We know what APIs are now, so what is API health?
Firstly, to recap, an API (Application Programming Interface) forms the communications links and bond between different elements of a software program, or different data streams.
A developer can use an API to code a link from one program (or program component, or operating system) to another to request services from that other application.
As the glue and between information components in any given software system, APIs have a required syntax and are implemented by function calls made up of verbs and nouns.
The required API syntax is described in the documentation of the application being called.
There are local APIs, web APIs and program APIs and you can read more here.
So then, back to our original question on API healthiness.
What is API health?
API testing, UI testing, code review and performance monitoring company SmartBear has defined API health for the Computer Weekly Developer Network as follows:
“API health refers to the availability, performance and functional correctness of not just your API endpoints, but also the business transactions that APIs support, whether they’re driven by the organisation [that owns the API] or a third-party. Full-visibility into API health means you [the developer] can gauge how the APIs are behaving in the both production environments and pre-production environments, so you can understand when and where degradation (of availability, speed, performance, functionality) occur.”
Why the fuss of focus here in the first place? Okay let’s give SmartBear its due… the company used the term in relation to several new API Monitoring capabilities in its AlertSite performance monitoring product.
The new API Monitoring capabilities in AlertSite are supposed to enable operations, DevOps and development teams to gain visibility into API performance, availability and functional correctness in real-time, across production, staging and engineering environments.
The company says that as APIs increasingly serve as the foundation for digital business, the need to ensure these APIs are highly reliable and performant is critical.
API monitoring challenges
However, SmartBear says there are several challenges with existing API monitoring tools that often prevent teams from being able to gain the visibility they need into the real-time health of their APIs.
- First, the complexity of APIs can make it overly complex and time-consuming to create API Monitors.
- Second, most API monitoring tools only provide basic insights into whether an API is up or down, but do not provide sufficient visibility into functional correctness or performance.
- Third, most API monitoring tools blast false alerts several times a day, which often results in real alerts being missed.
- Finally, API updates are released more frequently than ever as a result of microservices growth and monitors often require extensive re-work to prevent blind spots as APIs change.
SmartBear says that AlertSite has taken a different approach. It offers a way to create API monitors in a way that matches their operational cadence and technical skillset — if embracing DevOps and the Shift Left, simply upload existing OpenAPI Specification (OAS) files.
If running a rapid release cadence that doesn’t leave time to re-work monitors when APIs change, upload existing SoapUI test scripts. If teams are not highly technical, leverage an intuitive point-and-click wizard.