logging-best-practices

Wide events, or canonical log lines, are single comprehensive log events per request containing all relevant context. You need them when moving beyond scattered, low-context logs to a centralized, queryable data source for enhanced debugging and observability.

Implement canonical log lines by emitting a single, structured JSON log event per request. Include high-cardinality, high-dimensionality data, business context, and environment features to ensure logs capture comprehensive debugging information across services.

Include high cardinality, high dimensionality data, business context, and environmental features in structured logs. For a critical checkout failure, capturing user subscription level, cart value, and feature flag status enables rapid identification of the root cause and business impact.

Yes, utilizing middleware is recommended for consistent event generation across services. The approach also advises employing a single logger instance to standardize the JSON structured logs and maintain uniform observability data throughout the application.

The best way to structure application logs is using JSON format to create wide events. This approach centralizes queryable data by combining high-cardinality dimensions, business context, and environmental features into a single comprehensive log event per request.

Scattered, low-context logging fails because it lacks the high cardinality and business context needed to trace application behavior. Without structured wide events, identifying the root cause and business impact of critical failures requires manual correlation across fragmented logs.