Have you ever been notified of an API monitoring error late at night, only to find out that it was a temporary network error? Or do you have false positives in your API monitoring data that are affecting your customer reports?
Timeout errors can affect any API, maybe because of a temporary bad internet connection, or a complex database operation. And a lot of times, a simple re-run might return a 200 response. That's why we recently added a new feature called Retry on Failure.
How to Use Retry on Failure
Retry on Failure will automatically re-run a single test run, immediately after a test fails. Failed re-runs will not re-run again.
You can find this option in your test environment settings, under the Behaviors menu. It's set to off by default, and it's available for both test-specific or shared environments.
If you know that the API you're monitoring suffers from intermittent issues, and that it is not part of a bigger problem, definitely give this new feature a try. As we mentioned before, this can help you have a clearer picture of your API success/failure data, and get rid of false positives.
Another use case for this is if you're monitoring your tests using a longer timeframe, such as 15 minutes or an hour. With retries active, you can get more accurate data, without waiting for a test to be re-run only after a long period.
As a best practice, retry on failure works especially well if combined with threshold notifications. It will help you get to thresholds more quickly, so make sure to review your email and 3rd-party notification settings after activating it.
Questions and Feedback
Retry on failure can reduce the number of notifications sent to your team because of temporary network errors, and give you more accurate data about your APIs.