nsivabalan commented on a change in pull request #4789:
URL: https://github.com/apache/hudi/pull/4789#discussion_r812949213
##########
File path:
hudi-spark-datasource/hudi-spark-common/src/main/scala/org/apache/hudi/HoodieMergeOnReadRDD.scala
##########
@@ -314,35 +309,65 @@ class HoodieMergeOnReadRDD(@transient sc: SparkContext,
private def mergeRowWithLog(curRow: InternalRow, curKey: String) = {
val historyAvroRecord =
serializer.serialize(curRow).asInstanceOf[GenericRecord]
- logRecords.get(curKey).getData.combineAndGetUpdateValue(
- historyAvroRecord, tableAvroSchema, payloadProps)
+ logRecords.get(curKey).getData
+ .combineAndGetUpdateValue(historyAvroRecord, tableAvroSchema,
payloadProps)
}
}
- }
+}
private object HoodieMergeOnReadRDD {
val CONFIG_INSTANTIATION_LOCK = new Object()
def scanLog(split: HoodieMergeOnReadFileSplit, logSchema: Schema, config:
Configuration): HoodieMergedLogRecordScanner = {
val fs = FSUtils.getFs(split.tablePath, config)
- HoodieMergedLogRecordScanner.newBuilder()
- .withFileSystem(fs)
- .withBasePath(split.tablePath)
- .withLogFilePaths(split.logPaths.get.asJava)
- .withReaderSchema(logSchema)
- .withLatestInstantTime(split.latestCommit)
- .withReadBlocksLazily(
-
Try(config.get(HoodieRealtimeConfig.COMPACTION_LAZY_BLOCK_READ_ENABLED_PROP,
-
HoodieRealtimeConfig.DEFAULT_COMPACTION_LAZY_BLOCK_READ_ENABLED).toBoolean)
- .getOrElse(false))
- .withReverseReader(false)
- .withBufferSize(
- config.getInt(HoodieRealtimeConfig.MAX_DFS_STREAM_BUFFER_SIZE_PROP,
- HoodieRealtimeConfig.DEFAULT_MAX_DFS_STREAM_BUFFER_SIZE))
- .withMaxMemorySizeInBytes(split.maxCompactionMemoryInBytes)
- .withSpillableMapBasePath(
- config.get(HoodieRealtimeConfig.SPILLABLE_MAP_BASE_PATH_PROP,
- HoodieRealtimeConfig.DEFAULT_SPILLABLE_MAP_BASE_PATH))
- .build()
+
+ if (HoodieTableMetadata.isMetadataTable(split.tablePath)) {
+ val metadataConfig =
HoodieMetadataConfig.newBuilder().enable(true).build()
Review comment:
good catch. may be we need to instantiate using props. But this might
need fixes in other read code paths as well. For eg, HoodieInputFormatUtils.java
##########
File path:
hudi-common/src/main/java/org/apache/hudi/common/table/TableSchemaResolver.java
##########
@@ -526,6 +526,7 @@ private boolean hasOperationField() {
Schema tableAvroSchema = getTableAvroSchemaFromDataFile();
return tableAvroSchema.getField(HoodieRecord.OPERATION_METADATA_FIELD)
!= null;
} catch (Exception e) {
+ LOG.info(String.format("Failed to read operation field from avro schema
(%s)", e.getMessage()));
Review comment:
can we make this error log level instead of info?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
