Google Professional Cloud DevOps Engineer Question 153
Select 4Google Cloud PlatformYour team is running a containerized application on Google Kubernetes Engine (GKE) and has started receiving complaints from users about increased latency and occasional HTTP 500 errors. You notice that CPU usage on the pods is consistently high, and error logs point to database connection timeouts. What actions should you take to address the issue effectively?
- A
Increase the CPU limits for the pods in the deployment.
- B
Scale the number of replicas for the deployment to handle the increased load.
- C
Investigate and optimize the database queries to reduce connection timeouts.
- D
Enable pod auto-scaling based on CPU and memory usage.
- E
Switch the database to a lower-cost tier to reduce operational costs.
Show answer and explanation
Correct answers: A, B, C, D
Explanation
The issue involves high CPU usage on pods and database connection timeouts, leading to increased latency and errors. Addressing these problems requires a combination of actions: increasing resource limits to mitigate immediate CPU constraints, scaling replicas to distribute the load, optimizing database queries to reduce connection timeouts, and enabling auto-scaling to adapt to traffic patterns dynamically. Switching to a lower-cost database tier does not address the root cause and may exacerbate the problem.
- A. Correct.
Increasing the CPU limits for the pods can provide more resources to handle the current load, potentially reducing latency and errors caused by CPU exhaustion.
- B. Correct.
Scaling the number of replicas can distribute the load across more pods, improving application performance under high traffic conditions.
- C. Correct.
Database connection timeouts might indicate inefficient queries or a database under stress. Investigating and optimizing queries can help reduce the load on the database.
- D. Correct.
Enabling pod auto-scaling ensures that the application can dynamically adjust to traffic spikes, preventing resource exhaustion and improving reliability.
- E. Incorrect.
Switching to a lower-cost database tier could reduce costs but does not address the performance issues causing the latency and errors. It might even worsen the problem if the lower-cost tier has fewer resources.