Know which endpoints are slow or failing

Every route your app serves gets its own page: how much traffic it gets, how fast it answers, how often it fails and the errors behind the failures. Counted from every request, with each deploy marked on the charts.

What you get

Every route, measured from every request.

Traffic, latency and errors per route

Requests per minute, p50 and p95 latency, error rate and a breakdown of 4xx and 5xx responses for each endpoint.

Counted from every request

Endpoint numbers come from every request Vigilon receives, before any trace is dropped, so they are not estimates from a sample.

Routes, not raw URLs

Requests are grouped by route template, such as /users/:id in Express or /api/orders/[id] in Next.js, so URLs with different IDs land on one endpoint.

The errors behind the failures

Each endpoint lists its recent errors and sample traces, so a rising error rate leads straight to the cause.

Deploys on the charts

Each release is marked on the endpoint charts from the app version in your telemetry, with nothing to connect to source control.

Slower endpoints flagged on Team and Pro

Automatic insights flag an endpoint whose slowest requests move into a slower band, and an app whose error rate spikes.

How it works

From the first request to a page per route.

  1. 01

    Install the SDK

    Add the Node.js or Python package, or send OTLP traces from an existing OpenTelemetry setup.

  2. 02

    Requests are grouped by route

    Each request is counted on the route template it matched, so URLs with different IDs land on one endpoint.

  3. 03

    Health checks stay out

    /health, /ready and /favicon.ico are left out by default, and you can exclude more paths.

  4. 04

    Open the slow one

    Compare endpoints by traffic, latency and error rate, then open one to see its errors and traces.

Setup

Every route is measured once the SDK starts.

Install the SDK and start it with your app; there is nothing to add per route. Leave out the paths you do not want measured with VIGILON_EXCLUDED_URLS.

Docs What Node.js tracksWhat Python tracksApp version for deploy markers

npm install @vigilon/node
export VIGILON_API_KEY="[YOUR_API_KEY]"
export VIGILON_EXCLUDED_URLS="/metrics,/internal/status"
node --import @vigilon/node/register server.js

/health, /ready and /favicon.ico are excluded by default. Excluded routes do not appear in Traces or Endpoints.

Compared with Sentry

Endpoint monitoring in Vigilon and Sentry.

Sentry builds endpoint summaries from sampled traces. This is how the two compare, with a source for every Sentry fact.

CapabilitySentryVigilon
Numbers per endpoint: traffic, latency, errorsFROM SAMPLED TRACESTransaction summaries built from tracing data [10]FROM ALL REQUESTSTraffic, p50 and p95 latency, error rate and status codes
Deploy markers on chartsRELEASES YOU NAMESet in Sentry.init or at build time [11]AUTOMATICRead from your app version; nothing to connect
What the default install turns onERRORSTracing, profiling and logs are switched on in code [9]ERRORS, TRACES, ENDPOINTSOne package and one start flag
Spans included each month5MOn every plan, about 12 requests a minute at 10 spans each; then billed per span, and dropped past your budget [1] [2]UNLIMITEDVigilon counts requests, however many spans each one has
  1. [1] Sentry pricing: spans, logs and cron monitors included on each plan: sentry.io/pricing. Checked 2026-10-25.
  2. [2] Sentry pricing docs: span rates, and data dropped past the budget: docs.sentry.io/pricing. Checked 2026-10-25.
  3. [9] Sentry Express setup guide: docs.sentry.io/platforms/javascript/guides/express. Checked 2026-10-25.
  4. [10] Sentry transaction summary: docs.sentry.io/product/insights/overview/transaction-summary. Checked 2026-10-25.
  5. [11] Sentry releases in Node.js: docs.sentry.io/platforms/javascript/guides/node/configuration/releases. Checked 2026-10-25.

Read the full Vigilon and Sentry comparison

Questions

Endpoint monitoring questions.

What is API endpoint monitoring?

Endpoint monitoring measures each route your app serves: how many requests it gets, how long they take and how many fail. Vigilon builds one page per route from every request your app receives, with the errors and traces behind its failures.

Are endpoint numbers sampled?

No. They are counted from every request Vigilon receives, before any trace is dropped. Only traces are sampled: every error and every slow request is kept, plus a sample of the fast ones.

Which latency numbers do I get?

p50 and p95 latency per endpoint, with requests per minute, error rate and a breakdown of 4xx and 5xx responses.

Can I leave health checks out?

Yes. /health, /ready and /favicon.ico are excluded by default, and VIGILON_EXCLUDED_URLS adds more. Excluded routes do not appear in Traces or Endpoints.

How does Vigilon know which deploy changed an endpoint?

It reads your app version from your telemetry, from package.json on Node.js or a setting on Python, and marks each new version on the charts. Nothing connects to your source control.

Do I get alerted when an endpoint gets slower?

On Team and Pro, an automatic insight opens when an endpoint's slowest requests move into a slower band and it is clearly slower on average. Insights that turn critical are emailed; the rest stay in the dashboard.

Get started

Start free for two months.

Pick a plan, add the SDK, and Vigilon starts watching your app. Starter is free for your first two months, with no card needed.