Skip to content

chore: Reserve memory for native shuffle writer per partition #1022

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

Jump to bottom
Merged
andygrove merged 5 commits into from
Oct 19, 2024
Merged

chore: Reserve memory for native shuffle writer per partition #1022

andygrove merged 5 commits into from
Oct 19, 2024

Conversation

@viirya
Copy link
Member

@viirya viirya commented Oct 16, 2024 •
edited
Loading

Which issue does this PR close?

Closes #1019.

Rationale for this change

This restore the patch merged in #988. The patch causes the issue #1019. This patch includes a fix for that.

What changes are included in this PR?

How are these changes tested?

Manually run TPCH benchmark locally.

@viirya viirya force-pushed the revise_shuffle_memory_fix branch from ff3b262 to 851427f Compare October 16, 2024 20:11
@viirya viirya force-pushed the revise_shuffle_memory_fix branch from 72c7a0c to 2d5478a Compare October 17, 2024 19:53
@viirya viirya requested a review from andygrove October 17, 2024 21:15
@andygrove
Copy link
Member

andygrove commented Oct 18, 2024

I am testing this PR out now with benchmarks.

@andygrove
Copy link
Member

andygrove commented Oct 18, 2024

I am testing with TPC-H sf=100. I usually test with one executor and 8 cores, but with this PR I can only run with a single core. I tried with 2 cores with this config:

    --conf spark.executor.instances=1 \
    --conf spark.executor.memory=16G \
    --conf spark.executor.cores=2 \
    --conf spark.cores.max=2 \
    --conf spark.memory.offHeap.enabled=true \
    --conf spark.memory.offHeap.size=20g \

The job fails with:

org.apache.spark.SparkException: 
  Job aborted due to stage failure: Task 0 in stage 251.0 failed 4 times, most recent failure: 
  Lost task 0.3 in stage 251.0 (TID 2171) (10.0.0.118 executor 0): 
  org.apache.comet.CometNativeException: 
  External error: 
  Internal error: Partition is still not able to allocate enough memory for the array builders after spilling..

@viirya
Copy link
Member Author

viirya commented Oct 18, 2024

I will try it with sf=100.

andygrove
andygrove reviewed Oct 18, 2024
View reviewed changes
native/core/src/execution/datafusion/shuffle_writer.rs
Comment on lines +71 to +72
/// The difference in memory usage after appending rows
MemDiff(Result<isize>),
Copy link
Member

@andygrove andygrove Oct 18, 2024

Choose a reason for hiding this comment

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

Will this always be an increase in memory? Should this use usize?

Copy link
Member Author

@viirya viirya Oct 18, 2024

Choose a reason for hiding this comment

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

It could be decrease too, if flush happens.

viirya
viirya commented Oct 18, 2024
View reviewed changes
native/core/src/execution/datafusion/shuffle_writer.rs
// Cannot allocate enough memory for the array builders in the partition,
// spill partitions and retry.
self.spill().await?;
self.reservation.free();
Copy link
Member Author

@viirya viirya Oct 18, 2024

Choose a reason for hiding this comment

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

Forgot to free memory reservation in previous commit.

@andygrove Could you try run benchmarks again? Thanks.

Copy link
Member

@andygrove andygrove Oct 18, 2024

Choose a reason for hiding this comment

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

I no longer see the memory error, but there seems to be a significant performance regression. TPC_H q2 used to take 12 seconds and is now taking many minutes. I do not see spill happening in Spark UI. I am going to add some debug logging to try and understand what is happening.

Copy link
Member

@andygrove andygrove Oct 18, 2024

Choose a reason for hiding this comment

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

In native metrics, I do see excessive spilling:

 spill_count=8, spilled_bytes=19441254400, data_size=877436

Copy link
Member Author

@viirya viirya Oct 18, 2024

Choose a reason for hiding this comment

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

Okay. I will take a look it further.

Copy link
Member Author

@viirya viirya Oct 18, 2024

Choose a reason for hiding this comment

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

I guess that is because we silently use some memory but never report them into the reservation, like the memory usage on array builders, now we count for them. So under same memory settings, it is more likely you hit the bar of memory pool. Have you try to increase the Comet memory like spark.comet.memoryOverhead?

Copy link
Member

@andygrove andygrove Oct 18, 2024

Choose a reason for hiding this comment

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

I tried setting the overhead:

    --conf spark.executor.instances=1 \
   --conf spark.executor.memory=16G \
   --conf spark.executor.cores=8 \
   --conf spark.cores.max=8 \
   --conf spark.memory.offHeap.enabled=true \
   --conf spark.memory.offHeap.size=20g \
   --conf spark.comet.memoryOverhead=16g \

This did not help with performance:

Query 2 took 352.05119466781616 seconds

Copy link
Member Author

@viirya viirya Oct 19, 2024

Choose a reason for hiding this comment

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

With latest commit, I ran TPCH sf=100 locally and didn't see regression now. Can you verify it?

Copy link
Member

@andygrove andygrove Oct 19, 2024

Choose a reason for hiding this comment

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

Looks good!

Query 2 took 12.09787917137146 seconds

@codecov-commenter
Copy link

codecov-commenter commented Oct 19, 2024 •
edited
Loading

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 34.43%. Comparing base (591f45a) to head (fd78a74).
Report is 3 commits behind head on main.

Additional details and impacted files 
@@             Coverage Diff              @@
##               main    #1022      +/-   ##
============================================
+ Coverage     34.30%   34.43%   +0.13%     
- Complexity      887      898      +11     
============================================
  Files           112      112              
  Lines         43429    43538     +109     
  Branches       9623     9660      +37     
============================================
+ Hits          14897    14994      +97     
- Misses        25473    25479       +6     
- Partials       3059     3065       +6

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

andygrove
andygrove approved these changes Oct 19, 2024
View reviewed changes
Copy link
Member

@andygrove andygrove left a comment

Choose a reason for hiding this comment

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

Thanks @viirya

@andygrove andygrove merged commit 4033687 into apache:main Oct 19, 2024
@viirya
Copy link
Member Author

viirya commented Oct 19, 2024

Thanks @andygrove

@viirya viirya deleted the revise_shuffle_memory_fix branch October 19, 2024 18:31
coderfender pushed a commit to coderfender/datafusion-comet that referenced this pull request Dec 13, 2025
@andygrove
chore: bring in latest OSS comet changes (apache#498)
ccc7b06 
## Which issue does this PR close?

<!--
We generally require a GitHub issue to be filed for all bug fixes and
enhancements and this helps us generate change logs for our releases.
You can link an issue to this PR using the GitHub syntax. For example
`Closes apache#123` indicates that this PR will close issue apache#123.
-->

Closes #.

## Rationale for this change

<!--
Why are you proposing this change? If this is already explained clearly
in the issue then this section is not needed.
Explaining clearly why changes are proposed helps reviewers understand
your changes and offer better suggestions for fixes.
-->

## What changes are included in this PR?

<!--
There is no need to duplicate the description in the issue here but it
is sometimes worth providing a summary of the individual changes in this
PR.
-->

```
cb3e977 perf: Add experimental feature to replace SortMergeJoin with ShuffledHashJoin (apache#1007)
3df9d5c fix: Make comet-git-info.properties optional (apache#1027)
4033687 chore: Reserve memory for native shuffle writer per partition (apache#1022)
bd541d6 (public/main) remove hard-coded version number from Dockerfile (apache#1025)
e3ac6cf feat: Implement bloom_filter_agg (apache#987)
8d097d5 (origin/main) chore: Revert "chore: Reserve memory for native shuffle writer per partition (apache#988)" (apache#1020)
591f45a chore: Bump arrow-rs to 53.1.0 and datafusion (apache#1001)
e146cfa chore: Reserve memory for native shuffle writer per partition (apache#988)
abd9f85 fix: Fallback to Spark if named_struct contains duplicate field names (apache#1016)
22613e9 remove legacy comet-spark-shell (apache#1013)
d40c802 clarify that Maven central only has jars for Linux (apache#1009)
837c256 docs: Various documentation improvements (apache#1005)
0667c60 chore: Make parquet reader options Comet options instead of Hadoop options (apache#968)
0028f1e fix: Fallback to Spark if scan has meta columns (apache#997)
b131cc3 feat: Support `GetArrayStructFields` expression (apache#993)
3413397 docs: Update tuning guide (apache#995)
afd28b9 Quality of life fixes for easier hacking (apache#982)
18150fb chore: Don't transform the HashAggregate to CometHashAggregate if Comet shuffle is disabled (apache#991)
a1599e2 chore: Update for 0.3.0 release, prepare for 0.4.0 development (apache#970)
```

## How are these changes tested?

<!--
We typically require tests for all PRs in order to:
1. Prevent the code from being accidentally broken by subsequent changes
2. Serve as another way to document the expected behavior of the code

If tests are not included in your PR, please explain why (for example,
are they covered by existing tests)?
-->
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@andygrove andygrove andygrove approved these changes

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

CometNativeException: called Option::unwrap() on a None value

3 participants

@viirya @andygrove @codecov-commenter