I build digital products with AI.

Internal Server Error

The server encountered an internal error or misconfiguration
and was unable to complete your request.

Additionally, a 500 Internal Server Error error was encountered
while trying to use an ErrorDocument to handle the request.

What it is

A fatal PHP error, an .htaccess directive the server does not support, a missing Apache module, or files the web server is not allowed to read. That second paragraph is its own clue — the custom error page is broken too, so the real message never reaches the browser.

What I do

Read the server error log instead of the browser. Take .htaccess out of the path to isolate it, switch PHP error logging on, check ownership and permissions — 644 for files, 755 for directories — then fix the one failing line and put the error page back.

Forbidden

You don't have permission to access this resource.

Additionally, a 403 Forbidden error was encountered while
trying to use an ErrorDocument to handle the request.

What it is

Ownership or permissions left wrong by a migration, an index file missing while directory listing is switched off, a deny rule in .htaccess that is broader than it was meant to be, or a ModSecurity rule blocking a request that was perfectly legitimate.

What I do

Find the rule ID or the path in the log, restore correct ownership for the web user, put the index file or the Options directive back, and whitelist the single rule that fired rather than switching the firewall off.

502 Bad Gateway
nginx

504 Gateway Time-out
nginx

What it is

Nginx is healthy; the PHP-FPM pool behind it is not. It has crashed, run out of workers, or is taking longer to answer than fastcgi_read_timeout allows. The cause is almost always a memory ceiling or a slow query, not the web server.

What I do

Size the pool against the memory the machine actually has, turn on the FPM slow log to find what is hanging, fix that query or background job, and raise the timeout only where it is genuinely justified.

Error. Page cannot be displayed. Please contact your service
provider for more details.

What it is

The hosting platform’s own error page, shown in place of the real one. It is hiding a 500, an account resource limit, a suspended account, or a process the host killed — the vagueness is the whole point of the page.

What I do

Get the real status code and the real log: request the URL headers directly, then open the account error log. From there it stops being a mystery page and becomes whatever it actually is.

There has been a critical error on this website.

Learn more about troubleshooting WordPress.

What it is

A PHP fatal error inside a plugin or a theme, with WordPress catching it and showing this generic page instead. It turns up right after an update, a PHP version change, or a migration that was not finished cleanly.

What I do

Switch on WP_DEBUG_LOG and read the actual fatal in debug.log, disable the offending plugin from the filesystem or WP-CLI to bring the site back, then repair or replace it on a staging copy before it goes anywhere near production again.

Your server is running PHP version 7.3.33 but WordPress
requires at least 7.4.

What it is

PHP that has reached end of life. 7.3 stopped receiving security fixes in December 2021, so the site is unpatched at the language level, and new WordPress and plugin releases will simply refuse to install.

What I do

Check every plugin and the theme for compatibility first, take a full backup, move a staging copy to PHP 8.1 or newer, clear the deprecations that surface there, then switch production over with the old version kept available for one rollback window.

Error establishing a database connection

What it is

Credentials in wp-config.php that no longer match, a MySQL or MariaDB service that has stopped or hit max_connections, or a table that crashed during an unclean shutdown.

What I do

Verify the database user, password and host, bring the service back and read why it went down in the first place, repair the crashed tables, then set connection and memory limits so the same traffic does not take it down again.

Fatal error: Allowed memory size of 268435456 bytes exhausted
(tried to allocate 20480 bytes) in /home/site/public_html/wp-
includes/wp-db.php on line 2056

What it is

A single request asked for more memory than PHP is allowed to hand it — an oversized query, an image being resized, or an import running in one pass. With display_errors off you never see that line; you get a blank white page.

What I do

Raise the limit to something sane to restore service, then profile the request that caused it and fix the cause. A permanently high limit does not solve this, it just moves the crash further down the road.

Briefly unavailable for scheduled maintenance.
Check back in a minute.

What it is

An update that failed part way through and left the .maintenance file sitting in the site root. Nothing is updating any more, and nothing is going to clear that file on its own.

What I do

Remove the file, establish whether the update actually completed, repair any core, plugin or theme files left half written, and then run the update again properly.

example.com redirected you too many times.
ERR_TOO_MANY_REDIRECTS

What it is

siteurl and home no longer matching the domain or the scheme after a move to HTTPS or a new host, a corrupted .htaccess, or cookies still pinned to the address the site used to have.

What I do

Correct siteurl and home directly in the database, regenerate the rewrite rules, clear the object and page cache, and confirm the redirect chain now ends on exactly one canonical URL.

POST /wp-login.php    401    x 4,812
POST /xmlrpc.php       200    x 1,190
GET  /?p=1             302 -> pills-cheap.example

What it is

Either an outdated plugin that has already been exploited — redirects to spam, pages nobody published, admin users nobody created — or the relentless brute force and XML-RPC traffic that usually comes first.

What I do

Clean the site back to a known good copy, rotate every credential, patch whatever let them in, then harden it: rate limiting, fail2ban, two-factor on admin accounts, file editing disabled, XML-RPC closed.

Your connection is not private
Attackers might be trying to steal your information from
example.com (for example, passwords, messages or credit cards).
NET::ERR_CERT_DATE_INVALID

What it is

The certificate expired. The renewal almost always stopped quietly weeks earlier — a changed vhost, a cron job removed during a cleanup, or a validation path that no longer resolves.

What I do

Reissue the certificate, repair the renewal so it runs unattended again, and add expiry monitoring so the next one is caught by a machine rather than by a customer.

Mixed Content: The page at 'https://example.com/' was loaded
over HTTPS, but requested an insecure resource
'http://example.com/wp-content/uploads/logo.png'.
This request has been blocked.

What it is

The page is served over HTTPS while assets are still hardcoded to http — normally left behind by a migration, or by a theme that stored absolute URLs in the database.

What I do

Rewrite the stored URLs in the database, fix the hardcoded ones in the theme, and have the browser upgrade anything remaining, so the padlock stays and nothing silently fails to load.

This site can't be reached
Check if there is a typo in example.com.
DNS_PROBE_FINISHED_NXDOMAIN

What it is

Nameservers that were never switched, a missing A or CNAME record, or an old TTL still handing the previous IP address to part of the world while the rest already sees the new one.

What I do

Set the records at the nameservers that are genuinely authoritative, verify from several resolvers rather than from one browser, and lower TTL ahead of the next migration so cutover takes minutes instead of days.

550 5.7.26 Unauthenticated email from example.com is not
accepted due to the domain's DMARC policy. Please contact the
administrator of example.com domain.

What it is

Mail sent straight from the server with nothing the receiving side can verify — no SPF alignment, no DKIM signature, and a DMARC policy that now rejects exactly that.

What I do

Send over authenticated SMTP, publish SPF, DKIM and DMARC correctly, and then confirm placement with real test sends rather than assuming the problem is solved.