observability

Distributed tracing and structured logs are added by wrapping functions with trace() and using Pino for logging, eliminating the need to modify business logic. This provides end-to-end visibility and trace-log correlation across API calls.

OpenTelemetry semantic conventions are applied to background jobs and data processing pipelines by wrapping the logic with trace(). This generates spans, logs, and attributes to monitor async workflows and database queries.

Correlating Pino logs with OpenTelemetry spans is handled by the observability wrapper, which aligns structured logs with trace data. This enables cross-service diagnostics and performance monitoring without altering the core logic.

Autotel and pino are required dependencies to enable trace-log correlation. Installing these libraries allows you to wrap functions with trace() and produce structured logs following OpenTelemetry conventions.

Wrapping business logic with trace() provides visibility by capturing precise timing and error statuses across function executions. It solves the problem of unseen runtime behavior by generating spans and attributes for API calls and async workflows.

Trace wrappers for async workflows and database queries require the autotel and pino libraries to function. While they provide precise timing and error statuses, they are limited to producing OpenTelemetry-style spans and structured logs.