Microservices Architecture Patterns

Microservices, design is a great approach when some of our top and non-negotiable priorities align to below themes:

• Manageability -ability to correct and improve existing functionality with minimal impact to other components of the system
• Extensibility – Pluggable design with minimal impact to other components of the ecosystem
• Maintainability - scalability, reliability, availability, performance, and security
• Improved IT Governance, Faster Turnaround and Business Value Realisation- Lean Team with agile dynamics, small iterations delivering incremental value by way of implementing continuous delivery and quality control good practices
• Adaptability - Ability to invest and adopt modern technology trends while keeping technical debt in check

Service-oriented architecture (SOA) works on a broader enterprise level where each service is loosely coupled with the other and share a common Enterprise Service Bus (ESB) across all resources. Microservices, on the other hand, although loosely coupled come with a bounded context, meaning the relationship between each component and data is autonomous and communication takes place separately through application performance interfaces (APIs).

Since Microservices are fine grained services applications made using this architecture is quicker and easier to deploy. SOAs on the other hand, contain much larger and modular services giving lesser pliability during deployment.

Moving from the era of monolith, experiencing the evolution of Service Oriented Architectures, Microservices (fine-grained services with specific atomic responsibility) are a step further, a fork from SOA (which is most likely aligned to business processes) with distinct differences, primarily attributed to service structure, discovery & usage, asynchrony, decentralization of governance and data, eventually consistent distributed systems design principles.

If you are organization keeping microservices at the core of your software delivery plan of action, then here are a few things you may want to take note of:

• Utilizing your existing infrastructure and reducing costs: This can be done with microservices, using containers on a shared hardware you can run more services for your customers.
• Opting to go serverless: Apart from using containers, you can also leverage advancement in technology - serverless, a whole new set of frameworks where source codes, functions etc. can be directly used during your execution.
• Choosing the right abstraction layer: If you opted for containerization then Kubernetes and Docker act as your de facto container orchestration platforms where you can get a great deal of flexibility and speed. Going towards Functions as a service (FaaS) or serverless, you choose to create a whole set of Application Programming Interface (APIs) for larger and more complex applications.
• Researching on data service platforms: During the selection procedure, go for platforms that have good support to run data services. Serverless functions are stateless but in case your app requires a constant state then platforms that run on Kubernetes can be a choice.