AZ-305 Question 179
Single answerA logistics company needs to capture real-time delivery status updates from thousands of devices. The incoming events must be processed with low latency, visualized on a real-time dashboard, trigger notifications, and also be stored for long-term analytics. Which approach should you recommend to implement an event-driven architecture in Azure to meet these requirements?
- A
Use Azure Service Bus with a single queue to receive events, store events in Azure Table Storage, and poll the queue from a custom .NET service for dashboards.
- B
Use Azure Event Hubs for high-throughput event ingestion, Azure Stream Analytics for real-time analytics, and Azure Event Grid to route processed events to subscribers such as dashboards and notification systems.
- C
Use HTTP triggers in Azure Functions for all events, store them directly in Azure SQL Database, and use an Azure VM to poll for updates.
- D
Use Azure Data Factory pipelines triggered by scheduled intervals to fetch data from devices and load them into Azure Storage before notifying customers.
Show answer and explanation
Correct answer: B
Explanation
For real-time event processing and integration with multiple downstream consumers, combining Azure Event Hubs, Azure Stream Analytics, and Azure Event Grid is a recommended strategy. Event Hubs ingests high volumes of telemetry, Stream Analytics processes data in real time, and Event Grid routes processed events to relevant subscribers. Refer to Microsoft’s documentation on event-driven architectures (docs.microsoft.com/azure/event-grid) for more guidance on best practices and patterns.
- A. Incorrect.
Incorrect: While Azure Service Bus can handle enterprise messaging, relying on a single queue and polling from a custom service may introduce latency. Azure Table Storage does not provide real-time analytics capabilities, and you lose the flexible fan-out capabilities of an event-driven architecture.
- B. Correct.
Correct: Azure Event Hubs is designed for high-throughput event ingestion, making it ideal for device telemetry. Azure Stream Analytics can then process this data in near real-time, and Azure Event Grid can fan out the processed results to multiple subscribers, enabling dashboards, notifications, and downstream systems seamlessly.
- C. Incorrect.
Incorrect: Using only HTTP triggers in Azure Functions for massive device data ingestion can become a bottleneck, and storing everything directly in Azure SQL Database may not scale well for extremely high throughput. It also lacks the robust event routing that Azure Event Grid provides.
- D. Incorrect.
Incorrect: Azure Data Factory is an orchestration and data integration service that is primarily schedule-driven or pipeline-triggered. It is not optimized for near real-time event ingestion and processing, so this approach will not meet the low-latency requirements.