Hello, we are currently experiencing a high amount of false failure reports.
This occurs when many pull requests are being tested, tests will fail as if the environment was slow or simply get econnrefused message that selenium was unable to start.
We added extra heads for the tests, we also reduced the test size to see if it had any improvement, but it also got worst as soon as we increased pull request frequency.
Additionally, we are using suite and test retries as well as concurrent tests, to see if that makes anything better, without little or no success.
We are not sure if maybe it has to do with some sort of configuration error on our side, or if it is an issue by circle ci regarding high concurrence, or if it has to do with the fact that project may be too big for what circle ci can offer.
We would prefer to keep circle ci, since we are committed to its structure already, and we do like how it works.
Maybe you could provide some help or insight on this issue, thanks.