NVIDIA
[testing] [Python] Intermittent failure on 'remote' target test
During the development of PR https://github.com/NVIDIA/cuda-quantum/pull/1993/, it was noticed that appending a test in python/tests/kernel/test_kernel_features.py caused CI to crash on python/tests/remote/test_remote_platform.py::test_observe only in Python3.8, Fedora, x86 environment.
Sample pipeline - https://github.com/NVIDIA/cuda-quantum/actions/runs/10050329434/job/27778722969?pr=1993
This behavior was not reliably reproducible locally (@bmhowe23 managed to reproduce it sometimes).
Moving the test to new location allowed CI to pass in this case. (See: https://github.com/NVIDIA/cuda-quantum/pull/1993/commits/d43e4b67fba5784ad779985c6297b339c7bf7bd6) However, the root cause is unknown.
Hi @khalatepradnya ,
I'm interested in contributing to the CUDA Quantum library and came across issue [#1998](Intermittent failure on 'remote' target test). I noticed that appending a test in python/tests/kernel/test_kernel_features.py caused CI to crash on python/tests/remote/test_remote_platform.py::test_observe only in the Python 3.8, Fedora, x86 environment.
I have experience with Python and am eager to learn more about debugging and solving such issues. I would like to take on this challenge and help find the root cause of this problem. Could you please assign this task to me or provide some guidance on where to start?
Thank you!
Hi @khalatepradnya ,
I'm interested in contributing to the CUDA Quantum library and came across issue [#1998](Intermittent failure on 'remote' target test). I noticed that appending a test in python/tests/kernel/test_kernel_features.py caused CI to crash on python/tests/remote/test_remote_platform.py::test_observe only in the Python 3.8, Fedora, x86 environment.
I have experience with Python and am eager to learn more about debugging and solving such issues. I would like to take on this challenge and help find the root cause of this problem. Could you please assign this task to me or provide some guidance on where to start?
Thank you!
Thank you for your interest. I will assign this task to you.
I think the first step is to reproduce this issue reliably, outside of CI environment for debugging. You can start by reverting the change in commit https://github.com/NVIDIA/cuda-quantum/commit/d43e4b67fba5784ad779985c6297b339c7bf7bd6 and try to reproduce it by running whole suite of Python tests - python3 -m pytest -v python/tests/ --ignore python/tests/backends.
Hey @khalatepradnya , @sacpis I am new to open source contribution stuff and trying to cherry-pick and revert a commit with hash d43e4b6 in the cuda-quantum repository, but I'm encountering issues. Despite following the steps to fetch updates from the upstream repository, I receive the following error messages:
fatal: bad revision 'd43e4b6' Steps Taken:
- Added the upstream repository as a remote.
- Fetched the latest updates from the upstream repository.
- Attempted to cherry-pick and revert the commit d43e4b6.
Issue: The commit d43e4b6 does not seem to exist in the upstream repository or any of its branches, leading to errors during the cherry-pick and revert operations.
Attached Screenshot:
Could you please assist me in resolving this issue? Specifically, I am trying to understand why the commit is not found and if there is an alternative approach to achieve my goal.
thank you.
Hey @khalatepradnya , @sacpis I am new to open source contribution stuff and trying to cherry-pick and revert a commit with hash d43e4b6 in the cuda-quantum repository, but I'm encountering issues. Despite following the steps to fetch updates from the upstream repository, I receive the following error messages:
fatal: bad revision 'd43e4b6' Steps Taken: ... Could you please assist me in resolving this issue? Specifically, I am trying to understand why the commit is not found and if there is an alternative approach to achieve my goal.
thank you.
You can pull the changes from the remote repo to which the commit belongs to. I think it would be easier to manually revert the changes. i.e. simply undo what that commit did locally and proceed.