Observability Overview

Observability gives engineering visibility into production behavior — what broke, why it broke, and who was affected. We use three complementary tools, each with a distinct role.

The three pillars

Tool	What it captures	When you use it
OpenTelemetry	Traces, metrics, logs (raw signals)	Instrument code, propagate context across services
Sentry	Errors, crashes, performance issues	Debug exceptions, track release health
Axiom	Structured logs and events	Search logs, build dashboards, investigate incidents

How they work together

1. OpenTelemetry instruments your code and exports standardized telemetry (traces, metrics, logs).
2. Sentry receives error events and performance spans — surfaces stack traces, breadcrumbs, and release regressions.
3. Axiom ingests structured logs for search, alerting, and long-term retention.

Environment configuration

Variable	Alpha	Beta	Prod
SENTRY_DSN	Dev project DSN	Staging DSN	Production DSN
OTEL_EXPORTER_OTLP_ENDPOINT	Local collector or disabled	Staging collector	Prod collector
AXIOM_TOKEN	Dev dataset token	Beta dataset token	Prod dataset token
AXIOM_DATASET	logs-alpha	logs-beta	logs-prod

Never use production DSNs or tokens in local development.

Quick links

Guide	Description
OpenTelemetry	Instrumentation setup for Node.js and Next.js
Sentry	Error tracking and release monitoring
Axiom	Log shipping and query patterns

Official documentation

• OpenTelemetry Docs
• Sentry Docs
• Axiom Docs