Question

Your team runs a Python service on Compute Engine that occasionally crashes with unhandled exceptions. Support engineers currently grep through raw logs to find recurring failures, which is slow. You want a managed way to automatically group similar stack traces, show how often each error occurs, and notify the team when a new error type first appears. Which Google Cloud capability should you use?

Accepted Answer

Use Error Reporting, which aggregates and deduplicates exceptions into groups and can notify you when a new error is first seen.

Answer

Create a log-based metric that counts every log entry containing the word 'error' and build a Cloud Monitoring dashboard from it.

Answer

Enable an uptime check against the service endpoint and configure an alerting policy on its availability.

Answer

Export all logs to BigQuery and run scheduled queries that GROUP BY the exception message text.