More and more frequently, I have to re-run the pytest tests in CI due to a failure like:
=========================== short test summary info ============================
FAILED tests/test_api/test_api_spec.py::test_invalid_bearer - jwt.exceptions.PyJWKClientConnectionError: Fail to fetch data from the url, err: "<urlopen error [Errno 101] Network is unreachable>"
FAILED tests/test_api/test_auth.py::test_decode_edl_bearer_token - jwt.exceptions.PyJWKClientConnectionError: Fail to fetch data from the url, err: "<urlopen error [Errno 101] Network is unreachable>"
FAILED tests/test_api/test_auth.py::test_bad_bearer_token - jwt.exceptions.PyJWKClientConnectionError: Fail to fetch data from the url, err: "<urlopen error [Errno 101] Network is unreachable>"
=========== 3 failed, 234 passed, 120 warnings in 105.97s (0:01:45) ============
make: *** [Makefile:37: tests] Error 1
Error: Process completed with exit code 2.
https://github.com/ASFHyP3/hyp3/actions/runs/28409610457/job/84179591398?pr=3155
Generally, re-running makes them pass, so these don't seem like valid failures.
We should update our tests to make them less flaky.
More and more frequently, I have to re-run the pytest tests in CI due to a failure like:
https://github.com/ASFHyP3/hyp3/actions/runs/28409610457/job/84179591398?pr=3155
Generally, re-running makes them pass, so these don't seem like valid failures.
We should update our tests to make them less flaky.