HTTP Status Codes: A Practical Reference for Monitoring and Debugging.

Every HTTP response starts with a three-digit status code. It is the first thing a monitor looks at, and the first clue when something breaks. This reference covers the codes you will actually meet and what each one usually means in production.

Key takeaways

• 2xx success, 3xx redirect, 4xx client-side problem, 5xx server-side problem.
• 502, 503 and 504 usually point at a proxy or load balancer that can't reach the application.
• A 200 doesn't prove the page is right - check its content too.
• Decide per monitor whether redirects should be followed.
• Repeated 429 responses mean you are being rate-limited, not that the site is down.

The five classes

2xx: success

3xx: redirects

Redirect chains (HTTP to HTTPS to www to a language path) add latency to every first visit. Loops - A redirects to B, B back to A - break the site completely; browsers give up after about 20 hops.

4xx: client errors

Not every 4xx is the client's fault in practice. A sudden flood of 404s after a release usually means a broken route; a 403 for everyone often means a WAF rule went wrong.

5xx: server errors

When you see 502/504, look behind the proxy: is the application process running, listening on the expected port, and fast enough?

Configuring monitors

What counts as "up"

Most monitors treat 2xx and 3xx as up and 4xx/5xx as down. Adjust per endpoint:

• An admin page that should demand a login: expect 401 or 302 to the login page.
• A health endpoint: expect exactly 200.
• An API that returns 204 for health checks: allow it explicitly.

Follow redirects or not

Following redirects tests the page users finally land on. Not following them lets you verify the redirect itself - for example, that HTTP still redirects to HTTPS.

Check content too

A login page that says "Database connection failed" often still returns 200. Combine status checks with keyword monitoring for pages that matter.

You can check a site's current status code instantly with the website down checker.

Frequently asked questions

• Is a 404 an outage?

  +

  For one missing page, no. For your homepage, checkout or API route, yes - a deploy that breaks routing produces exactly that.

• What is the difference between 502 and 504?

  +

  502 means the proxy got a bad or no response at all (often a refused connection). 504 means it waited and the application took too long.

• Should 429 trigger an alert?

  +

  Usually not as an outage. If your monitor gets 429, ask the site owner to allow the monitoring IPs or lower the check frequency.