Skip to content

[SPARK-31226][CORE][TESTS] SizeBasedCoalesce logic will lose partition #27988

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
AngersZhuuuu wants to merge 4 commits into from
Closed

[SPARK-31226][CORE][TESTS] SizeBasedCoalesce logic will lose partition #27988

Changes from 2 commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
007f49c
Update RDDSuite.scala
AngersZhuuuu Mar 23, 2020
2c4ef14
Update RDDSuite.scala
AngersZhuuuu Mar 23, 2020
d7764ac
Update RDDSuite.scala
AngersZhuuuu Jul 10, 2020
20b1183
Update RDDSuite.scala
AngersZhuuuu Jul 10, 2020
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
14 changes: 7 additions & 7 deletions core/src/test/scala/org/apache/spark/rdd/RDDSuite.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1298,18 +1298,18 @@ class SizeBasedCoalescer(val maxSize: Int) extends PartitionCoalescer with Seria
val splitSize = fileSplit.getLength
if (currentSum + splitSize < maxSize) {
addPartition(partition, splitSize)
index += 1
if (index == partitions.size) {
updateGroups
}
} else {
if (currentGroup.partitions.size == 0) {
if (currentGroup.partitions.isEmpty) {
addPartition(partition, splitSize)
index += 1
} else {
updateGroups

@cloud-fan cloud-fan Jul 9, 2020

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

can you point out which line/lines cause the bug?

@AngersZhuuuu AngersZhuuuu Jul 10, 2020

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

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

can you point out which line/lines cause the bug?

In the pr description I have show the case this bug happen. Not one line or some lines cause this error.

@AngersZhuuuu AngersZhuuuu Jul 10, 2020

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

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

can you point out which line/lines cause the bug?

Change the desc make it more clear.

updateGroups()
addPartition(partition, splitSize)
}

@dongjoon-hyun dongjoon-hyun Jul 10, 2020

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

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

Instead of the following,

if (currentGroup.partitions.isEmpty) {
  addPartition(partition, splitSize)
} else {
  updateGroups()
  addPartition(partition, splitSize)
}

The following will be better.

if (currentGroup.partitions.nonEmpty) {
  updateGroups()
}
addPartition(partition, splitSize)

@AngersZhuuuu AngersZhuuuu Jul 10, 2020

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

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

Updated, thanks

}
index += 1
if (index == partitions.length) {
updateGroups()

@cloud-fan cloud-fan Jul 10, 2020

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

nit: can we do this after the loop exit? There is no early stop in this loop.

@AngersZhuuuu AngersZhuuuu Jul 10, 2020

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

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

nit: can we do this after the loop exit? There is no early stop in this loop.

Sure, updated, since before exit loop, the last action is addPartition so currentGourps if not empty, we don't need to check if currentGroup is empty, just updateGroups

}

@dongjoon-hyun dongjoon-hyun Apr 1, 2020

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

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

BTW, the previous code was added at 2.0.0. Did you hit a test flakiness due to this?

@AngersZhuuuu AngersZhuuuu Apr 2, 2020 •
edited
Loading

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

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

BTW, the previous code was added at 2.0.0. Did you hit a test flakiness due to this?

Yeah, these days I am testing a way to fix spark small out put file problem, and I use this logic, during test, found that this logic is wrong.
#27248

}
groups.toArray
}
Expand Down