WordPress Errors

A 500 Internal Server Error means the WordPress application errored before it could produce a valid response. This article explains the exact mechanism, the five real causes, how to tell them apart in the logs, and how to fix each one.

This error means WordPress can't connect to your database, which takes your entire site offline.

WordPress shows this screen when a PHP fatal error stopped the request. The recovery email is the fastest way back in, and the debug log tells you which plugin or theme to remove.

A completely blank WordPress page in 2026 almost always means a PHP fatal fired before WordPress could render the critical error screen. Here is how to turn that blank page into a readable error and fix the cause.

A redirect loop makes your WordPress site unreachable. This guide names the five real causes in order of likelihood and fixes each one with a verification you can check in the browser.

PHP killed the request because it tried to allocate more memory than your server allowed. Fix it by raising WP_MEMORY_LIMIT or the underlying PHP memory_limit, and track down the plugin, upload, or admin task that is asking for too much.

WordPress shows this message when an update is running. If it stays visible, a file called .maintenance was left behind. Here is how to remove it safely.

A PHP script ran longer than the server allowed. Fix it by raising max_execution_time, calling set_time_limit(), or breaking the task into smaller batches.

A 403 Forbidden on a WordPress site means the server understood the request and refused it on purpose. This article explains the five real causes, how to tell which is yours, and how to fix each one.

A single missing bracket or semicolon in a PHP file can take your entire WordPress site down. The error message tells you exactly where it is, and the fix is usually two edits away.

A WordPress login redirect loop throws you back to wp-login.php after you submit valid credentials. This article walks through the real causes in order of likelihood, shows you how to verify each one, and gives you a targeted fix per case.

WordPress shows this message when a capability check has rejected your request after you logged in. The fix is almost never the login form: it is your stored role, a plugin that rewrote your capabilities, or a database mismatch that broke the lookup.

A 502 Bad Gateway means the proxy in front of your WordPress site got an invalid response (or no response at all) from the application server, usually because a PHP-FPM worker crashed mid-request. This article explains the exact mechanism, the five real causes, how to tell them apart in the logs, and how to fix each one.

A 503 Service Unavailable means the server in front of your WordPress site is currently unable to handle the request, almost always because something is overloaded, rate-limited, or in maintenance. This article explains the exact mechanism, the five real causes, how to tell them apart, and how to fix each one.

A 504 Gateway Timeout means the proxy in front of your WordPress site waited too long for the application server to answer and gave up. This article explains the exact mechanism, the four real causes, how to tell them apart, and how to fix each one.

WordPress shows a bare "HTTP error" in the media library when the server rejects an upload mid-request. The cause is almost always a size limit, an Imagick memory spike, a request timeout, or a WAF rule. Diagnose with the server error log and fix the right limit.

A WordPress page returns 404 even though it exists in the admin. This article explains why that happens, how to tell which of the five causes is yours, and how to fix each one.

You enabled SSL and now your WordPress site shows a broken padlock or 'Not Secure' warning. That is mixed content, not a broken certificate. This article explains the difference, how to find the offending resources, and the four real fixes in order of thoroughness.

Your SSL certificate is active (for the primer on certificate types, see [SSL certificates in WordPress]({{ '/en/knowledge-base/wordpress/security/ssl-certificate-in-wordpress/' | locale_url }})) but WordPress still loads over HTTP, or you enabled a redirect and now see ERR_TOO_MANY_REDIRECTS. This guide covers the four methods that actually work, in order of safety, and the two configurations (Cloudflare Flexible and untrusted reverse proxies) that will burn you if you skip them.

WordPress shows 'Missed schedule' when the publish time passed but the cron runner never fired. The post is not lost. This article gets the stuck post live in the next minute and fixes the underlying cause so the next one publishes on time.

A 429 Too Many Requests in WordPress almost never means your site is under attack. In most cases the site is rate-limiting itself, usually because WP-Cron or a runaway Action Scheduler job is hammering an external API. This article explains which layer is actually saying 'slow down', how to tell them apart from the logs, and how to fix each root cause without breaking anything else.

WordPress refuses an upload with "exceeds the upload_max_filesize directive" because the limit comes from PHP, not WordPress. Raise upload_max_filesize and post_max_size together in the right place for your stack, then verify in Site Health.

Another KB article told you to turn on WP_DEBUG and read debug.log. Here is exactly how to do it safely, where the file lives, what each constant really does, and how to keep the log private so visitors never see your errors.

You clicked update, the page reloaded, and now your site is down. This article walks the safe recovery path: get back into wp-admin, identify the culprit, roll back the right way, and avoid the database migration trap that breaks naive rollbacks.

You click Install Plugin, and instead of installing, WordPress shows a Connection Information form asking for an FTP host, username, and password. This article explains the file ownership check that triggers it, walks through the correct fix (chown to the PHP user), shows the wp-config.php shortcuts and when each one is safe, and lists the common Nginx and Docker scenarios where it shows up.

When the WordPress media library shows blank tiles, broken-image icons, or generic placeholders instead of thumbnails, the cause is usually one of seven things: file permissions, a missing image library, stale post meta after a migration, an SSL switch that left HTTP URLs behind, the WP 5.3 -scaled trap, or a format the server cannot decode. Diagnose with Site Health and the browser network tab before regenerating anything.

WordPress emailed you to say the automatic update did not finish. The site might be fine, the site might be locked in maintenance mode, or the site might be running the new version with cached old code. This article walks the safe recovery path: verify the site state, clear the .maintenance file, work through the five real causes, and confirm the update actually completed.

WordPress refuses an upload with "Sorry, you are not allowed to upload this file type" because the file's MIME type is not in the default allow-list. The fix depends on the format. SVG needs Safe SVG (sanitisation is mandatory). WebP needs WordPress 5.8+ and a server with WebP support. HEIC needs WordPress 6.7+ with Imagick. ALLOW_UNFILTERED_UPLOADS is an emergency escape hatch, not a permanent fix.

After a WordPress update, every visit to wp-admin sends you to the Database Update Required screen. You click the button, it says it succeeded, and the next page load shows the same screen again. Here is why the loop happens and how to break it without touching the schema by hand.

Your WooCommerce add-to-cart button does nothing, the cart empties itself between pages, or the checkout page spins without completing the order. The cause is almost never 'a plugin conflict' in the way most forums suggest. In the majority of cases it is a caching configuration that serves a stale page where it should serve a dynamic one. This article walks through every common cause in order of likelihood, with a browser-side diagnostic for each.

WordPress shows "The link you followed has expired. Please try again." with a blank white page and a 403 status code. The message sounds like a session timeout, but the most common cause is a file upload that exceeded PHP's post_max_size, which silently empties the form data and makes WordPress think the security nonce was never sent.

Your WooCommerce checkout is broken and you need to know whether the cause is caching, a JavaScript conflict, a payment gateway, a Web Application Firewall, or something else entirely. The hardest part is not the fix. It is figuring out which of seven possible failure modes you are actually looking at. This article disambiguates the symptom first, then routes you to the right diagnostic and the right sibling article when one exists.

This error means WordPress had a working database connection, but MySQL or MariaDB closed it before the current query could finish. The most common triggers are an idle connection that exceeded wait_timeout, or a query packet larger than max_allowed_packet.

You enabled a payment gateway in WooCommerce settings but it does not appear on the checkout page. The cause is usually one of five things: the gateway does not support the Checkout Block, the site lacks a valid SSL certificate, the gateway is in test mode with live credentials (or vice versa), a plugin conflict prevents the gateway JavaScript from loading, or geographic restrictions silently exclude the customer's country. This article walks through each cause with a diagnostic for every step.

Your WooCommerce checkout shows "No shipping options were found for your location" or displays the wrong shipping rate. The cause is usually zone ordering (WooCommerce matches the first zone in the list, not the most specific), a free shipping method whose conditions are not all met, or cached shipping rates that persist after a settings change. This article walks through the zone priority model, the most common misconfigurations, and the debug tools WooCommerce provides.

Your site is broken, you suspect a plugin, and you do not know which one. This article walks the structured diagnosis: safe-mode for one user, folder-rename bisection when wp-admin is gone, and the debug log read that names the offender.