Question

A payments team runs a critical microservice on Cloud Run. The application writes structured JSON logs to Cloud Logging, and certain business failures are logged with a field severity=ERROR and a jsonPayload field txn_status="declined_downstream". Operations wants to be alerted whenever the rate of these specific downstream-decline errors exceeds a threshold within a rolling window, without changing application code and without exporting logs to an external system. What is the most appropriate approach?

Accepted Answer

Create a log-based metric (counter) filtered on the specific jsonPayload field and severity, then build a Cloud Monitoring alerting policy on that metric's rate over a rolling window.

Answer

Configure a log sink to export all logs to BigQuery and schedule a query that emails the team when the decline count is high.

Answer

Enable Error Reporting and rely on its automatic grouping to notify the team when the decline error group grows.

Answer

Set up a log sink to Pub/Sub and write a Cloud Function that parses each message and calls an alerting webhook when a decline is detected.