1
0

[HUDI-3838] Implemented drop partition column feature for delta streamer code path (#5294)

* [HUDI-3838] Implemented drop partition column feature for delta streamer code path

* Ensure drop partition table config is updated in hoodie.props

Co-authored-by: Sagar Sumit <sagarsumit09@gmail.com>
This commit is contained in:
Vinoth Govindarajan
2022-04-12 05:40:30 -07:00
committed by GitHub
parent 101b82a679
commit d16740976e
3 changed files with 70 additions and 2 deletions

View File

@@ -105,6 +105,7 @@ import java.util.stream.Collectors;
import scala.collection.JavaConversions;
import static org.apache.hudi.common.table.HoodieTableConfig.ARCHIVELOG_FOLDER;
import static org.apache.hudi.common.table.HoodieTableConfig.DROP_PARTITION_COLUMNS;
import static org.apache.hudi.config.HoodieClusteringConfig.ASYNC_CLUSTERING_ENABLE;
import static org.apache.hudi.config.HoodieClusteringConfig.INLINE_CLUSTERING;
import static org.apache.hudi.config.HoodieCompactionConfig.INLINE_COMPACT;
@@ -280,6 +281,7 @@ public class DeltaSync implements Serializable {
.setPreCombineField(cfg.sourceOrderingField)
.setPartitionMetafileUseBaseFormat(props.getBoolean(HoodieTableConfig.PARTITION_METAFILE_USE_BASE_FORMAT.key(),
HoodieTableConfig.PARTITION_METAFILE_USE_BASE_FORMAT.defaultValue()))
.setDropPartitionColumnsWhenWrite(isDropPartitionColumns())
.initTable(new Configuration(jssc.hadoopConfiguration()),
cfg.targetBasePath);
}
@@ -375,6 +377,7 @@ public class DeltaSync implements Serializable {
SimpleKeyGenerator.class.getName()))
.setPartitionMetafileUseBaseFormat(props.getBoolean(HoodieTableConfig.PARTITION_METAFILE_USE_BASE_FORMAT.key(),
HoodieTableConfig.PARTITION_METAFILE_USE_BASE_FORMAT.defaultValue()))
.setDropPartitionColumnsWhenWrite(isDropPartitionColumns())
.initTable(new Configuration(jssc.hadoopConfiguration()), cfg.targetBasePath);
}
@@ -478,13 +481,14 @@ public class DeltaSync implements Serializable {
boolean shouldCombine = cfg.filterDupes || cfg.operation.equals(WriteOperationType.UPSERT);
JavaRDD<GenericRecord> avroRDD = avroRDDOptional.get();
JavaRDD<HoodieRecord> records = avroRDD.map(gr -> {
JavaRDD<HoodieRecord> records = avroRDD.map(record -> {
GenericRecord gr = isDropPartitionColumns() ? HoodieAvroUtils.removeFields(record, getPartitionColumns(keyGenerator, props)) : record;
HoodieRecordPayload payload = shouldCombine ? DataSourceUtils.createPayload(cfg.payloadClassName, gr,
(Comparable) HoodieAvroUtils.getNestedFieldVal(gr, cfg.sourceOrderingField, false, props.getBoolean(
KeyGeneratorOptions.KEYGENERATOR_CONSISTENT_LOGICAL_TIMESTAMP_ENABLED.key(),
Boolean.parseBoolean(KeyGeneratorOptions.KEYGENERATOR_CONSISTENT_LOGICAL_TIMESTAMP_ENABLED.defaultValue()))))
: DataSourceUtils.createPayload(cfg.payloadClassName, gr);
return new HoodieAvroRecord<>(keyGenerator.getKey(gr), payload);
return new HoodieAvroRecord<>(keyGenerator.getKey(record), payload);
});
return Pair.of(schemaProvider, Pair.of(checkpointStr, records));
@@ -727,6 +731,9 @@ public class DeltaSync implements Serializable {
private void reInitWriteClient(Schema sourceSchema, Schema targetSchema) throws IOException {
LOG.info("Setting up new Hoodie Write Client");
if (isDropPartitionColumns()) {
targetSchema = HoodieAvroUtils.removeFields(targetSchema, getPartitionColumns(keyGenerator, props));
}
registerAvroSchemas(sourceSchema, targetSchema);
HoodieWriteConfig hoodieCfg = getHoodieClientConfig(targetSchema);
if (hoodieCfg.isEmbeddedTimelineServerEnabled()) {
@@ -898,4 +905,24 @@ public class DeltaSync implements Serializable {
return Option.empty();
}
}
/**
* Set based on hoodie.datasource.write.drop.partition.columns config.
* When set to true, will not write the partition columns into the table.
*/
private Boolean isDropPartitionColumns() {
return props.getBoolean(DROP_PARTITION_COLUMNS.key(), DROP_PARTITION_COLUMNS.defaultValue());
}
/**
* Get the list of partition columns as a list of strings.
*
* @param keyGenerator KeyGenerator
* @param props TypedProperties
* @return List of partition columns.
*/
private List<String> getPartitionColumns(KeyGenerator keyGenerator, TypedProperties props) {
String partitionColumns = HoodieSparkUtils.getPartitionColumns(keyGenerator, props);
return Arrays.asList(partitionColumns.split(","));
}
}