Collects detailed information about individual events — Every action inside the system — such as HTTP requests, database queries, errors, or authentication attempts — can generate a log entry.
Logs are centralized for analysis — They are typically shipped to centralized storage systems like Elasticsearch or Grafana Loki, where they can be searched, filtered, and correlated.
Provides very granular visibility into system behavior — Logs allow engineers to reconstruct exactly what happened during an incident, making them extremely valuable for debugging and root cause analysis.
Downside: high cost and complexity at scale — Storing and indexing every event generates massive volumes of data. As traffic grows, storage, processing, and retention costs increase significantly.