Docs

A website shows a 500, 503 or white page

Find the cause of a 500 Internal Server Error, 503 Service Unavailable or blank page on an intoPanel server: the error_log file, the web server's error log, PHP-FPM and resource limits.

AlmaLinux, Rocky Linux, Ubuntu, Debian 2 min read Updated

A 500 error, a 503 error and a blank white page usually have different causes. Work through them in this order.

1. Read the website's error_log#

PHP writes its own errors to a file named error_log in the folder of the script that failed: an error in public_html/index.php goes to public_html/error_log, one in public_html/wp-admin/admin.php to public_html/wp-admin/error_log. Clients find it in the File Manager; over SSH:

find /home/USERNAME/public_html -name error_log -mmin -60 -exec tail -n 20 {} \;

Look for a line starting with PHP Fatal error at the time of the failure. The most common ones:

  • Allowed memory size exhausted: raise memory_limit for the website in its PHP settings.
  • Call to undefined function or Class not found: a missing PHP extension, or a plugin that does not support the PHP version. See How to add a PHP extension.
  • Maximum execution time exceeded: a slow script; raise max_execution_time or find what is slow.

2. Read the web server's error log#

The client area's Error Log page shows the last 200 lines for each website. Over SSH, the web server's per-website logs and Apache's own error log tell you about refused requests, .htaccess mistakes and ModSecurity:

tail -n 50 /var/log/httpd/error_log     # AlmaLinux, Rocky Linux
tail -n 50 /var/log/apache2/error.log   # Ubuntu, Debian
  • Invalid command or RewriteRule errors: a broken .htaccess file. Rename it to test.
  • ModSecurity: Access denied: a rule matched the request. Check Security → ModSecurity Hit Log, and switch ModSecurity off for that website only if you are sure the request is safe.

3. A 503: is PHP-FPM running?#

A 503 usually means Apache could not reach the website's PHP. Check the PHP-FPM service of the version the website uses (here 8.3):

systemctl status php83-php-fpm --no-pager
journalctl -u php83-php-fpm -n 50 --no-pager

On Ubuntu and Debian the service is named php8.3-fpm.

With Account Isolation on, the account's PHP runs in a service of its own; the admin area's Service Status and the account's page show its state. Restarting PHP from the panel is the safest way to bring it back.

4. Is the account at its resource limit?#

When an account reaches its memory or process limit, new requests can fail until usage drops. Open Resource Usage in the admin area and look at the account's history. If it is at the limit every day, raise the limit in its package, or find the plugin or bot that causes the load.

5. Is the account suspended or over quota?#

A suspended account shows a suspension page; a full disk can make PHP fail to write sessions and caches. Check the account's page in All Accounts.

Still stuck?#

Ask in the community with the exact error line, the PHP version and what you already tried, or open a ticket if you hold a licence.

Frequently asked questions

Should I turn on "display_errors" to see the error?

Only briefly while testing, with Show errors to visitors in the website's PHP settings. The error_log file is safer: visitors cannot open it.

The error_log file is huge. Can I delete it?

Yes. It is created again with the next error. Fix the cause first, or it fills up again.

Was this article helpful?

Something missing? Ask in the community or open a ticket.