As a daily commuter and an engineer, I have often thought there is some parallel that can be drawn between traffic on the roads and the digital services being used in corporate networks. Speed, congestion, accidents and any number of factors could impact your trip. Unlike the traffic helicopters reporting to help a driver steer clear of hazards, many in IT are caught by surprise when slow traffic or a problem is ahead.
Let me explain, I live in state where most residents view the speed limit as a suggested starting point, when the traffic is moving along at steady pace of 75MPH, everything is flowing smoothly. However, when a few drivers decide to obey the speed limit of 65MPH, then the rhythm of the commute becomes as congested as the coffee shop line on free donut day.
Today’s most modern applications use more than one digital service. These services are tightly choreographed to achieve whatever ultimate goal the business needs to keep running smoothly. Like with an accident on the highway, when one service completely fails it is pretty apparent what the issue is. Yet, when a service – or set of services – simply starts to slow down, the root cause of the failing infrastructure is harder to determine.
As digital services are consumed in an enterprise, there is an implicit service contract between the application and the external services. Application developers often assume that if no error occurs on an external transaction, then everything is working correctly. Simply slowing down services over a small period of time can cause a catastrophic failure of an application ecosystem. This applies to a single service slowing down or a set of services slowing down sporadically which unfortunately is occurring as more and more systems and networks become overloaded or under attack.
At Trapize, our digital services broker monitors all the services that your enterprise is using, think of us as your traffic helicopter reporting on the health of your digital traffic. We provide sophisticated monitoring of external services providing key performance metrics across a wide range of functionality. Alerts can be passed to the enterprise infrastructure or triggered in-band to the applications when services start to misbehave.
Like Sammy Hagar, I will freely admit, “I can’t drive 55”. Enterprises need a way to ensure their services aren’t as well.