Skip to content

[SPARK-29552][SQL] Execute the "OptimizeLocalShuffleReader" rule when creating new query stage and then can optimize the shuffle reader to local shuffle reader as much as possible. #26207

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
JkSelf wants to merge 3 commits into from
Closed

[SPARK-29552][SQL] Execute the "OptimizeLocalShuffleReader" rule when creating new query stage and then can optimize the shuffle reader to local shuffle reader as much as possible. #26207

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
997e994
fix the flaky test in multi-joins with local shuffle reader
JkSelf Oct 22, 2019
1bc418e
update the comments
JkSelf Oct 23, 2019
b372636
polish the comment for why execute the OptimizaLocalShuffleReader rul…
JkSelf Oct 23, 2019
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
Next Next commit
fix the flaky test in multi-joins with local shuffle reader
  • Loading branch information
@JkSelf
JkSelf committed Oct 22, 2019
commit 997e994d322dd334c4574d6b42811df024fac019
3 changes: 3 additions & 0 deletions sql/core/src/main/scala/org/apache/spark/sql/execution/adaptive/AdaptiveSparkPlanExec.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -92,6 +92,9 @@ case class AdaptiveSparkPlanExec(
// optimizations should be stage-independent.
@transient private val queryStageOptimizerRules: Seq[Rule[SparkPlan]] = Seq(
ReuseAdaptiveSubquery(conf, subqueryCache),
// Here we need put the OptimizeLocalShuffleReader rule before
// ReduceNumShufflePartitions rule to avoid the further optimizaiton.
Copy link
Contributor

@cloud-fan cloud-fan Oct 22, 2019

Choose a reason for hiding this comment

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

I think the comment needs to explain 2 things:

  1. why execute this rule twice
  2. why it must be run before OptimizeLocalShuffleReader

Copy link
Contributor Author

@JkSelf JkSelf Oct 23, 2019

Choose a reason for hiding this comment

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

Updated.

OptimizeLocalShuffleReader(conf),
ReduceNumShufflePartitions(conf),
ApplyColumnarRulesAndInsertTransitions(session.sessionState.conf,
session.sessionState.columnarRules),
Expand Down
2 changes: 1 addition & 1 deletion sql/core/src/test/scala/org/apache/spark/sql/execution/adaptive/AdaptiveQueryExecSuite.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -164,7 +164,7 @@ class AdaptiveQueryExecSuite
val bhj = findTopLevelBroadcastHashJoin(adaptivePlan)
assert(bhj.size == 3)
// additional shuffle exchange introduced, only one shuffle reader to local shuffle reader.
checkNumLocalShuffleReaders(adaptivePlan, 1)
checkNumLocalShuffleReaders(adaptivePlan, 2)
Copy link
Contributor

@HeartSaVioR HeartSaVioR Oct 22, 2019 •
edited
Loading

Choose a reason for hiding this comment

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

Just to confirm, would the change make this value consistently be 2? Because the value has changed to 2 but the value was actually flaky (neither 1 or 2 consistently) depending on the situation/randomness (maybe).

You may want to run the same for what I've discovered, 1) solely in local dev, 2) test suite in local dev, 3) trigger CI for 5 times or alike.

Copy link
Contributor Author

@JkSelf JkSelf Oct 23, 2019

Choose a reason for hiding this comment

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

@HeartSaVioR With this patch, the value will consistently be 2. Because we already optimize all the possible local shuffle reader. And I have run in local dev and also the test suite, the value are all 2. Thanks.

Copy link
Contributor

@HeartSaVioR HeartSaVioR Oct 23, 2019

Choose a reason for hiding this comment

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

OK thanks for confirming.

}
}

Expand Down