Github user JoshRosen commented on a diff in the pull request:
https://github.com/apache/spark/pull/9276#discussion_r43719127
--- Diff:
sql/core/src/main/scala/org/apache/spark/sql/execution/Exchange.scala ---
@@ -209,6 +299,98 @@ private[sql] case class EnsureRequirements(sqlContext:
SQLContext) extends Rule[
}
}
+ /**
+ * Adds [[ExchangeCoordinator]] to [[Exchange]]s if adaptive query
execution is enabled
+ * and partitioning schemes of these [[Exchange]]s support
[[ExchangeCoordinator]].
+ */
+ private def withExchangeCoordinator(
+ children: Seq[SparkPlan],
+ requiredChildDistributions: Seq[Distribution]): Seq[SparkPlan] = {
+ val supportsCoordinator =
+ if (children.exists(_.isInstanceOf[Exchange])) {
+ // Right now, ExchangeCoordinator only support HashPartitionings.
+ children.forall {
+ case e @ Exchange(hash: HashPartitioning, _, _) => true
+ case child =>
+ child.outputPartitioning match {
+ case hash: HashPartitioning => true
+ case collection: PartitioningCollection =>
+
collection.partitionings.exists(_.isInstanceOf[HashPartitioning])
--- End diff --
Is it possible for a `PartitioningCollection` to contain partitioning of
different classes (e.g. a hash partitioning _and_ a range partitioning)? I
don't think so, so maybe it'd be clearer to use `forall` instead of `exists`?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
