AZ-305 Designing Microsoft Azure Infrastructure Solutions Exam

Recommend a Messaging Architecture

Align Messaging Patterns with Azure Services

Azure messaging services offer various patterns for the processing and decoupling of workloads. These messaging patterns include high-volume telemetry, ordered delivery, publish-subscribe distribution, and simple queue-based decoupling. To provide the optimal messaging architecture, architects must map these requirements such as throughput, latency, and delivery guarantees to services like Event Hubs, Service Bus, Event Grid, and Storage Queues.

The workload characteristics influence the selection of service:

• Event Hubs are ideal for handling high-volume telemetry with real-time streaming pipelines.
• Service Bus is suitable for ordered transactional delivery and reliable enterprise messaging.
• Event Grid effectively handles publish-subscribe event distribution with low-latency notifications.
• Storage Queues work best for simple queue-based decoupling and cost-effective buffering.

Throughput is a critical factor when choosing a service:

• Event Hubs can support millions of events per second.
• Storage Queues and Service Bus are more appropriate for smaller volumes.

When considering latency, note that:

• Event Grid delivers events nearly instantly.
• Queues require polling or long polling methods.

For delivery guarantees, each service has its specifics:

• Service Bus provides at-least-once delivery, FIFO ordering via sessions, and duplicate detection.
• Event Hubs and Storage Queues offer at-least-once delivery without default ordering.

Architects can combine these services to cater to complex requirements. For example:

• Use Service Bus for order processing.
• Utilize Event Hubs for telemetry ingestion.
• Deploy Event Grid to react to status changes promptly.

Furthermore, hybrid scenarios can connect on-premises systems with Azure using the Service Bus. Additionally, serverless compute options like Azure Functions or Logic Apps can orchestrate message flows and enrich data streams efficiently.

Conclusion

In conclusion, selecting the appropriate Azure messaging service depends on workload characteristics, including throughput, latency, and delivery guarantees. High-volume telemetry is best handled by Event Hubs, ordered transactional delivery is suited for Service Bus, and Event Grid excels in low-latency publish-subscribe scenarios. Simple queue-based decoupling is most cost-effective with Storage Queues. By understanding these services and how they can be combined, architects can recommend an effective messaging architecture that meets complex business needs seamlessly.