Scalable If microservices receive praise for nothing else, they have undoubtedly been hailed as the liberator of scalability. A microservice based system offers unique scalability because parts of an application scale independently. If there is a heavy load on a payment system, for example, it can be scaled on its own leaving the rest of the system untouched. By contrast, the monoliths of old tend to be stateful. Thus, scalability issues are solved: 1. V ertically, by throwing more compute resources at them, which ends up becoming expensive and limited in success. 2. H orizontally, by making use of sticky sessions, something else which can be quite brittle in the event of failure. Stateless microservices don’t encounter either of these challenges. In fact, it becomes very trivial to scale as needed horizontally. DevOps Oriented Conway’s Law promotes the idea that a team will produce a design reflective of its communication structure. Depending on your present organizational structure, this could be a lesson or a warning. That is to say, if your developers don’t mingle with operations outside of the annual company party, you’re in for some challenges. The microservices-related advantage here lies in the organization of teams around business functionality rather than technical concerns, leading to more efficient application design. Remember when we talked about the Data Scientist who pushes R models like a developer pushes Java? In a DevOps or BizDevOps model, that Data Scientist could sit, stand, or walk on a treadmill desk next to your QA engineer. This cross-functional team structure is breaking down the silos that are all too familiar with financial services, and are widely known to cause inefficiencies and lackluster customer experience.
A microservice based system offers unique scalability because parts of an application scale independently.
© 2017 PROVENIR ALL RIGHTS RESERVED
Made with FlippingBook Digital Publishing Software