AZ-305 Question 169
Select 2You are designing a microservices-based e-commerce application that needs to handle large spikes in traffic during flash sales. The architecture must be able to scale automatically, support asynchronous communication between services, and cost-effectively run periodic background processing tasks. You plan to use Azure Kubernetes Service (AKS) for container orchestration of core microservices, and Azure Functions for event-driven scenarios such as lightweight background jobs. Which two additional design considerations should you incorporate to achieve these requirements?
- A
Use Azure Service Bus to decouple microservices and handle message-based communication asynchronously.
- B
Install a custom Windows service on each container to handle all background processing jobs.
- C
Implement the Horizontal Pod Autoscaler (HPA) in AKS to automatically scale microservices based on CPU or memory usage.
- D
Store and analyze logs locally on container disk to minimize storage costs and avoid dependency on external services.
- E
Run Azure Functions on a Consumption plan to automatically scale out and handle intermittent workloads.
Show answer and explanation
Correct answers: A, C
Explanation
To build a resilient, scalable microservices solution on Azure, it is crucial to incorporate asynchronous messaging for loosely coupled services and implement autoscaling mechanisms. Azure Service Bus helps handle asynchronous communication and ensures each microservice remains independent, allowing for better fault isolation. Using the Horizontal Pod Autoscaler in AKS automatically adjusts the number of pods to match resource demand, essential for unpredictable workloads. Refer to the Microsoft Azure Architecture Center for detailed best practices: https://docs.microsoft.com/azure/architecture/.
- A. Correct.
Correct. Using Azure Service Bus (or another messaging service) provides a robust way to decouple microservices and implement asynchronous communication. This is essential for handling spikes in traffic and reducing the risk of bottlenecks in a microservices architecture.
- B. Incorrect.
Incorrect. Installing a custom Windows service on each container contradicts the microservices approach and introduces unnecessary maintenance overhead. It is more efficient to leverage Azure Functions or dedicated microservices for background tasks rather than bundling everything into each container.
- C. Correct.
Correct. Implementing the Horizontal Pod Autoscaler in AKS is a best practice for automatically scaling out or in based on resource usage, such as CPU or memory. This ensures the microservices can handle spike loads during flash sales without manual intervention.
- D. Incorrect.
Incorrect. Storing logs solely on local container disks is risky because logs can be lost if containers are redeployed or scaled down. Centralized logging (e.g., Azure Monitor, Azure Log Analytics) is recommended for durability and ease of analysis.
- E. Incorrect.
Incorrect. While running Functions on a Consumption plan can be appropriate for many scenarios, in this case the question specifically focuses on adding messaging and autoscaling for the core microservices. Using Functions on the Consumption plan by itself is not enough to ensure asynchronous communication or AKS scaling.