wanted faster feedback on tests, using the pytest-xdist with pytest lets us spread tests across available cores (4 on GH runners)
When using all 4, I saw an --iterative-timeout, so dropped the GH action specifically to use 1 cores. Honestly didn't see much speedup on GH, it's nice to have locally though so I think this is still worth it.
realized we had a CNAME_LOOP_ANSWERS constant in integration_tests but no test actually used it, added one
Performance
Running the integration_tests.py on a MacBook Air:
Description
pytest-xdistwithpytestlets us spread tests across available cores (4 on GH runners)--iterative-timeout, so dropped the GH action specifically to use 1 cores. Honestly didn't see much speedup on GH, it's nice to have locally though so I think this is still worth it.integration_testsbut no test actually used it, added onePerformance
Running the
integration_tests.pyon a MacBook Air:main- 20.9 s.