nianton / activity-stream-processing

Activity Stream Processing

This is a templated deployment of a sample Azure architecture for implementing an User activity/event stream processing, having all the related PaaS components.

The architecture of the solution is as depicted on the following diagram:

The role of each component

• Ingestion Function App -public facing HTTP endpoint (or Kafka consumer)
• Event Hubs -event streaming service
• Steam Analytics -to aggregate specific event that needed to be further consolidated before being processed
• Processor & Notification Function Apps -Function Applications to process the incoming events and notify downstream systems respectively
• Application Insights to provide monitoring and visibility for the health and performance of the application
• Service Bus -reliable message broker to decouple processing and notification flows, external APIs may need throttling etc
• Azure Key Vault responsible to securely store the secrets/credentials for the PaaS services

In an enterprise grade environment, all PaaS services can leverage Private Endpoints to be accessible only within a private Virtual Network and App Gateway (WAF) or Azure Front Door can be used to accept the incoming requests (not included in the present deployment).

---

Based on the template repository (https://github.com/nianton/bicep-starter) to get started with an bicep infrastructure-as-code project, including the azure naming module to facilitate naming conventions.

For the full reference of the supported naming for Azure resource types, head to the main module repository: https://github.com/nianton/azure-naming