Infrastructure monitoring has been around since the advent of IT itself. Knowing the state of your network devices, servers, and the applications that run on top of it all is paramount to ensuring that business critical functions are available. Over the past several years there has been a large shift in monitoring technologies from simply up/down polling and notification, to complete monitoring and analytics of the various components of an application. It’s not enough just to know that your network switch, server, or VM are reachable via ping or that your can access ports 80 and 443 on your web server.
The complete collection of logs and performance metrics, coupled with analytics and AI allows modern platforms to identify usage patterns, pinpoint bottlenecks, and provide full visibility into the cause of any outages or performance issues. Using this information not only provides greater visibility into multiple layers of an application and infrastructure stack, but also allows application owners and operations teams to become more proactive in the way they handle application and infrastructure performance.
Moving Monitoring into the Cloud Age
One of the leaders in the space of infrastructure and application performance monitoring (APM) is Datadog. After being founded in 2010 and rapidly adding support for multiple cloud platforms including Amazon Web Services (AWS), Microsoft Azure, and Google Cloud Platform (GCP), Datadog quickly established themselves as a leader in cloud infrastructure monitoring. For complete monitoring and analytics of an application and its infrastructure, Datadog’s features are quite robust.
Yet despite this maturity, there is still no capability within the Datadog platform to identify what issues may exist on the front end of an application. If the development team responsible for an application’s front-end discovers a bug or performance issue, everything that Datadog has visibility into can quickly be investigated and eliminated as the root cause. That could still leave the root cause undiscovered and the front-end developers responsible for hunting it down. To solve this problem and provide full stack visibility, Datadog has recently announced the acquisition of AI powered startup Madumbo.
Madumbo co-founders Gabriel-James Safar (CEO) and Sébastien Deprez (CTO)
Becoming a Full Stack Solution
Founded in 2017, Madumbo’s first product was released with the goal of automating end to end tests of an application. This is accomplished in two ways. First, the Madumbo test recorder can watch as a test is performed manually and repeat it later in an automated fashion. If elements within a page are renamed or moved around, Madumbo’s AI will recognized these changes and adapt the test automatically. Any problems with code or errors that are revealed during testing can be exported to a ticketing system or integrations with a CI/CD system could prevent a problem build from deploying.
Additionally Madumbo’s “roamer” allows the product to automatically find and interact with all elements on a page. From there, it can follow all the potential paths in an application and essentially act as a smoke test. This will in effect trigger all states and ensure that they do not result in Javascript, network, or language errors. The roamer and recorder functionality within Madumbo provides the full visibility and testing needed to completely understand the working state of an application front-end. Integrations of Madumbo into the Datadog platform are currently being developed to give customers and end to end view of their entire application through a single utility.
Ken’s Conclusion
Given the lack of features to provide visibility of an application’s front-end within Datadog’s existing platform, the acquisition of Madumbo is a logical step in elevating the product to be a full stack solution. As great as APM is at identifying the cause of application performance issues or outages, it is still somewhat reactive. Having a tool like Madumbo integrated with Datadog will provide a more proactive solution to application monitoring and troubleshooting. Allowing developers to identify issues before they hit production and remediate before major issues arise is a powerful capability. Full stack visibility is a logical next step in the evolution of APM as a technology. With the acquisition of Madumbo, Datadog is poised to maintain a top spot in the APM space.
