Skip to content
Hosting Operations8 min read

Fix Website Down Not Working (2026)

Website down not working? Follow this security-hardening checklist to audit logs, lock down access, and verify your server is safe and reachable.

Written by Abdul AbrorTechnical Hosting Support Engineer
a computer screen with the words nothing great is made alone
On this page

TL;DR — Key takeaways

  • Start with a quick health check: confirm DNS, HTTP status, firewall rules, and recent config changes before assuming a breach.
  • Audit access paths first: check SSH, SFTP, database credentials, and API keys for unauthorized changes or expired tokens.
  • Harden the surface: disable unused services, enforce TLS, restrict admin panels, and keep the OS and web server patched.
  • Verify with safe tests: use a staging environment, take a snapshot, and roll back any change that breaks the live site.
  • Document the incident: record the timeline, actions taken, and who made each change to speed up recovery and review.

A website that is down not working usually points to a simple misconfiguration, a resource limit, or a security event. I have seen the same pattern repeat across shared hosting, VPS, and cloud instances. The fix is rarely a single magic command. It is a calm sequence of checks, a security audit, and a controlled rollout.

This guide covers the threat model, an audit checklist, hardening steps, and a safe way to verify the configuration. Use it as an evergreen operational reference when the site goes quiet.

Why the Site Goes Down: The Threat Model

A website stops responding for three broad reasons: the network path is broken, the application or server process is exhausted, and an attacker or misconfiguration has changed the environment. Each reason requires a different fix, so the first step is to separate them.

Network issues include DNS failures, expired domain registrations, or firewall rules that block the web port. Application issues include crashed processes, full disk space, or database connection limits. Security events include brute force attempts, unauthorized access, or a compromised plugin that drops the site.

Do not assume the worst. Check the basics first, then move to access controls, and only then treat the event as a security incident. This order saves time and prevents you from locking yourself out of a server that is already struggling.

  • Network path: DNS, firewall, load balancer, and ISP routing
  • Application state: process crashes, disk full, database locks
  • Security surface: unauthorized logins, changed files, new accounts

Quick Health Check Before You Change Anything

Run these checks in order. Each one takes under a minute and tells you which layer is failing.

First, confirm the domain resolves to the correct IP address. Use a public DNS lookup tool or run dig and nslookup from the command line. If the record points to the wrong server, the site will not load regardless of how healthy the origin is.

Second, check the HTTP status code. A 503 or 502 means the web server is up but cannot reach the application. A 404 usually means the domain is correct but the file path is missing. A 200 with a blank page often means the process is alive but the output is empty.

Third, review the web server error logs. Look for permission denied, missing modules, or database connection failures. These lines usually name the exact file or service that broke.

Fourth, verify disk space and memory. A full disk will stop the web server from writing logs or serving files. Run df and free on Linux, or check the equivalent in your control panel.

  • DNS lookup: confirm the domain points to the correct IP
  • HTTP status: 503/502 means upstream failure, 404 means missing file
  • Error logs: permission denied, missing modules, database failures
  • Resource limits: disk space, memory, and process count

Security Audit Checklist

Once the site is reachable, audit the access paths. Security hardening is not about adding more tools. It is about removing what should not be there.

Start with SSH and SFTP. Check the authentication logs for repeated failed logins, new key pairs, or accounts that were created recently. Disable password authentication if you use keys, and restrict SSH to known IP ranges when possible.

Review database credentials and API keys. Search the configuration files for hardcoded passwords, tokens, or connection strings that were exposed in a public repository. Rotate any key that may have been shared outside the team.

Check file permissions and ownership. Web files should not be writable by the web server user. A writable config file is a common entry point for attackers who want to inject code or change the site behavior.

Look for unexpected processes or scheduled tasks. A hidden cron job can phone home, mine cryptocurrency, or deface the site. Compare the running processes against your known application stack.

  • SSH: disable password auth, review failed logins, remove unknown keys
  • Database and API: rotate credentials, check for hardcoded secrets
  • File permissions: web user must not write config or code files
  • Processes and cron: verify no hidden tasks or unexpected binaries

Hardening Steps That Do Not Break the Site

Apply hardening in small batches. Each change should be tested before the next one is enabled. This prevents a single bad patch from taking the site down again.

Enable TLS and redirect HTTP to HTTPS. Use a valid certificate and force the redirect at the web server level. This stops downgrade attacks and protects user data in transit.

Restrict admin panels and management interfaces. Move them behind IP allowlists, require strong passwords, and disable directory listing. If the admin panel is exposed to the public, it becomes a target.

Disable unused services and modules. Every open port is a potential entry point. Turn off FTP if you use SFTP, remove unused plugins, and stop background workers that are not needed.

Keep the operating system, web server, and application dependencies patched. Set up automatic security updates for non-critical packages, but test them in a staging environment first. A broken kernel update can take the site down just as easily as a vulnerability.

  • TLS: enforce HTTPS and use a valid certificate
  • Admin panels: IP allowlist, strong passwords, disable directory listing
  • Services: disable FTP, remove unused plugins, stop unused workers
  • Patching: automate updates but test in staging before production

How to Verify the Configuration Is Secure

Verification is where most guides stop. Do not stop there. Confirm the changes actually improved the posture without breaking availability.

Run a port scan from an external network. Confirm only the required ports are open. If the web server listens on 443 and 80, that is normal. Anything else needs a reason.

Test the admin panel from a different network. It should not be reachable unless you are on the allowed IP range. If it is public, close it or move it behind a reverse proxy.

Check the logs after the changes. Look for a drop in failed login attempts and no new user accounts or file modifications. A secure server should show a quiet, predictable log pattern.

Perform a rollback if anything looks wrong. Keep the backup and snapshot available. Restore the previous configuration, verify the site works, and then decide whether the change was worth the risk.

  • External port scan: only required ports should be open
  • Admin panel: unreachable from public networks unless allowed
  • Log review: fewer failed logins, no new accounts or file changes
  • Rollback: restore the previous state if verification fails

Safe Testing Boundaries and Rollback

Risky operations need a safety net. Never apply a security change directly to production without a backup and a rollback plan.

Take a full backup of the site files, database, and configuration. Store the backup in a separate location. If the change breaks the site, restore the files and restart the service.

Create a snapshot or snapshot tag before the change. Some cloud platforms allow you to snapshot the instance. If the site does not recover, revert to the snapshot and investigate.

Test the change in a staging environment that mirrors production. Run the same commands, apply the same firewall rules, and check the same logs. Only move to production when the staging test passes.

Keep a change log. Record the date, the person who made the change, the command used, and the expected outcome. This helps you trace a future outage and proves you followed a controlled process.

  • Full backup: files, database, and configuration stored separately
  • Snapshot: create a rollback point before any risky change
  • Staging test: mirror production before touching the live site
  • Change log: date, person, command, and expected outcome

Quick troubleshooting checklist

  • Run a DNS lookup and confirm the domain resolves to the correct IP address
  • Check HTTP status codes and review the web server error logs for the last 24 hours
  • Verify firewall rules allow only required ports and block unnecessary inbound traffic
  • Review SSH and SFTP logs for failed logins, unexpected keys, or new user accounts
  • Check database credentials and API keys for unauthorized changes or expired tokens
  • Disable unused services, plugins, and modules that are not required for the site
  • Enforce TLS and redirect HTTP traffic to HTTPS before any further changes
  • Take a full backup and create a rollback snapshot before applying security patches
  • Test the site from a staging environment before applying hardening changes to production
  • Record the incident timeline, actions taken, and the person responsible for each step

FAQ

What should I check first when my website is down not working?

Check DNS resolution, HTTP status codes, firewall rules, and recent configuration changes. These four items explain most outages and are the fastest way to isolate the problem.

How do I harden a server without breaking the live site?

Take a full backup, create a rollback snapshot, and test all changes in a staging environment. Apply one security control at a time and verify the site stays reachable after each step.

What signs indicate a security incident rather than a simple outage?

Look for unexpected admin logins, new user accounts, changed file permissions, unusual outbound traffic, and error logs that show repeated failed access attempts.