Skip to content

[SPARK-52767][SQL] Optimize maxRows and maxRowsPerPartition for join and union #51451

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

Closed
zml1206 wants to merge 1 commit into from
Closed

[SPARK-52767][SQL] Optimize maxRows and maxRowsPerPartition for join and union #51451

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
44 changes: 21 additions & 23 deletions ...st/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/basicLogicalOperators.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -580,39 +580,37 @@ case class Union(
allowMissingCol: Boolean = false) extends UnionBase {
assert(!allowMissingCol || byName, "`allowMissingCol` can be true only if `byName` is true.")

override def maxRows: Option[Long] = {
var sum = BigInt(0)
children.foreach { child =>
if (child.maxRows.isDefined) {
sum += child.maxRows.get
if (!sum.isValidLong) {
return None
override lazy val maxRows: Option[Long] = {
Comment thread
cloud-fan marked this conversation as resolved.
val sum = children.foldLeft(Option(BigInt(0))) {
case (Some(acc), child) =>
child.maxRows match {
case Some(n) =>
val newSum = acc + n
if (newSum.isValidLong) Some(newSum) else None
case None => None
}
} else {
return None
}
case (None, _) => None
}
Some(sum.toLong)
sum.map(_.toLong)
}

final override val nodePatterns: Seq[TreePattern] = Seq(UNION)

/**
* Note the definition has assumption about how union is implemented physically.
*/
override def maxRowsPerPartition: Option[Long] = {
var sum = BigInt(0)
children.foreach { child =>
if (child.maxRowsPerPartition.isDefined) {
sum += child.maxRowsPerPartition.get
if (!sum.isValidLong) {
return None
override lazy val maxRowsPerPartition: Option[Long] = {
val sum = children.foldLeft(Option(BigInt(0))) {
case (Some(acc), child) =>
child.maxRowsPerPartition match {
case Some(n) =>
val newSum = acc + n
if (newSum.isValidLong) Some(newSum) else None
case None => None
}
} else {
return None
}
case (None, _) => None
}
Some(sum.toLong)
sum.map(_.toLong)
}

private def duplicatesResolvedPerBranch: Boolean =
Expand Down Expand Up @@ -666,7 +664,7 @@ case class Join(
hint: JoinHint)
extends BinaryNode with PredicateHelper {

override def maxRows: Option[Long] = {
override lazy val maxRows: Option[Long] = {
joinType match {
case Inner | Cross | FullOuter | LeftOuter | RightOuter | LeftSingle
if left.maxRows.isDefined && right.maxRows.isDefined =>
Expand Down