-
-
Notifications
You must be signed in to change notification settings - Fork 18.4k
CI: Bump timeouts Python dev builds #48973
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
Conversation
390abd5
to
504a480
Compare
I just ran the tests (--skip-slow --skip-db) from April 1 and they took 19:30, compared to 30:06 from yesterday. Of that half hour, 8 minutes is in tests/extension/test_arrow.py. Not sure what to do with this, just making sure its written down. |
(FWIW: The Python dev builds don't run the optional deps) Each test in |
I think the main issue is the build step. Building pandas shouldn't take 25 minutes like it does on the macOS job. |
At least this gives us a buffer (other builds are at 120) @lithomas1 any quick suggestions how we might speed things up or should that be addressed in a followup? |
Try changing this line https://github.com/pandas-dev/pandas/pull/48973/files#diff-f6f33d5fc5cfd739f0d6efd63ea9dff886abe35cdfff5672e9a863ec8da8e72cR78 to Numpy should be uploading Python 3.11 wheels so, we'll shave off ~5 minutes for sure from not compiling numpy. We might get a little more speed-up in the tests too from this. |
@@ -75,7 +75,7 @@ jobs: | |||
run: | | |||
python --version | |||
python -m pip install --upgrade pip setuptools wheel | |||
python -m pip install git+https://github.com/numpy/numpy.git | |||
python -m pip install -i https://pypi.anaconda.org/scipy-wheels-nightly/simple numpy |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
python -m pip install -i https://pypi.anaconda.org/scipy-wheels-nightly/simple numpy | |
python -m pip install numpy==1.23.3 |
They have wheels for mac, windows, and linux:
https://pypi.org/project/numpy/1.23.3/#files
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hm, it might be worth it to stay with numpy main, until Python 3.11 is out, just in case compat issues arise between numpy and Python 3.11.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM. Can you open a follow up issue so we can make a plan to get these timeouts back down?
…ds) (#49028) Backport PR #48973: CI: Bump timeouts Python dev builds Co-authored-by: Matthew Roeschke <[email protected]>
* Bump timeout * add --no-index * install numpy nightlies
e.g. https://github.com/pandas-dev/pandas/actions/runs/3193038414/jobs/5211185245