Scalable Cloud Architecture for Synchronizing Pharmacy Inventory Between Central and Local Systems

This article demonstrates how to build a pharmacy inventory synchronization application between a central system and localish branches onMicrosoft Azure (instead of AWS Lambda, SQS, DynamoDB).The solution uses Azure Virtual Machines, Azure App Services, Azure Functions, Azure Kubernetes Service (AKS), Azure Cosmos DB, Azure SQL Database, Azure Service Bus, Azure Cache for Redis, and Azure Logic Apps.It enables near real-time inventory synchronization, conflict resolution, and scalable loading for various loads.Local stores further promote inventory events through Azure Functions that are dispatched by Service Bus queues or Event Grid.The heart of the inventory system is a database in Azure SQL (relational) and Cosmos DB (distributed cache world-wide).AKS microservices orchestrate synchronization logic; hot-data access is accelerated by Azure Cache for Redis.Telemetry in Azure Monitor and Application Insights getch the data.We did experiments with 20 simulated branch nodes where we varied the transaction load (50-1000 ops/min) for 24 hours.At peak throughput (1000 ops/min across branches), 99.9% over 24h Auto-scaling in App Services and AKS kept CPU utilization around ~60-70% On both cases, Azure Service Bus queue length never exceeded 500 messages.Azure Monitor dashboards made it Rajesh Karvannan https://iaeme.com/