You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
While working on #7864 over the weekend, the Linux/Windows JS CI tests each had at least half a dozen spurious failures. Every single time the failure occurred during the run of the test-services package. The test-whatever packages are run in alphabetical order, so test-services isn't run until near the end of the tests, at around the ~30-35 min mark.
I think it would make the CI slightly less painful to deal with if test-services ran first. The tests would still fail the same, of course, but at least that way we wouldn't have to wait 30 minutes just to see the test die. Is this a sensible proposal?
The text was updated successfully, but these errors were encountered:
While working on #7864 over the weekend, the Linux/Windows JS CI tests each had at least half a dozen spurious failures. Every single time the failure occurred during the run of the test-services package. The test-whatever packages are run in alphabetical order, so test-services isn't run until near the end of the tests, at around the ~30-35 min mark.
I think it would make the CI slightly less painful to deal with if test-services ran first. The tests would still fail the same, of course, but at least that way we wouldn't have to wait 30 minutes just to see the test die. Is this a sensible proposal?
The text was updated successfully, but these errors were encountered: