API Monitoring is a synthetic approach to monitoring an API (Application Programming Interface) for availability, proper responses, and performance. API monitoring helps identify outages and poor performing API calls that cause applications to fail, dependent services and websites to fail, and adversely affect the user experience.

What is an API?

An API is a software product that receives calls or requests from another piece of software (website, application, operating system) providing a service. For example, when using a banking app, the user can take a photo of a check and submit it for deposit. A bank may use an API that takes the photo and pulls the needed information off the picture such as amount and account number. Another API may validate the account number and verify funds. The API handles the processing without the calling application needing to know how it works beyond the provided interface.

Why monitor APIs?

If a business provides an API or relies on one, making sure it is available and functioning well is a must to preserve brand reputation and revenue. If an API fails, the applications and websites that rely on them also fail. API monitoring makes sure everything is working properly 24/7.

How does API monitoring work?

An API monitor uses a remote computer to send requests to the API. The computer evaluates the API response for speed, content, and response codes. If anything about the response doesn't meet expectations, the monitoring service records an error. Frequently the monitoring service immediately issues a second test from a different monitoring location. If that test also fails, the monitoring service sends out an alert to notify the API provider or user that the API no longer works properly. An API monitor may check a single response or make several requests to test entire user scenarios.

What type of API Monitors are there?

You will find two types of API monitors, basic and multi-step monitors. Both monitor types can evaluate and track uptime and performance as frequently as once per minute.

Basic API monitors

Basic API monitors test a single API call. The checkpoint computer sends a request and receives the response. The checkpoint measures the time from sending the request to receiving the response, checks the response for a success code, and possibly checks the response for a specified string of characters.

Multi-step API Monitors

Multi-step API Monitoring goes beyond checking a single response and checks complete API interactions. An API might respond well in a single response, but how does the API perform with subsequent calls that require the reuse of values like product IDs, geolocation information, and user responses? What about when the API requires authentication or contains page redirects? Multi-step API Monitoring can handle most use cases from the single call to many while checking responses, extracting values, and building the next request.

Conclusion

If an API is crucial to a business or its customers, the API needs monitoring. Businesses should also test the third-party APIs that they rely on; failure or slow performance from a third-party API is just as damaging to a brand as a failure in their own product. Customers don't care who the API belongs to; they just expect the product to work.