Thanks to visit codestin.com
Credit goes to github.com

Skip to content

BLD: Upload wheel artifacts separately [wheel build] #20747

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

Merged
merged 1 commit into from
Jan 14, 2022

Conversation

lithomas1
Copy link
Collaborator

No description provided.

@github-actions github-actions bot added the 36 - Build Build related PR label Jan 6, 2022
@lithomas1 lithomas1 marked this pull request as draft January 6, 2022 03:00
@lithomas1 lithomas1 marked this pull request as ready for review January 6, 2022 05:18
@lithomas1 lithomas1 marked this pull request as draft January 6, 2022 05:19
@lithomas1 lithomas1 marked this pull request as ready for review January 6, 2022 14:51
Copy link
Member

@rgommers rgommers left a comment

Choose a reason for hiding this comment

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

Thanks @lithomas1!

This seems to work. On macOS:

Artifact name is valid!
Container for artifact "cp39-macosx" successfully created. Starting upload of file(s)
Uploaded /Users/runner/work/numpy/numpy/wheelhouse/numpy-1.23.0.dev0+350.g404bd8802-cp39-cp39-macosx_11_0_arm64.whl (66.6%) bytes 0:8388607
Uploaded /Users/runner/work/numpy/numpy/wheelhouse/numpy-1.23.0.dev0+350.g404bd8802-cp39-cp39-macosx_10_9_x86_64.whl (48.1%) bytes 0:8388607
Uploaded /Users/runner/work/numpy/numpy/wheelhouse/numpy-1.23.0.dev0+350.g404bd8802-cp39-cp39-macosx_10_9_x86_64.whl (96.3%) bytes 8388608:16777215
Uploaded /Users/runner/work/numpy/numpy/wheelhouse/numpy-1.23.0.dev0+350.g404bd8802-cp39-cp39-macosx_10_9_x86_64.whl (100.0%) bytes 16777216:17404266
Uploaded /Users/runner/work/numpy/numpy/wheelhouse/numpy-1.23.0.dev0+350.g404bd8802-cp39-cp39-macosx_11_0_arm64.whl (100.0%) bytes 8388608:12591682
Total size of all the files uploaded is 29995950 bytes
File upload process has finished. Finalizing the artifact upload

On Windows:

Artifact name is valid!
Container for artifact "cp310-win_amd64" successfully created. Starting upload of file(s)
Uploaded D:\a\numpy\numpy\wheelhouse\numpy-1.23.0.dev0+350.g404bd8802-cp310-cp310-win_amd64.whl (57.5%) bytes 0:8388607
Uploaded D:\a\numpy\numpy\wheelhouse\numpy-1.23.0.dev0+350.g404bd8802-cp310-cp310-win_amd64.whl (100.0%) bytes 8388608:14584226
Total size of all the files uploaded is 14584227 bytes
File upload process has finished. Finalizing the artifact upload
Artifact has been finalized. All files have been successfully uploaded!

PyPy on Linux:

Artifact name is valid!
Container for artifact "pp38-manylinux_x86_64" successfully created. Starting upload of file(s)
Uploaded /home/runner/work/numpy/numpy/wheelhouse/numpy-1.23.0.dev0+350.g404bd8802-pp38-pypy38_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (51.9%) bytes 0:8388607
Uploaded /home/runner/work/numpy/numpy/wheelhouse/numpy-1.23.0.dev0+350.g404bd8802-pp38-pypy38_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (100.0%) bytes 8388608:16148486
Total size of all the files uploaded is 16148487 bytes
File upload process has finished. Finalizing the artifact upload

It isn't entirely clear to me what this one line change does exactly though. Can you elaborate? Why is macOS special-cased here?

@lithomas1
Copy link
Collaborator Author

Basically, if the artifacts have the same name, they get combined into one zip. By default, the name for uploads is "artifact", so that's why they appear as a huge ~300 MB zip file.

Since it is a pain(and uses a lot of bandwidth) to download the whole thing, and most of the time you only need a subset of wheels for testing/debugging, I use the job name as the artifact name, which splits everything up.

macos is special cased since it has a "*" symbol in the job name e.g. cp38-macosx_*, which the uploading action doesn't like.
(It's also slightly different in that the x86_64 and arm64 wheels are in the same zip, since they are produced by the same job.)

@mattip
Copy link
Member

mattip commented Jan 14, 2022

Ahh, I see. Now there are 14 separate artifacts at the bottom of the wheel build summary page, and in the macOS one there are actually two wheels: x86_64 and arm64. The sizes seem to match the wheel sizes on PyPI

@mattip mattip merged commit ae270c3 into numpy:main Jan 14, 2022
@mattip
Copy link
Member

mattip commented Jan 14, 2022

Thanks @lithomas1

@lithomas1 lithomas1 deleted the patch-2 branch January 14, 2022 15:09
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
36 - Build Build related PR
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants