logging-best-practices

Wide events in structured logging emit a single, complete JSON log per request per service to minimize noise and maximize observability. This approach captures comprehensive business and environment context to support unknown-unknown debugging.

To standardize structured logging across distributed APIs, configure a single shared logger and middleware to automatically emit one JSON wide event per request. This ensures consistent context collection and simplifies tracing across service hops.

The best way to capture high cardinality and dimensionality in logs is by using middleware to auto-collect environment data and timing. This enforces standardized wide events that include business and environment context for deep observability.

Yes, you need middleware to implement canonical log lines effectively. Middleware integrates with a shared logger to automatically collect environment and request context, ensuring a single queryable log per request for unified observability.

You should not use a single wide event per request when your services generate exceptionally high request volumes or when logging complex asynchronous background jobs, as this approach is designed specifically for request-bound endpoints to simplify tracing.