You are viewing a plain text version of this content. The canonical link for it is here.
Posted to github@arrow.apache.org by GitBox <gi...@apache.org> on 2022/11/24 16:34:02 UTC
[GitHub] [arrow-datafusion] xudong963 opened a new pull request, #4361: Support to use Schedular in tpch benchmark
xudong963 opened a new pull request, #4361:
URL: https://github.com/apache/arrow-datafusion/pull/4361
# 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 #123` indicates that this PR will close issue #123.
-->
No
# 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.
-->
Support to test `Schedular`(the newest executor designed by @tustvold ) performance in tpch benchmark
# 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.
-->
Support to use `Schedular` in tpch benchmark
# 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)?
-->
No need, I have checked in my local.
```
โ benchmarks git:(master) โ cargo run --release --bin tpch -- benchmark datafusion --iterations 1 --path ~/parquet_data --format parquet --enable-scheduler --query 1
Finished release [optimized] target(s) in 0.13s
Running `/arrow-datafusion/target/release/tpch benchmark datafusion --iterations 1 --path /Users/xudong/parquet_data --format parquet --enable-scheduler --query 1`
Running benchmarks with the following options: DataFusionBenchmarkOpt { query: Some(1), debug: false, iterations: 1, partitions: 2, batch_size: 8192, path: "/Users/xudong/parquet_data", file_format: "parquet", mem_table: false, output_path: None, disable_statistics: false, enable_scheduler: true }
Query 1 iteration 0 took 3847.1 ms and returned 4 rows
Query 1 avg time: 3847.15 ms
```
# Are there any user-facing changes?
<!--
If there are user-facing changes then we may require documentation to be updated before approving the PR.
-->
<!--
If there are any breaking changes to public APIs, please add the `api change` label.
-->
No
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow-datafusion] xudong963 commented on pull request #4361: Support to use Schedular in tpch benchmark
Posted by GitBox <gi...@apache.org>.
xudong963 commented on PR #4361:
URL: https://github.com/apache/arrow-datafusion/pull/4361#issuecomment-1326937365
@isidentical It's expected.
There is a tracking issue: https://github.com/apache/arrow-datafusion/issues/2504
I add the `Option` because I plan to iterate our new executor with @tustvold, the option will help check some iterations quickly.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow-datafusion] ursabot commented on pull request #4361: Support to use Schedular in tpch benchmark
Posted by GitBox <gi...@apache.org>.
ursabot commented on PR #4361:
URL: https://github.com/apache/arrow-datafusion/pull/4361#issuecomment-1327408579
Benchmark runs are scheduled for baseline = 58b43f5c0b629be49a3efa0e37052ec51d9ba3fe and contender = 010aded5d524d90e8b96520457321250f1cc4c1d. 010aded5d524d90e8b96520457321250f1cc4c1d is a master commit associated with this PR. Results will be available as each benchmark for each run completes.
Conbench compare runs links:
[Skipped :warning: Benchmarking of arrow-datafusion-commits is not supported on ec2-t3-xlarge-us-east-2] [ec2-t3-xlarge-us-east-2](https://conbench.ursa.dev/compare/runs/ebfe3b53553f4bd5bca428db8129812b...16469809f6114a299a854df7cef8611a/)
[Skipped :warning: Benchmarking of arrow-datafusion-commits is not supported on test-mac-arm] [test-mac-arm](https://conbench.ursa.dev/compare/runs/60522d960ef1410bae315ef9838b97fa...97f0d768330c44669224919de9ed50f0/)
[Skipped :warning: Benchmarking of arrow-datafusion-commits is not supported on ursa-i9-9960x] [ursa-i9-9960x](https://conbench.ursa.dev/compare/runs/e3a304672e394f998f4d852dfb73ecfa...fccca4a14b534ebfaad13d8b409dedf2/)
[Skipped :warning: Benchmarking of arrow-datafusion-commits is not supported on ursa-thinkcentre-m75q] [ursa-thinkcentre-m75q](https://conbench.ursa.dev/compare/runs/322033f0914141549d351a533ca76cf4...ce88368e26bd455883327099188da5c0/)
Buildkite builds:
Supported benchmarks:
ec2-t3-xlarge-us-east-2: Supported benchmark langs: Python, R. Runs only benchmarks with cloud = True
test-mac-arm: Supported benchmark langs: C++, Python, R
ursa-i9-9960x: Supported benchmark langs: Python, R, JavaScript
ursa-thinkcentre-m75q: Supported benchmark langs: C++, Java
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow-datafusion] xudong963 commented on pull request #4361: Support to use Schedular in tpch benchmark
Posted by GitBox <gi...@apache.org>.
xudong963 commented on PR #4361:
URL: https://github.com/apache/arrow-datafusion/pull/4361#issuecomment-1327211397
> Perhaps when I finish up integrating the row format everywhere I'll have some more time to work on this ๐
I'll start to do it after work and look forward to seeing you back!
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow-datafusion] xudong963 merged pull request #4361: Support to use Schedular in tpch benchmark
Posted by GitBox <gi...@apache.org>.
xudong963 merged PR #4361:
URL: https://github.com/apache/arrow-datafusion/pull/4361
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow-datafusion] xudong963 commented on a diff in pull request #4361: Support to use Schedular in tpch benchmark
Posted by GitBox <gi...@apache.org>.
xudong963 commented on code in PR #4361:
URL: https://github.com/apache/arrow-datafusion/pull/4361#discussion_r1031694401
##########
benchmarks/src/bin/tpch.rs:
##########
@@ -337,7 +344,13 @@ async fn execute_query(
);
}
let task_ctx = ctx.task_ctx();
- let result = collect(physical_plan.clone(), task_ctx).await?;
+ let result = if enable_scheduler {
+ let scheduler = Scheduler::new(num_cpus::get());
Review Comment:
I use `num_cpus` to initial the size of thread pool.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [arrow-datafusion] isidentical commented on pull request #4361: Support to use Schedular in tpch benchmark
Posted by GitBox <gi...@apache.org>.
isidentical commented on PR #4361:
URL: https://github.com/apache/arrow-datafusion/pull/4361#issuecomment-1326835681
This is a cool option! Just out of my own curiosity I've compared `--enable-scheduler` vs not to see if there are any clear wins and there are a few queries where the scheduler outperforms by %10. Obviously, there are still some others with major degradations but with this option, it should be much easier to identify those cases and see the underlying cause.
```
โโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโ
โ Query โ Baseline โ Comparison โ Change โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Q1 โ 643.67ms โ 592.49ms โ +1.09x faster โ
โ Q2 โ 291.85ms โ 262.98ms โ +1.11x faster โ
โ Q3 โ 373.14ms โ 1028.82ms โ 2.76x slower โ
โ Q4 โ 103.54ms โ 105.93ms โ no change โ
โ Q5 โ 363.02ms โ 714.48ms โ 1.97x slower โ
โ Q6 โ 370.04ms โ 374.17ms โ no change โ
โ Q7 โ 745.62ms โ 1316.95ms โ 1.77x slower โ
โ Q8 โ 471.89ms โ 867.03ms โ 1.84x slower โ
โ Q9 โ 651.69ms โ 998.41ms โ 1.53x slower โ
โ Q10 โ 432.42ms โ 1165.39ms โ 2.70x slower โ
โ Q11 โ 204.11ms โ 173.80ms โ +1.17x faster โ
โ Q12 โ 186.74ms โ 194.05ms โ no change โ
โ Q13 โ 404.37ms โ 379.72ms โ +1.06x faster โ
โ Q14 โ 337.40ms โ 413.87ms โ 1.23x slower โ
โ Q15 โ 325.42ms โ 624.71ms โ 1.92x slower โ
โ Q16 โ 153.15ms โ 155.41ms โ no change โ
โ Q17 โ 1931.96ms โ 2051.97ms โ 1.06x slower โ
โ Q18 โ 1349.79ms โ 1280.70ms โ +1.05x faster โ
โ Q19 โ 621.30ms โ 683.34ms โ 1.10x slower โ
โ Q20 โ 609.66ms โ 642.90ms โ 1.05x slower โ
โ Q21 โ 628.37ms โ 710.87ms โ 1.13x slower โ
โ Q22 โ 65.89ms โ 65.55ms โ no change โ
โโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโ
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@arrow.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org