BigSmokes Cigar is a leader in elite premium cigars that are sold around the globe. Headquartered in Florida, USA, they have been using an on-prem solution for their ecommerce website. This website is their primary source of orders and is backed by microservices. During cyclone Nirma, they discovered how vulnerable their data center is. The data center was wiped out due to the floods and BigSmokes went offline and hence no revenue generated for 2 weeks. This loss is perceived to be a tiny fraction of what BigSmokes will have to spend on Cloud Migration. They have chosen AWS as their cloud partner and are looking for a proposal from Cognizant to migrate their On-prem site to the cloud. The solution must be best-in-class, but at the same time convince the “very cost conservative” CFO who has to approve the finance. The decisive factors of a winning architecture will be
- Highly Resilient: Even if the American Continent is engulfed in floods, they expect to be in business. The have factories worldwide, so manufacturing isn’t a problem
- Highly Available: 99.9% available. No noticeable outage and hence no lost orders
- Performance: especially response to user requests should never go over a second
- Cost should not be prohibitive: While they can spend any amounts of money, the CFO is very particular that the solution MUST cost lesser than the On-Prem Solution
- Value additions: Innovation that would buy a better bang for the bucks shelled out by BigSmokes.
This is a high-level view of how the different microservices interact with each other. Each service diagram illustrates individual service stack along with endpoints meant to interact with other services.
Communication/Messaging:
- AWS AppSync for interactions between users and the ecommerce platform.
- Amazon API Gateway for service-to-service synchronous communication (request/response).
- Amazon EventBridge for service-to-service asynchronous communication (emitting and reacting to events).
Authentication/Authorization:
- Amazon Cognito for managing and authenticating users, and providing JSON web tokens used by services.
- AWS Identity and Access Management for service-to-service authorization, either between microservices (e.g. authorize to call an Amazon API Gateway REST endpoint), or within a microservice (e.g. granting a Lambda function the permission to read from a DynamoDB table).
Compute:
- AWS Lambda as serverless compute either behind APIs or to react to asynchronous events.
Storage:
- Amazon DynamoDB as a scalable NoSQL database for persisting informations.
Monitoring:
- Amazon CloudWatch for metrics, dashboards, log aggregation.
| Services | Description |
|---|---|
| users | Provides user management and authentication. |
| products | Source of truth for products information. |
| orders | Manages order creation and status. |
| warehouse | Manages inventory and packaging orders. |
| delivery | Manages shipping and tracking packages. |
| payment | Manages payment collection and refunds. |
| Services | Description |
|---|---|
| frontend-hosting | Hosting platform to serve static HTML, CSS, JS, images and other resouces. |
| frontend-api | User-facing API for interacting with the services. |
| Services | Description |
|---|---|
| platform | Core platform resources for deploying backend services. |
| monitor & log | Log service metrics for future diagnostics. |
| management & governance | IAM Roles for maintenance, management, governance and configuration of the architecture |
The orders service acts a the single source of truth for data related to an order: delivery address, products, user, etc.
As orders are (somewhat) immutables, information about the delivery address and products are replicated within the order to ensure consistency over time. In some cases (e.g. if a product cannot be packaged, is substituted to another item, etc.), the order might be modified. This service also monitors the change in state from other services (mainly delivery and warehouse) so that users can probe the status of their order.
When a user creates an order, this service also acts as a central gate that contacts other services to verify that the user input is valid. For example, this checks that the products exist and that the prices are correct. If any check fails, this will return an error to the end-user.
The warehouse service handles packaging incoming orders into packages ready for delivery.
