Earlier this week, we talked about the varying quality of web APIs. Reliability can be incredibly important in the world of APIs. Just as a chain can only be as strong as its weakest link, an application can only be as reliable as the least reliable API it utilizes.
Hence, it’s vital that a developer ensures their API meets, if not surpasses, their standards. Why? Because a small mistake here and there can lead to the catastrophic failures of countless applications. A developer’s job does not simply end when he releases his final version. What comes after development, that is, servicing, managing, and fixing the API, is arguably much more important than the development itself. A simple but well-maintained API is infinitely more useful than a powerful but buggy one. Features can bring in customers, but consistency makes them stay.
This raises the question, how can we check to make sure our APIs stay up to par? That’s where API monitoring comes in. Even the simplest software will provide a wealth of invaluable information, such as success rate, response time, and status codes. There’s no excuse not to utilize API monitoring. Thanks to the SaaS business model, there are literally hundreds upon hundreds of API monitoring systems available online for you to choose from, and many of them are free.
In my opinion, the toughest part of utilizing these services was creating an account. After creating an API Science account, you’ll find that they are extremely simple to use. Just plug in your API endpoint and just like that, you’ve got data!
I recommend all three services. Keep in mind that Uptime Robot offers a freemium package, while Runscope and API Science offer free trials.
The tools that Uptime Robot offers are a bit bare-bones, but sometimes that’s all you need. You have limited control over how it forms its requests and when it tests its requests (other than the interval of time between each test). You still get all the basic information you can expect from any other API monitoring software, such as response time and status code records, all color-coded and nicely plotted.
Runscope is more powerful. Not only does it give you full control over request formation and testing schedule, but it also allows you to create complex, multi-step, dynamic tests. Runscope offers a number of unique debugging tools as well.
API Science is a nice mix between the two. With API Science, you have the the test control that Runscope offers without its special functionalities such as multi-tiered tests and debugging tools.
API monitoring is, without a doubt, essential to using APIs. It’s cost effective and easy to set up too. With that in mind, you have no reason to run away from using them. It’ll make you happy, it’ll make your clients happy, and it’ll make your API happy.
How do you monitor your APIs?