Skip to content
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.

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

[CHORE] Refactor local hash joins + pipeline connections #2719

Merged
merged 6 commits into from
Aug 29, 2024

Conversation

colin-ho
Copy link
Contributor

@colin-ho colin-ho commented Aug 23, 2024

  1. Separate hash join into build sink + probe op.
  2. Pipeline nodes now return a Pipeline Channel instead of nothing.
  3. Put stats counting in the counting senders / receivers.
Screenshot 2024-08-28 at 5 00 14 PM

@github-actions github-actions bot added the chore label Aug 23, 2024
Copy link

codspeed-hq bot commented Aug 23, 2024

CodSpeed Performance Report

Merging #2719 will improve performances by 51.6%

Comparing colin/separate-hash-joins (7c76de2) with main (dee1af8)

Summary

⚡ 1 improvements
✅ 15 untouched benchmarks

Benchmarks breakdown

Benchmark main colin/separate-hash-joins Change
test_count[1 Small File] 20.5 ms 13.6 ms +51.6%

Ok(Box::pin(mp_stream))
}

fn chunk_tables_into_micropartition_stream(
Copy link
Contributor Author

Choose a reason for hiding this comment

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

No need to chunk in source anymore, since ops do their own buffering.

@colin-ho colin-ho marked this pull request as ready for review August 27, 2024 01:11
@colin-ho colin-ho requested a review from samster25 August 27, 2024 01:15
@colin-ho colin-ho force-pushed the colin/separate-hash-joins branch from 177c072 to 7c76de2 Compare August 29, 2024 21:07
&self,
idx: usize,
input: &PipelineResultType,
state: Option<&mut Box<dyn IntermediateOperatorState>>,
Copy link
Member

Choose a reason for hiding this comment

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

I think this api can also be Option<&mut dyn IntermediateOperatorState> instead of the Box of one

Copy link
Contributor Author

Choose a reason for hiding this comment

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

I tried, but had a bit of difficulty. I believe this is because performing a as_deref_mut on Option<Box<dyn IntermediateOperatorState>> returns Option<&mut dyn IntermediateOperatorState + 'static>>, which outlives the original Option<Box<dyn IntermediateOperatorState>>. https://users.rust-lang.org/t/whats-the-lifetime-of-the-output-of-as-deref-mut-of-option-box-dyn-trait/105002/2

@colin-ho colin-ho merged commit 554856d into main Aug 29, 2024
35 checks passed
@colin-ho colin-ho deleted the colin/separate-hash-joins branch August 29, 2024 23:38
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants