Sub-domains and bounded contexts

2024-01-01 | 📑 206 words | ⏱ 3 mins | 🧾 History | ✍ NHS Notify | 🔑 NHS Notify

Known Issues / Todo

• ⛅ This page is draft and is subject to rapid change, and may not be fully accurate or complete

Sub-domains and bounded contexts

Bounded contexts are a pattern used in DDD to model parts of a system which operate on a single “unified” model of part of a business domain.

For NHS Notify we are attempting to keep our architecture modular following DDD principles. As a starting point when creating new bounded contexts these are some suggestions for structuring the AWS accounts and components which are deployed within them.

• By default, each sub-domain has its own AWS accounts
  • <sub-domain>-preprod (AKA dev)
  • <sub-domain>-prod
• Each sub-domain will have a top-level GitHub repo which defines the account-level infrastructure
• Each sub-domain will have one or more bounded contexts
  • Usually this revolves around some store of state
  • Subscriptions for events which affect that state
  • Handlers and processing for changes to state
  • APIs for direct actions on the state
  • Events published to reflect changes to the state
  • Auditing and logging to wrap the service and provide observability
• A sub-domain may contain one or more microservices
  • Each microservice should be individually deployable with defined contracts between services
  • Each microservice should be contained in its own GitHub repo