nsivabalan commented on code in PR #5337:
URL: https://github.com/apache/hudi/pull/5337#discussion_r851655246
##########
hudi-spark-datasource/hudi-spark-common/src/main/scala/org/apache/hudi/BaseFileOnlyRelation.scala:
##########
@@ -84,21 +84,24 @@ class BaseFileOnlyRelation(sqlContext: SQLContext,
protected def collectFileSplits(partitionFilters: Seq[Expression],
dataFilters: Seq[Expression]): Seq[HoodieBaseFileSplit] = {
val partitions = listLatestBaseFiles(globPaths, partitionFilters,
dataFilters)
- val fileSplits = partitions.values.toSeq.flatMap { files =>
- files.flatMap { file =>
- // TODO move to adapter
- // TODO fix, currently assuming parquet as underlying format
- HoodieDataSourceHelper.splitFiles(
- sparkSession = sparkSession,
- file = file,
- // TODO clarify why this is required
- partitionValues = getPartitionColumnsAsInternalRow(file)
- )
+ val fileSplits = partitions.values.toSeq
+ .flatMap { files =>
+ files.flatMap { file =>
+ // TODO fix, currently assuming parquet as underlying format
+ HoodieDataSourceHelper.splitFiles(
+ sparkSession = sparkSession,
+ file = file,
+ partitionValues = getPartitionColumnsAsInternalRow(file)
+ )
+ }
}
- }
+ // NOTE: It's important to order the splits in the reverse order of their
+ // size so that we can subsequently bucket them in an efficient
manner
+ .sortBy(_.length)(implicitly[Ordering[Long]].reverse)
Review Comment:
I see within the spark adaptor we use next fit bin packing algo. If thats
never going to change, we can leave it here. if not, we can move the sorting
within adaptor.getFilePartitions(...)
--
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]
