2V0-71.23 Question 293
Select 4A development team has deployed a microservices-based application to a Kubernetes cluster managed by VMware Tanzu. The application experiences varying levels of traffic throughout the day, and the team wants to ensure it scales efficiently to handle demand while minimizing resource costs. Which autoscaling functionality should the team configure, and under what circumstances should it be used?
- A
Horizontal Pod Autoscaler (HPA) to scale pods based on CPU or memory usage
- B
Vertical Pod Autoscaler (VPA) to adjust container resource requests and limits based on observed usage
- C
Cluster Autoscaler to add or remove worker nodes based on pending pods
- D
Static scaling by predefining the number of replicas for each pod
- E
Custom Metrics Autoscaler to scale pods based on application-defined metrics like request throughput
Show answer and explanation
Correct answers: A, B, C, E
Explanation
To effectively manage traffic fluctuations and minimize resource costs, the team should leverage multiple autoscaling functionalities. HPA handles scaling pods based on resource metrics, VPA ensures optimal resource allocation at the container level, and the Cluster Autoscaler ensures the cluster has enough capacity by adjusting worker nodes. Additionally, the Custom Metrics Autoscaler can scale workloads based on application-specific metrics, making it a versatile solution for dynamic traffic patterns. Static scaling, while simpler, does not dynamically adjust to traffic and is not relevant to the scenario.
- A. Correct.
Horizontal Pod Autoscaler (HPA) is suitable for dynamically scaling the number of pods based on resource metrics such as CPU or memory usage, allowing the application to handle demand spikes.
- B. Correct.
Vertical Pod Autoscaler (VPA) is used to adjust container resource requests and limits dynamically, ensuring optimal resource utilization without requiring manual intervention.
- C. Correct.
Cluster Autoscaler is designed to scale the number of worker nodes in the cluster when there are pending pods that cannot be scheduled due to insufficient resources.
- D. Incorrect.
Static scaling does not involve dynamic adjustments to traffic changes and is not considered an autoscaling mechanism. It requires manual intervention to resize workloads, making it unsuitable for the scenario described.
- E. Correct.
Custom Metrics Autoscaler allows scaling based on application-defined metrics such as request throughput or latency, providing a tailored approach to meet specific application needs.