lokesh-lingarajan-0310 commented on code in PR #9852:
URL: https://github.com/apache/hudi/pull/9852#discussion_r1366100184
##########
hudi-utilities/src/main/java/org/apache/hudi/utilities/streamer/StreamSync.java:
##########
@@ -597,16 +578,56 @@ private Pair<SchemaProvider, Pair<String,
JavaRDD<HoodieRecord>>> fetchFromSourc
(SchemaProvider) new DelegatingSchemaProvider(props,
hoodieSparkContext.jsc(), dataAndCheckpoint.getSchemaProvider(),
new SimpleSchemaProvider(hoodieSparkContext.jsc(),
targetSchema, props)))
.orElse(dataAndCheckpoint.getSchemaProvider());
- // Rewrite transformed records into the expected target schema
- avroRDDOptional = transformed.map(t -> getTransformedRDD(t,
reconcileSchema, schemaProvider.getTargetSchema()));
+ }
+
+ // Short circuit for bulk insert to gain performance benefits of row
writing
+ if (recordType == HoodieRecordType.SPARK) {
+ rowDatasetOptional = transformed;
+ noNewData = (!rowDatasetOptional.isPresent()) ||
(rowDatasetOptional.get().isEmpty());
+ } else {
+ // If the target schema is specified through Avro schema,
+ // pass in the schema for the Row-to-Avro conversion
+ // to avoid nullability mismatch between Avro schema and Row schema
+ if (errorTableWriter.isPresent()
Review Comment:
good catch Siva
--
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]
