Github user sraghunandan commented on a diff in the pull request:
https://github.com/apache/carbondata/pull/1841#discussion_r163139005
--- Diff:
streaming/src/main/scala/org/apache/carbondata/streaming/StreamHandoffRDD.scala
---
@@ -99,6 +101,38 @@ class StreamHandoffRDD[K, V](
formatter.format(new Date())
}
+ private def setTempDirLocation(split: Partition) : Unit = {
--- End diff --
SparkPartitionLoader, DataLoadProcessorStepOnSpark, CarbonFileFormat uses
array of locations to speed up the loading, can't use the same function to
achieve the functionality
---