import org.apache.hudi.common.util.queue.BoundedInMemoryQueueConsumer;
 import org.apache.hudi.config.HoodieWriteConfig;
 import org.apache.hudi.exception.HoodieException;
 import org.apache.hudi.testutils.HoodieClientTestHarness;
 
 import org.apache.avro.generic.IndexedRecord;
+import org.apache.spark.TaskContext;
+import org.apache.spark.TaskContext$;
 import org.junit.jupiter.api.AfterEach;
 import org.junit.jupiter.api.BeforeEach;
 import org.junit.jupiter.api.Test;
@@ -44,7 +47,7 @@
 import static org.mockito.Mockito.mock;
 import static org.mockito.Mockito.when;
 
-public class TestSparkBoundedInMemoryExecutor extends HoodieClientTestHarness {
+public class TestBoundedInMemoryExecutorInSpark extends HoodieClientTestHarness {
 
   private final String instantTime = HoodieActiveTimeline.createNewInstantTime();
 
@@ -58,6 +61,11 @@ public void tearDown() throws Exception {
     cleanupResources();
   }
 
+  private Runnable getPreExecuteRunnable() {
+    final TaskContext taskContext = TaskContext.get();
+    return () -> TaskContext$.MODULE$.setTaskContext(taskContext);
+  }
+
   @Test
   public void testExecutor() {
 
@@ -85,10 +93,10 @@ protected Integer getResult() {
           }
         };
 
-    SparkBoundedInMemoryExecutor<HoodieRecord, Tuple2<HoodieRecord, Option<IndexedRecord>>, Integer> executor = null;
+    BoundedInMemoryExecutor<HoodieRecord, Tuple2<HoodieRecord, Option<IndexedRecord>>, Integer> executor = null;
     try {
-      executor = new SparkBoundedInMemoryExecutor(hoodieWriteConfig, hoodieRecords.iterator(), consumer,
-          getTransformFunction(HoodieTestDataGenerator.AVRO_SCHEMA));
+      executor = new BoundedInMemoryExecutor(hoodieWriteConfig.getWriteBufferLimitBytes(), hoodieRecords.iterator(), consumer,
+          getTransformFunction(HoodieTestDataGenerator.AVRO_SCHEMA), getPreExecuteRunnable());
       int result = executor.execute();
       // It should buffer and write 100 records
       assertEquals(100, result);
@@ -131,11 +139,11 @@ protected Integer getResult() {
           }
         };
 
-    SparkBoundedInMemoryExecutor<HoodieRecord, Tuple2<HoodieRecord, Option<IndexedRecord>>, Integer> executor = null;
+    BoundedInMemoryExecutor<HoodieRecord, Tuple2<HoodieRecord, Option<IndexedRecord>>, Integer> executor = null;
     try {
-      executor = new SparkBoundedInMemoryExecutor(hoodieWriteConfig, hoodieRecords.iterator(), consumer,
-          getTransformFunction(HoodieTestDataGenerator.AVRO_SCHEMA));
-      SparkBoundedInMemoryExecutor<HoodieRecord, Tuple2<HoodieRecord, Option<IndexedRecord>>, Integer> finalExecutor = executor;
+      executor = new BoundedInMemoryExecutor(hoodieWriteConfig.getWriteBufferLimitBytes(), hoodieRecords.iterator(), consumer,
+          getTransformFunction(HoodieTestDataGenerator.AVRO_SCHEMA), getPreExecuteRunnable());
+      BoundedInMemoryExecutor<HoodieRecord, Tuple2<HoodieRecord, Option<IndexedRecord>>, Integer> finalExecutor = executor;
 
       Thread.currentThread().interrupt();
 

diff --git a/hudi-common/src/main/java/org/apache/hudi/common/util/Functions.java b/hudi-common/src/main/java/org/apache/hudi/common/util/Functions.java
@@ -25,6 +25,11 @@
  */
 public interface Functions {
 
+  static Runnable noop() {
+    return () -> {
+    };
+  }
+
   /**
    * A function which has not any parameter.
    */