Google Professional Cloud DevOps Engineer Question 158
Select 3Google Cloud PlatformYour team is responsible for managing a complex microservices-based application running on Google Kubernetes Engine (GKE). Recently, you have received intermittent reports of high latency from users, but you cannot identify the root cause using your current monitoring setup. After investigating, you find that logs are scattered across multiple services, metrics don't provide sufficient granularity, and there is no unified tracing system in place. What actions should you take to improve observability and diagnose the issue more effectively?
- A
Implement distributed tracing using OpenTelemetry and integrate it with Google Cloud Trace.
- B
Set up structured logging for all microservices and centralize logs in Google Cloud Logging.
- C
Increase the frequency of metrics collection to capture more granular data points.
- D
Deploy a network monitoring tool to analyze packet-level traffic between microservices.
- E
Use Google Cloud Profiler to identify latency bottlenecks in your application code.
Show answer and explanation
Correct answers: A, B, E
Explanation
To improve observability in a microservices-based application, you need tools that provide visibility across service interactions, application logs, and performance bottlenecks. Distributed tracing (e.g., OpenTelemetry with Cloud Trace) enables end-to-end tracking of requests across services. Structured logging and centralized logging in Google Cloud Logging make it easier to correlate events across services. Lastly, Google Cloud Profiler can pinpoint latency issues in the application code itself. These solutions work together to address the root cause of the observability issues.
- A. Correct.
Implementing distributed tracing with OpenTelemetry allows you to track requests as they traverse multiple services, making it easier to identify latency bottlenecks and dependencies.
- B. Correct.
Structured logging and centralizing logs in Google Cloud Logging improve observability by providing a unified and searchable source for logs from all microservices.
- C. Incorrect.
While increasing the frequency of metrics collection may provide more data, it does not address the root problem of insufficient granularity in existing metrics. Better logging and tracing tools are more effective in this scenario.
- D. Incorrect.
Deploying a network monitoring tool can be useful in specific cases (e.g., network-related issues), but it is not the most effective solution for diagnosing application latency in a microservices architecture.
- E. Correct.
Google Cloud Profiler helps identify performance bottlenecks in application code and can help diagnose issues contributing to high latency.