Google Professional Cloud DevOps Engineer Question 137
Select 3Google Cloud PlatformYour team recently deployed a new application to Google Cloud, and you suspect that an increase in latency is causing degraded user experience. You want to use Metrics Explorer to perform an ad hoc analysis of the application's latency metrics to identify patterns or anomalies. Which of the following steps should you take to investigate this issue effectively?
- A
Filter the metrics by specific labels such as instance ID or region to isolate the affected resources.
- B
Use the 'Group By' functionality to aggregate metrics by labels like service name or zone.
- C
Set up an alerting policy in Metrics Explorer to notify you of future latency spikes.
- D
Adjust the time range to include both recent and historical data to identify trends.
- E
Export the metrics to BigQuery to perform ad hoc analysis using SQL queries.
Show answer and explanation
Correct answers: A, B, D
Explanation
To effectively use Metrics Explorer for ad hoc latency analysis, you should filter the metrics by labels to focus on the impacted resources, group the data by meaningful dimensions to uncover patterns, and adjust the time range to analyze both recent and historical data. These steps allow you to leverage Metrics Explorer's capabilities to identify patterns and anomalies in real-time data without requiring external tools or configurations.
- A. Correct.
Filtering metrics by specific labels helps narrow down the analysis to the affected resources, making it easier to identify the root cause of latency issues.
- B. Correct.
Grouping metrics by labels allows you to compare aggregated data across different dimensions (e.g., service name or zone), which can help identify patterns or anomalies.
- C. Incorrect.
Setting up an alerting policy is not part of ad hoc analysis; it is a proactive monitoring step that is configured separately from Metrics Explorer.
- D. Correct.
Adjusting the time range to include both recent and historical data enables you to compare current performance against past trends, which is crucial for identifying anomalies.
- E. Incorrect.
Exporting metrics to BigQuery is not necessary for ad hoc analysis in Metrics Explorer; this would be more appropriate for advanced or long-term analysis.