Solution architecture
Download as PPTX, PDF0 likes1,019 views
- The solution proposes a cloud-based e-commerce application using a microservices architecture hosted on Azure. Key services include Azure WAF, VPN, subnets, API Management, Azure AD/OAuth 2.0, Azure Cosmos DB, and Azure Media Services. - The application would be broken into bounded contexts and microservices for functions like search, browse, cart, orders, recommendations, and administration. Services like Elasticsearch, Redis, Cassandra, and SQL would be used for data storage. - High risks include cost optimization on the cloud, testing environments, infrastructure as code, microservices communication complexity, training on cloud technologies, and implementing continuous integration/deployment pipelines.
Solution architecture
- 1. Solution Approach - Cloud based application - Secured from Internet through Azure WAF, VPN, subnets & API Management using Azure AD/OAuth 2.0 - Microservices architecture - NoSQL datastore like Azure Cosmos, RavenDB with features like external replication for horizontal scaling/scale-out functionality to support millions and millions of customers hitting the website.
- 2. Logical architecture Bounded Context Search Browse Cart Orders Recommendations Wish List Admin .WAR or .EAR file Search – 5 million requests per minute Recommendations – Compute heavy Orders – Mostly Writes, few Queries Admin – 5 requests per minute Search ElasticSearch Bounded Context Browse Redis Bounded Context Recommendations Cassandra Bounded Context Reviews SQL Bounded Context Cart Redis Bounded Context Orders SQL Modeling Product group category SKU Monolith application Microservices application Shipping address customer Shipping company Order sku customer address record Performance Scalability Capacity Availability Reliability Recoverability Maintainability Servicability Security Regulatory Manageability Environmental Data Integrity Usability Interoperability NFR’s
- 3. Technical Architecture Azure Media Service for audio/video streaming
- 4. High level risks in such a program - Cost optimization (TCO) on Cloud - Testing environments - Infrastructure-as-a-code - Complexity in microservices communication - Need to go for Service Mesh/Istio for cross-cutting concerns like logging, configuration, security, service discovery etc. - Training on Cloud/New Technologies for developers - Continuous Integration/Continuous Depoyment(CI/CD) using Azure DevOps. Setting up build and release pipeline. - DevSecOps with code on demand and cloud native applications - 12 factor cloud native application - Inventory availability from various vendors - API integration issues with various vendors