API Monitors: A False Sense of Security

APIs are the backbone of our web and mobile platforms today. They literally run the world, yet we don’t monitor them properly. Traditional approaches to monitoring have focused too much on performance and ping tests. That approach has proven to be highly unreliable, failing to alert teams when an API is not functioning correctly. Now, with more and faster releases thanks to agile, the problem of unreliable APIs is becoming worse. That’s why a new measure is needed that goes beyond uptime - Functional Uptime. Monitors should reproduce actual user flows that validate that the entire chain of calls works as expected. Thanks to the proliferation of new automated testing tools and techniques, this has never been easier to set up. In this demo-focused session, we will reveal the devastating impact that “false uptime” from traditional monitoring has had in five real-life examples from customers. We will show what the bug was, what caused it, why it wasn’t caught, and finally how to make sure you don’t get caught off guard as well.