How-to guides¶

Each guide solves one problem and assumes you already have a provider building. If none of that is true yet, start with Getting started.

• Wire OTLP logs, metrics and traces — all three providers against one resource, and resolving one shared endpoint with per-signal overrides.
• Correlate logs with traces — make records carry trace_id and span_id, keep stderr output alongside OTLP, and put a level floor on the bridge.
• Diagnose telemetry that never arrives — the checklist for a service whose providers build without error and whose backend stays empty.

Guides that do not exist yet¶

Worth saying plainly, so you do not go looking:

• Exporting over OTLP/gRPC — not supported; only OTLP/HTTP exporters are built.
• Switching to a Prometheus, stdout or Jaeger exporter — not supported; the exporter is built internally and cannot be substituted.
• Instrumenting HTTP handlers, databases or queues — this module exports telemetry but produces none. Use otelhttp, otelgrpc and the OTel contrib instrumentation.

What this module does not do covers the full list and the reasoning.