From Log Beginner to Pro: A QA’s Journey in Game Log Management and Monitoring

Introduction – The author shares a personal journey from being a complete newcomer to logs to becoming a skilled practitioner, aiming to help other beginners understand and master game logging.

What is a Log? – Logs are records of events, similar to diaries, containing time, actor, action, method, and result. In games, logs like GameRewardLog capture detailed player actions and outcomes, enabling reconstruction of events for QA and debugging.

Types of Logs – Two main categories are operational logs (structured, standardized, for cross‑team analysis) and developer/debug logs (free‑form, primarily for troubleshooting). Operational logs must be clear, single‑event focused, and include essential fields such as unique player ID and server ID.

Log Retrieval – Client logs reside in the application’s log directory; server logs are in the server’s log folder. In production, logs are accessed via a centralized log storage and query platform.

Operational Log Standards – The article outlines log grading (P1 mandatory, P2 recommended) and the basic JSON format: [logtime][operation],JSON with required fields like unique ID and server ID.

QA Workflow for Log Management – Steps include: (1) collecting external test logs, (2) reviewing code and specifications to identify missing or inconsistent fields, (3) consolidating best practices from external sources, (4) documenting and sharing standards via PPT and meetings, (5) creating tickets for each QA to track log issues, (6) ensuring new features include proper log requirements, and (7) performing detailed verification during delivery.

Log Monitoring – Monitoring is essential for post‑release stability. It involves configuring single‑log (e.g., GetLog) and multi‑log (e.g., money_update_log + item_buy) aggregations, defining thresholds, and setting alert frequencies. The author describes building a monitoring framework, handling long logs by splitting them, and using real‑world examples of abnormal behavior detection.

Collaboration with Planning and Development – Establishing a release‑wide “投放计划” (distribution plan), aligning log fields with design, and maintaining a shared monitoring checklist ensure consistency across versions.

Conclusion – Logging extends beyond basic record‑keeping to data analysis and automated testing; the author encourages readers to continue exploring and applying log‑driven practices to become proficient log engineers.