A data team is operating an automated data processing pipeline deployed on Google Cloud. During a scheduled run, a task fails prematurely because custom transformation code encounters an unexpected schema variation and raises an unhandled exception. Which type of log in Cloud Logging should the data engineer inspect to examine the stack trace and debugging messages generated by the code?

Application logs: ### What Application Logs Are **Application logs** capture runtime events, diagnostics, status messages, and unhandled errors emitted directly by user code, container processes, or application software during execution. In Google Cloud services such as Cloud Run, Compute Engine, and Cloud Functions, standard output (`stdout`) and standard error (`stderr`) streams are automatically collected by **Cloud Logging** as application-level logs. ### How It Meets the Requirements - **Stack trace analysis (✓):** When custom code raises an unhandled exception or encounters runtime errors, the complete traceback and exception details are printed to standard error and preserved directly in the application log stream. - **Diagnostic context (✓):** Developers embed contextual log statements (e.g., `logger.error()`, `logger.info()`) within data processing scripts, allowing engineers to trace the exact input data or schema anomaly that caused the job to abort. - **Root-cause investigation (✓):** Application logs provide the granular, code-level visibility required to debug application logic flaws rather than cloud infrastructure or authentication issues. ### Key Advantages and Capabilities - **Granular severity levels:** Records events across standardized levels including `DEBUG`, `INFO`, `WARNING`, `ERROR`, and `CRITICAL`. - **Structured logging support:** Allows custom code to emit structured JSON payloads containing custom dimensions, such as pipeline task IDs, record batches, and runtime execution metrics. - **Log-based metrics:** Enables teams to create custom metrics and alerts in **Cloud Monitoring** based on specific application log error patterns. - **Real-time log exploration:** Supports filtering in the Logs Explorer by resource type, execution ID, and log payload text. ### Why This Is Optimal Application logs directly capture software behavior and code output. While platform audit logs or network logs show administrative API requests and network connections, only application logs contain the internal programmatic details and error stack traces necessary to fix data transformation failures.