1. Home
  2. Guides
  3. Website down: what to check
Performance guide

Website down? Common causes and what to check first

When a website stops loading, the cause is almost always one of a few things: the domain, the connection, the certificate, or the server itself. Work through them in order and you can usually find the failing layer in minutes.

Key points

  • First confirm the site is down for everyone, not just you, using a check from outside your own network.
  • Then test the layers in order: DNS, connection, HTTPS certificate, HTTP status. The first one that fails points to the cause.
  • Short outages rarely have a lasting search impact, but repeated or long ones can. Use a 503 for planned maintenance and set up monitoring.

Is it down for everyone?

Your own network, browser cache, VPN or DNS cache can make a working site look dead. Before changing anything, run the address through the website down checker, which tests from our server rather than your connection. Also try your phone on mobile data. If the site loads elsewhere, the problem is local: clear the cache, try another browser, or flush your DNS cache.

The layers that can fail

  1. DNS. The domain name must resolve to a server address. If the domain has expired, the DNS records were deleted or the name servers are wrong, nobody can reach the site. Symptom: a "server not found" or "DNS_PROBE_FINISHED_NXDOMAIN" error.
  2. Connection. The server must accept connections on ports 80 and 443. A stopped web server, a full machine, a cloud firewall rule or a provider outage all cause timeouts or "connection refused".
  3. HTTPS certificate. An expired or mismatched certificate triggers a full-screen browser warning even if the server is healthy. Visitors cannot easily continue.
  4. HTTP response. The server answers, but with an error code. This is where most "site is broken" reports end up.

What the error codes tell you

CodeMeaningTypical causes
500Internal Server ErrorA code or configuration error: a bad deploy, a broken plugin, a syntax error in a config file.
502Bad GatewayA proxy or CDN got an invalid reply from the application behind it, often because the app crashed.
503Service UnavailableOverloaded or in maintenance, or the application is not running.
504Gateway TimeoutThe application or database took too long to answer.
403ForbiddenFile permissions, a firewall, or a bot-protection rule blocking the request.
404Not FoundThe page or its route is missing, for example after a bad deploy or a rewritten URL rule.

A checklist to work through

  1. Did anything change recently? A deploy, plugin update, DNS change or certificate renewal in the last hours is the most likely cause. Roll it back if you can.
  2. Domain. Check the registrar account: is the domain active and paid up? Do the name servers match what your host gave you?
  3. Hosting. Look at your provider's status page and dashboard. Is the server or container running? Has it hit CPU, memory or disk limits, or a bandwidth quota?
  4. Certificate. Check its expiry date and that auto-renewal still works.
  5. Logs. The web server error log and the application log usually name the failure directly: a missing file, a database connection error, an out-of-memory kill.
  6. Database and dependencies. A stopped or overloaded database makes many sites return 500 or 503 errors.
  7. CDN or firewall. If you use one, bypass it temporarily or check its rules. A wrongly set rule can block real visitors or crawlers.
  8. Restart carefully. Restarting the web server or application often restores service, but note what you saw first so the cause is not lost.

If the checker shows a firewall or bot-protection page instead of your site, the site itself may be fine: the protection is blocking automated tests. Compare with a normal browser visit.

Does downtime hurt search rankings?

  • A brief outage typically has no lasting effect. Google retries pages later.
  • When a site returns server errors, Google slows its crawling. If errors continue for a long time, pages can drop out of the index until the site recovers.
  • For planned maintenance, return a 503 status with a Retry-After header, rather than a 200 page saying "back soon". This tells crawlers the outage is temporary.
  • After recovery, check your key pages load, and check redirects and robots.txt were not changed by accident.

How to reduce the risk

  • Monitor from outside. Use an uptime monitor that checks your site every few minutes and alerts you by email or phone. Many hosts and services offer free plans.
  • Turn on auto-renewal for the domain and the certificate, and keep a current payment method on both.
  • Keep backups and know how to restore them. Test a restore at least once.
  • Stage changes. Test updates on a copy of the site before applying them live, and keep a rollback path.
  • Keep headroom. Caching, a CDN and enough memory help a site survive traffic spikes. See compression and caching.
  • Look at the raw response when something looks wrong: status, redirects and timing often show the problem at once.

FAQ

How can I tell if my website is down for everyone?

Test it from a network other than your own, for example with an online checker or your phone on mobile data. If those load the site, the problem is local to your device or network.

What does a 502 or 504 error mean?

Both come from a server that sits in front of your application, such as a proxy or CDN. A 502 means it received an invalid reply, often because the application crashed. A 504 means the application did not answer in time.

My site shows a certificate warning. Is it down?

The server is probably running, but visitors are blocked by the browser warning, which is effectively an outage. Renew the certificate or fix the name mismatch.

Will downtime remove my site from Google?

Short outages do not. Long or repeated server errors can lead to pages being dropped from the index until the site is stable again, and rankings can take time to recover.

What should I show visitors during maintenance?

A simple page that explains the work and when you expect to be back, served with a 503 status and a Retry-After header so search engines know the outage is temporary.

WebsiteHealthLabs Editorial
Guides explain audit results in plain English. Facts are checked against public documentation from Google, W3C and the browser vendors, and we update them when guidance changes.

↑ ↓ to moveEnter to openEsc to closeAll results