README.md

November 10, 2025 · View on GitHub

Cloud Messaging Component

This component provides the real-time messaging infrastructure that enables scalable, event-driven communication between edge devices and cloud services in the Edge AI Accelerator solution. It implements a modern messaging architecture using Azure Event Grid and Event Hubs.

Purpose and Role

The Cloud Messaging component serves as the communication backbone for the Edge AI Accelerator:

  • Edge-to-Cloud Communication: Facilitates real-time data flow from edge devices to cloud services
  • Event-Driven Architecture: Enables decoupled, scalable system design with publish-subscribe patterns
  • Data Streaming: Provides high-throughput streaming capabilities for telemetry and AI inference data
  • Integration Hub: Connects edge devices, IoT Operations, data persistence, and analytics services

Component Resources

This component creates the following Azure messaging resources:

Event Grid Infrastructure

  • Event Grid Namespace: Modern Event Grid with enhanced capabilities and MQTT broker support
  • Event Topics: Organized event channels for different types of messages (telemetry, alerts, commands)
  • Event Subscriptions: Routing rules that direct events to appropriate consumers
  • MQTT Broker: Enables direct MQTT communication from edge devices

Event Hubs Infrastructure

  • Event Hubs Namespace: Container for high-throughput event streaming
  • Event Hubs: Individual event hubs for different data streams and processing requirements
  • Consumer Groups: Enable multiple independent consumers to process the same event stream
  • Partitioning: Provides scalable parallel processing of event streams

Integration Components

  • Azure Dataflow Examples: Sample dataflow configurations for common edge-to-cloud scenarios
  • Managed Identity Access: Secure authentication and authorization for messaging operations
  • Role Assignments: RBAC permissions for producers and consumers

Messaging Architecture

Data Flow Patterns

  1. Edge Publishing: Edge devices and Azure IoT Operations publish events to Event Grid topics
  2. Event Routing: Event Grid routes messages to appropriate Event Hubs based on subscription rules
  3. Stream Processing: Event Hubs provide buffering and fan-out for real-time and batch processing
  4. Consumer Integration: Downstream services consume events for storage, analytics, and response actions

Supported Message Types

  • Telemetry Data: Device sensors, system metrics, and operational data
  • AI Inference Results: Output from edge AI models and processing
  • Commands and Control: Cloud-to-edge messaging for device management
  • Alerts and Notifications: System health and anomaly detection events

Integration with Edge AI Accelerator

This messaging infrastructure integrates with all major components:

  • Edge Devices: Receive events via MQTT and HTTP protocols
  • Data Persistence: Events flow to storage and analytics systems
  • Observability: Message metrics and logs feed into monitoring dashboards
  • Applications: Custom applications can publish and consume events

Scalability and Performance

  • Auto-scaling: Both Event Grid and Event Hubs automatically scale based on load
  • Throughput Units: Configurable performance tiers for different workload requirements
  • Partitioning: Enables parallel processing and horizontal scaling
  • Global Distribution: Multi-region support for disaster recovery and low latency

Deployment Options

Terraform

Refer to Terraform Components - Getting Started for deployment instructions.

Learn more about the required configuration by reading the ./terraform/README.md

Bicep

Refer to Bicep Components - Getting Started for deployment instructions.

Learn more about the required configuration by reading the ./bicep/README.md

🤖 Crafted with precision by ✨Copilot following brilliant human instruction, then carefully refined by our team of discerning human reviewers.