Google Professional Cloud DevOps Engineer Question 151
Select 3Google Cloud PlatformYour team is managing a production environment hosted on Google Cloud. Recently, users have reported intermittent service outages. Upon investigating, you notice that the Compute Engine instances hosting critical services are occasionally becoming unresponsive due to resource exhaustion. To address this issue, which actions should you take to improve the reliability of your infrastructure?
- A
Set up Managed Instance Groups (MIGs) with autoscaling enabled based on CPU or memory utilization thresholds.
- B
Use Cloud Monitoring to establish alerts for high resource utilization and integrate them with Cloud Operations Suite for further analysis.
- C
Manually resize the Compute Engine instances to larger machine types whenever resource exhaustion occurs.
- D
Implement preemptible VM instances in your environment to handle resource spikes.
- E
Configure load balancing to distribute traffic across multiple instances and regions for improved fault tolerance.
Show answer and explanation
Correct answers: A, B, E
Explanation
To address infrastructure issues such as resource exhaustion, it's important to implement scalable and proactive solutions. Managed Instance Groups with autoscaling dynamically adjust resources, while monitoring and alerting help you detect and resolve issues before they impact users. Load balancing ensures traffic is distributed, improving fault tolerance and reliability. Reactive or unsuitable solutions, such as manually resizing instances or using preemptible VMs for critical workloads, do not effectively address the problem.
- A. Correct.
Correct. Managed Instance Groups (MIGs) with autoscaling can dynamically adjust the number of instances based on resource utilization, ensuring your infrastructure scales to meet demand.
- B. Correct.
Correct. Monitoring and alerting help you detect and respond to resource exhaustion issues proactively, improving overall system reliability.
- C. Incorrect.
Incorrect. Manually resizing instances is reactive, labor-intensive, and not an efficient long-term solution for resource exhaustion issues.
- D. Incorrect.
Incorrect. Preemptible VMs are low-cost but are not suitable for handling critical production workloads as they can be terminated at any time.
- E. Correct.
Correct. Load balancing improves fault tolerance by distributing traffic across multiple instances and regions, reducing the risk of unresponsiveness due to resource exhaustion on a single instance.