Skip to content

[HUDI-7118] set conf 'spark.sql.parquet.enableVectorizedReader' to true automatically only if the value is not explicitly set #10134

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
danny0405 merged 2 commits into from
Nov 21, 2023
Merged

[HUDI-7118] set conf 'spark.sql.parquet.enableVectorizedReader' to true automatically only if the value is not explicitly set #10134

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
ed3c70d
[HUDI-7118] set conf 'spark.sql.parquet.enableVectorizedReader' to tr…
beyond1920 Nov 17, 2023
557ed31
Simplify code because the default value of enableVectorizedReader is …
beyond1920 Nov 18, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 0 additions & 8 deletions ...rk-datasource/hudi-spark-common/src/main/scala/org/apache/hudi/BaseFileOnlyRelation.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -77,14 +77,6 @@ case class BaseFileOnlyRelation(override val sqlContext: SQLContext,
override def updatePrunedDataSchema(prunedSchema: StructType): Relation =
this.copy(prunedDataSchema = Some(prunedSchema))

override def imbueConfigs(sqlContext: SQLContext): Unit = {
super.imbueConfigs(sqlContext)
// TODO Issue with setting this to true in spark 332
if (HoodieSparkUtils.gteqSpark3_4 || !HoodieSparkUtils.gteqSpark3_3_2) {
sqlContext.sparkSession.sessionState.conf.setConfString("spark.sql.parquet.enableVectorizedReader", "true")
Copy link
Contributor

@xiarixiaoyao xiarixiaoyao Nov 18, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

should not use contains.
spark always inject this config.

The reason why this place is forcibly set to true is because the parent class of BaseFileOnlyRelationship previously forced set spark.sql.parquet.enableVectorizedReader=false. users unable to turn on vectorization, resulting in a significant decrease in query performance. So at that time, it was forcibly set to true in this location.

however https://issues.apache.org/jira/browse/HUDI-3639 This PR removes the behavior of the parent class of BaseFileOnlyRelationship.
Therefore, we no longer need to force set spark.sql.parquet.enableVectorizedReader=true in BaseFileOnlyRelationship; Whether to enable vectorization is left to the user for decision

Copy link
Contributor

@xiarixiaoyao xiarixiaoyao Nov 18, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

just remove line 83. to line 85

Copy link
Contributor Author

@beyond1920 beyond1920 Nov 18, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for reminder. Updated.

}
}

protected override def composeRDD(fileSplits: Seq[HoodieBaseFileSplit],
tableSchema: HoodieTableSchema,
requiredSchema: HoodieTableSchema,
Expand Down
6 changes: 0 additions & 6 deletions ...asource/hudi-spark-common/src/main/scala/org/apache/hudi/HoodieBootstrapMORRelation.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -67,12 +67,6 @@ case class HoodieBootstrapMORRelation(override val sqlContext: SQLContext,

override lazy val mandatoryFields: Seq[String] = mandatoryFieldsForMerging

override def imbueConfigs(sqlContext: SQLContext): Unit = {
super.imbueConfigs(sqlContext)
sqlContext.sparkSession.sessionState.conf.setConfString("spark.sql.parquet.enableVectorizedReader", "true")
}


protected override def getFileSlices(partitionFilters: Seq[Expression], dataFilters: Seq[Expression]): Seq[FileSlice] = {
if (globPaths.isEmpty) {
fileIndex.listFileSlices(HoodieFileIndex.
Expand Down