Open anissa111 opened 9 months ago
Definitely looks related to the 0.6.2 release
(asv-bench) ➜ benchmarks git:(main) ✗ asv --version
Couldn't load asv.plugins._mamba_helpers because
No module named 'libmambapy'
asv 0.6.2
(asv-bench) ➜ benchmarks git:(main) ✗ asv check
Couldn't load asv.plugins._mamba_helpers because
No module named 'libmambapy'
· Discovering benchmarks
· Checking benchmarks
·· No problems found.
(asv-bench) ➜ benchmarks git:(main) ✗ pip install asv=="0.6.1"
Collecting asv==0.6.1
Using cached asv-0.6.1-cp310-cp310-macosx_11_0_arm64.whl.metadata (3.5 kB)
Requirement already satisfied: asv-runner>=v0.1.0 in /Users/anissaz/miniconda3/envs/asv-bench/lib/python3.10/site-packages (from asv==0.6.1) (0.2.1)
Requirement already satisfied: json5 in /Users/anissaz/miniconda3/envs/asv-bench/lib/python3.10/site-packages (from asv==0.6.1) (0.9.14)
Requirement already satisfied: tabulate in /Users/anissaz/miniconda3/envs/asv-bench/lib/python3.10/site-packages (from asv==0.6.1) (0.9.0)
Requirement already satisfied: pyyaml in /Users/anissaz/miniconda3/envs/asv-bench/lib/python3.10/site-packages (from asv==0.6.1) (6.0.1)
Requirement already satisfied: pympler in /Users/anissaz/miniconda3/envs/asv-bench/lib/python3.10/site-packages (from asv==0.6.1) (1.0.1)
Requirement already satisfied: importlib-metadata in /Users/anissaz/miniconda3/envs/asv-bench/lib/python3.10/site-packages (from asv-runner>=v0.1.0->asv==0.6.1) (7.0.1)
Requirement already satisfied: zipp>=0.5 in /Users/anissaz/miniconda3/envs/asv-bench/lib/python3.10/site-packages (from importlib-metadata->asv-runner>=v0.1.0->asv==0.6.1) (3.17.0)
Using cached asv-0.6.1-cp310-cp310-macosx_11_0_arm64.whl (295 kB)
Installing collected packages: asv
Attempting uninstall: asv
Found existing installation: asv 0.6.2
Uninstalling asv-0.6.2:
Successfully uninstalled asv-0.6.2
Successfully installed asv-0.6.1
(asv-bench) ➜ benchmarks git:(main) ✗ asv check
· Discovering benchmarks
· Checking benchmarks
·· No problems found.
@philipc2, I think this is what's causing your asv ci issues, too (https://github.com/UXARRAY/uxarray/issues/699).
Looking at the logs it is because build
is missing, not libmambapy
(that's just a logging note). However, this is still an upstream bug, unless the pyproject
is malformed somehow.
Well. Unpinning did not appear to have gone according to plan (see logs). I, uh, don't know why exactly.
The benchmarks ran successfully right before the unpinning (see working logs). We did not get any sort of memory warning on that (or previous) benchmark run. I'm going to re-open this and re-pin asv for now.
First occurrence (2/12/24): https://github.com/NCAR/geocat-comp/actions/runs/7881645531/job/21505603710#step:8:15
Likely related to the asv 0.6.2 release type from yesterday.
This is also confusing because it fails on the commit step instead of from the benchmarks. I also think we should add some sort of
asv check
step to the workflow