Dashboard Get started

Quick Start

One command sets everything up. You'll have live traces, logs, and exceptions in the dashboard in under two minutes — with zero code changes.

Registration is open — sign up for free at lightarch.io/sign-up. No credit card required.
1

Get your DSN

After signing up, your DSN is shown in the onboarding checklist at the top of the dashboard, and permanently under Settings → DSN. It looks like:

text
la_live_4d87ed8393887db2dd0c0843f31f498e
2

Run la init

In your project's root directory:

bash
npm install -g @lightarch/cli
la init --dsn la_live_YOUR_TOKEN_HERE

la init detects your runtime — Node.js, Express, Fastify, NestJS, Next.js, Python, Go, .NET, Java, Cloudflare Workers, and more — and does the entire setup:

  • Installs the OpenTelemetry auto-instrumentation package
  • Writes the required env vars to .env
  • Adds .env to .gitignore (your DSN is a secret)
  • Patches your npm start / dev scripts so it all loads automatically
Change OTEL_SERVICE_NAME in .env to your real service name (e.g. api-gateway) — it's how spans appear in the dashboard.
3

Start your app and make a request

bash
npm start

# In another terminal — any request works
curl http://localhost:3000/api/users

Open the dashboard — your first span appears within ~5 seconds and the onboarding checklist ticks itself. Or verify straight from the terminal:

bash
la query "SELECT service, severity, body, duration_ms FROM recent_spans ORDER BY timestamp DESC LIMIT 5"

Screenshot needed

Logs page showing first span: time=08:32:11, level=INFO, body='GET /api/users', service=my-app, status=200, 142ms. Onboarding checklist shows '🎉 First span received!'

What you get automatically

No code changes needed for any of this:

  • HTTP requests — every route, method, status code, and duration
  • Database queries — Postgres, MySQL, MongoDB, Redis with query timing
  • Outbound calls — fetch, http/https, gRPC, AWS SDK
  • Exceptions — unhandled errors with type, message, and stack trace, flagged in the Errors page
  • Logs — pino, winston, and bunyan output ships to the Logs page automatically
console.log is not captured — that's true of every OTel-based tool, including Datadog. Use a structured logger like pino and it flows automatically.

Next steps

Now that data is flowing, explore the platform:

  • Traces — Explore distributed traces and waterfall flamegraphs
  • Logs — Search logs with FTS5, live tail, detail panel
  • Metrics — Latency p50/p99, error rate, memory, CPU charts
  • Alerts — Set threshold and anomaly detection rules
  • Dashboards — Build custom panels from any SQL query
  • CLI — Full reference for la query, la trace fetch, la logs tail