operationalizing microservices

Over the past few years, microservices have evolved from an innovator-only architecture to a practice used to some degree at most companies. However, scaling up from a proof-of-concept project to a production-grade, enterprise-scale software platform built on microservices requires serious planning, dedication and time. The companies that have invested heavily in creating stable microservices architectures have learned many lessons on overcoming the challenges involved in operating complex, distributed systems.

Operationalizing Microservices includes:

• Mature Microservices and How to Operate Them - Microservices have allowed The Financial Times to release code to production 250 times more frequently than for their previous monolithic platform. There are many steps involved when optimizing for speed, with continuous delivery providing the foundation. But faster delivery is only one-third of the problem. In her role overseeing operations and reliability at FT, Sarah Wells has learned that the other two puzzles to solve are how to operate microservices, and what to do when people move on, leaving behind legacy microservices to be maintained.

• Design Microservice Architectures the Right Way - A handful of key decisions directly impact the quality and maintainability of a microservice architecture. Michael Bryzek believes investing time and effort into good tools, practices, and automation up front is critical to ensure that teams and systems remain productive and scale. At Flow Commerce, this philosophy covers infrastructure, continuous deployment, communication, event streaming, language choice and more.

• Monitoring and Managing Workflows across Collaborating Microservices - Proponents of microservices like to cite loose coupling as a major benefit of the architecture. But, as with any architectural decision, loose coupling comes with trade-offs. By allowing for flexible choreography between services, you lose some ability to ensure those services are interacting according to established business processes. Bernd Rücker describes the challenges of balancing this trade-off and covers solutions ranging from monitoring to full orchestration.

• Lessons from 300k+ Lines of Infrastructure Code - Becoming an expert in anything requires repetitive practice. Writing hundreds of thousands of lines of infrastructure code should, therefore, provide quite a bit of expert guidance. Yevgeniy Brikman shares some of those lessons through a four-part infrastructure cookbook. He believes too many teams only focus on the first part, involving configuration and deployment, and fail to realize how much additional, time-consuming work is still required to create production-grade infrastructure. The checklist includes everything from security and backups to documentation and tests.

• Using Golang to Build Microservices at The Economist: A Retrospective - While it is useful to understand some of the high-level challenges of operating microservices, it can be equally useful to do a deep-dive into the day-to-day process of creating microservices. K Jonas provides a retrospective over three years of using Golang to build microservices at The Economist. Go was selected from a handful of possible languages because it closely aligned with the platform goals of moving quickly, helped enforce best practices around fast-failing services, and facilitated a platform that could perform at scale.

Free download