Closed pglass closed 5 years ago
Not a worthwhile difference on Jenkins: https://jenkins.openrepose.org/job/repose-gradle-pull_request-core/70/testReport/features.core.configloadingandreloading/history/ and it actually took a bit longer than usual for these tests. I'm wondering if the waits take longer than the sleeps they replaced.
Could just be noisy neighbor or the particular slave you ended up on.
Retest this please
Whoo, so build 74 is the fastest it's ever been: https://jenkins.openrepose.org/job/repose-gradle-pull_request-core/74/testReport/features.core.configloadingandreloading/history/
Seems like there are improvements then, and it was just a cloud resource problem on the first run.
This removes or replaces some hardcoded sleeps in the
TransitionBadToGoodConfigsTest
andTransitionGoodToBadConfigsTest
.I recorded the runtimes before/after these changes in the commit messages, so please look through those. Adding up the differences for the modified tests, this saved 2 minutes out of about 19 minutes on my laptop.