Design Principles For Microservices Architecture

Kanika Modi
3 min readApr 13, 2022

We’ve all experienced the disparity between the monolithic architecture vs microservices architecture. This blog covers the fundamentals of microservices architecture and design principles to keep in mind when working on microservices patterns and its challenges.

What Is Microservices Architecture?

Microservices architecture is a service-oriented architecture in which an application is structured as a collection of small autonomous services designed around a business domain. Microservices can be compared to an organ, with the total system representing the human body. It’s a “share as little as possible” design in which each service is self-contained and implements a single business capability with clear accountability. The entire system is built on a high cohesion and low coupling philosophy to provide a more scalable and available system; the unavailability of one component or service does not affect all of the system’s use cases.

Why Microservices Architecture?

Monolithic services have a number of drawbacks, including being inflexible for diverse technologies, being unreliable as a whole system, being unscalable, and preventing parallel development. There are numerous benefits to using the microservices pattern to handle these issues:

  • Fault Isolation: Since the system’s services are mostly detached, even if one of the application’s services fails, few of the system’s capabilities will continue to function, allowing for more resilient systems to be built.
  • Loosely Coupled: The services are as independent as possible from one another, allowing the entire program to be constructed and modified without affecting any unrelated code components or services.
  • Agility and Speed of Iteration: Multiple developers and teams can deliver relatively independently of each other under this model, thus increasing development parallelism, increases feature delivery speed.
  • Ownership: Applications are treated as products by microservices, for which they are responsible. Each service is in charge of a single business capability.
  • Decentralized Governance: Because there is no need for a standardized pattern, developers can choose the optimal hardware/tool/technology for each service independently.
  • Granular Scaling: The individual service can be scaled as required.

Why Not Microservices Architecture?

Before you start designing microservices architecture, you should be aware of the following caveats:

  • Operational Load: Due to fine-grained components, the system must be operationally ready.
  • Difficult to monitor: Without a request tracing method across components, monitoring the entire system and identifying a specific bug is challenging.
  • Testing scope: It is hard to set up an end-to-end integration test environment covering all components.
  • Eventual Consistency: Since microservices don’t share data store, in case the data across data stores need be synchronized, this process yields an eventual consistency design.

Tips For Designing Microservices

For a well-designed microservices architecture, follow the below principles:

  • Functional Cohesion: Each microservice has its own set of responsibilities and functions. The same functionality is not duplicated in more than one microservice.
  • Common Closure Principle: Classes that change for the same reason should be in the same package. Perhaps, for instance, two classes implement different aspects of the same business rule.
  • Data Isolation: Each microservice is responsible for its own data (both database and cache) and does not have direct access to the data of other microservices. The persistent data of each microservice must be private to that service and only available via its API.
  • Clean interfaces: Microservice APIs should be clean and not include implementation details. The message structure in async messaging must be explicitly defined.
  • Compute Isolation: Each microservice owns its compute resources. Two microservices do not run on the same instance.
  • Deployment Isolation: Each microservice can be tested and deployed to production individually.
  • Backward compatibility: Microservices must ensure that their APIs are backward compatible so that depended microservices do not break. Microservices should have a version for breaking changes.

Choosing the right size of component is the hardest task in Microservices Architecture, one has to choose the right fit size so that it should not result low performance system overall and more operational burden to the team.

Thank you for reading! If you found this helpful, here are some next steps you can take:

  1. This blog is part of my System Design Series. Check out the other blogs of the series!
  2. Send some claps my way! 👏
  3. Follow me on Medium & subscribe below to get a notification whenever I publish! 📨

--

--

Kanika Modi

Software Engineer @amazon | Designing Systems At Scale | Tech & Career Blogs | She/Her | ViewsMine()