Skip to content

Add benchmark for async tree workloads#187

Merged
ericsnowcurrently merged 8 commits intopython:mainfrom
arielin3:bm_async_tree
May 16, 2022
Merged

Add benchmark for async tree workloads#187
ericsnowcurrently merged 8 commits intopython:mainfrom
arielin3:bm_async_tree

Conversation

@arielin3
Copy link
Contributor

@arielin3 arielin3 commented Apr 27, 2022 •
edited
Loading

What

Add a benchmark for testing async workloads, specifically an async tree workload that simulates simpler versions of a typical Instagram endpoint. ("typical implementation of endpoint in IG is an async function that in turn calls into numerous other async functions to generate an output." - see context below)

Motivation/Context

See python/cpython#91121

Testing

Running pyperformance locally:

% python3 -m pyperformance list
'<default>' benchmarks:
- 2to3
- async_tree
- async_tree_cpu_io_mixed
- async_tree_io
- async_tree_memoization
- chameleon
- chaos
...

% python3 -m pyperformance run -o ~/python_efficiency_hackamonth/pyperformance_async_tree_output.json --benchmarks=async_tree,async_tree_cpu_io_mixed,async_tree_memoization,async_tree_io
Python benchmark suite 1.0.5

Creating the virtual environment /Users/arielin/python_efficiency_hackamonth/pyperformance_git/pyperformance/venv/cpython3.10-38b51ccaf402-compat-c0d88e07feb9

...

Performance version: 1.0.5
Python version: 3.10.4 (64-bit) revision 9d38120e33
Report on macOS-12.3.1-x86_64-i386-64bit
Number of logical CPUs: 16
Start date: 2022-05-04 18:06:49.877751
End date: 2022-05-04 18:13:51.586338

### async_tree_none ###
Mean +- std dev: 830 ms +- 59 ms

### async_tree_cpu_io_mixed ###
Mean +- std dev: 1.10 sec +- 0.05 sec

### async_tree_io ###
Mean +- std dev: 2.00 sec +- 0.05 sec

### async_tree_memoization ###
Mean +- std dev: 963 ms +- 37 ms

(EDIT: Testing updated on 5/4 after rename changes)

arielin3
arielin3 commented Apr 27, 2022
View reviewed changes
gvanrossum
gvanrossum reviewed Apr 28, 2022
View reviewed changes
@itamaro itamaro mentioned this pull request May 2, 2022
Merged
@ericsnowcurrently ericsnowcurrently self-requested a review May 3, 2022 19:43
ericsnowcurrently
ericsnowcurrently reviewed May 9, 2022
View reviewed changes
Copy link
Member

@ericsnowcurrently ericsnowcurrently left a comment

Choose a reason for hiding this comment

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

Mostly LGTM.

ericsnowcurrently
ericsnowcurrently approved these changes May 13, 2022
View reviewed changes
Copy link
Member

@ericsnowcurrently ericsnowcurrently left a comment

Choose a reason for hiding this comment

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

I'll merge this tomorrow, after you've had a change to make any final changes.

@arielin3
Copy link
Contributor Author

arielin3 commented May 14, 2022

@ericsnowcurrently Thanks so much for the review! Updated and addressed all the comments :)

@arielin3
Copy link
Contributor Author

arielin3 commented May 16, 2022

@ericsnowcurrently It looks like the most recent test failure (test / ubuntu-latest - 3.11-dev) is due to the "sqlalchemy" benchmarks and not async tree related - it failed at creating venvs for that benchmarks, at

(43/53) creating venv for benchmark (sqlalchemy_declarative)
(44/53) creating venv for benchmark (sqlalchemy_imperative)

Are you able to rerun the failed test?

@ericsnowcurrently
Copy link
Member

ericsnowcurrently commented May 16, 2022

Yeah, it's not this PR.

@ericsnowcurrently ericsnowcurrently merged commit 6e7b445 into python:main May 16, 2022
@ericsnowcurrently
Copy link
Member

ericsnowcurrently commented May 16, 2022

Thanks for working on this!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@gvanrossum gvanrossum gvanrossum left review comments

@ericsnowcurrently ericsnowcurrently ericsnowcurrently 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.

3 participants

@arielin3 @ericsnowcurrently @gvanrossum