Bad Gateway and php5-fpm.sock fail

Hi, I’m experiencing the exact same problem, not sure how to fix it. I already restart a couple of times service php5-fpm restart.

the ls -lah /var/run/php5-fpm.sock is returning: www-data www-data 0 Dec 22 22:18 /var/run/php5-fpm.sock

and the status is running.

I already have cloudflare and 5 droplets connected. And I’m getting same error.

This is affecting our current website a lot.

Any help on this?

Thanks

Heya,

A 502 Bad Gateway error typically indicates that the gateway/proxy (in your case, Nginx) received an invalid response from the upstream server it tried to access (which seems to be PHP-FPM in your situation).

The error message you’re seeing suggests that Nginx is attempting to connect to PHP-FPM using a Unix socket at /var/run/php5-fpm.sock, but it can’t establish a connection. This could be due to PHP-FPM not running, the socket file not existing, or permission issues.

Here are some steps to troubleshoot the issue:

1. Check PHP-FPM Status: Check if PHP-FPM is running:

sudo service php5-fpm status

If it’s not running, try to start it:

sudo service php5-fpm start

2. Check PHP-FPM Socket File: If PHP-FPM is running, check if the socket file exists at /var/run/php5-fpm.sock. If it doesn’t, there might be a configuration issue.

3. PHP-FPM Configuration:

   • Verify that the PHP-FPM pool configuration (usually located at /etc/php5/fpm/pool.d/www.conf or similar) has the listen directive set to /var/run/php5-fpm.sock.
   • Ensure the user and group in the PHP-FPM configuration match the user and group that Nginx is running under.
   • After making changes to the configuration, restart PHP-FPM.

4. Nginx Configuration:

   • Check the Nginx configuration for your site (typically found in /etc/nginx/sites-available/ directory) and ensure that the fastcgi_pass directive points to the correct socket file:

fastcgi_pass unix:/var/run/php5-fpm.sock;

• After making changes, test the configuration with sudo nginx -t and if successful, reload Nginx: sudo service nginx reload.

5. File Permissions:

   • Check the permissions of the socket file and ensure that Nginx has permission to read and write to it.
   • The permissions should allow the user that Nginx runs as (usually www-data) to access the socket.

6. Logs:

   • Check both the Nginx error logs (/var/log/nginx/error.log) and the PHP-FPM logs for any additional information.
   • The PHP-FPM log might give you more insights into why the connection is failing.

7. Resources:

   • A “Resource temporarily unavailable” error might indicate a resource limit issue. Check the server’s resources (CPU, memory) and consider whether the PHP-FPM pm.max_children setting is too low.

8. Cloudflare:

   • Make sure that Cloudflare’s IP ranges are whitelisted in your firewall.
   • Verify that there’s no IP conflict between Cloudflare’s settings and your server.

If you have recently migrated your droplet or changed configurations, ensure that all paths and settings are consistent with the new environment. If you keep hitting a wall, consider reaching out to DigitalOcean support again with the latest details and steps you have taken.