The frequency of flaky tests interrupting the build seems to have increased, and disrupts review and deployment of otherwise qualified changes. These fixes address two flaky tests in gardener:
TestStandardMonitor - could fail when some integration test operations took longer than 30s to complete. Anecdotally, tis can sometimes be multiple minutes. Since this completion time is not under our control, this change increases the timeout and ignores operations that are still in their final state ("joining").
Coverage decreased (-0.06%) to 60.566% when pulling 71f22069d0d17d629888d65b9e26d6575bf7cbcc on sandbox-soltesz-flaky-tests into 420be6fca3ee24427fba425d287d7bf696405b23 on master.
The frequency of flaky tests interrupting the build seems to have increased, and disrupts review and deployment of otherwise qualified changes. These fixes address two flaky tests in gardener:
Fixes: https://github.com/m-lab/etl-gardener/issues/337
Addresses: https://github.com/m-lab/etl-gardener/issues/328
This change is![Reviewable](https://reviewable.io/review_button.svg)