It is difficult to help without the ability to read all configuration files.
You just placed one, not included or php-fpm.conf. This is not disapproval (the wall of configuration files is not entirely appropriate in the question), but simply to indicate that the configuration file that we "donβt see" may differ depending on the installation.
In any case, I see some differences from the configuration file that I have on the server for the wordpress site.
Here are some tips to consider that since you are not getting any php-fpm errors , and nginx can "communicate" with it through a socket (otherwise you will get a bad gateway error).
At the beginning...
server { listen 80; server_name wordpress.bob; root /Users/mark/Sites/wordpress; index index.php;
Make sure you have wordpress.conf enabled
location / { try_files $uri $uri/ /index.php?$args; }
The last part...
location ~ \.php$ { fastcgi_split_path_info ^(.+\.php)(/.+)$; fastcgi_buffer_size 128k; fastcgi_buffers 256 16k; fastcgi_busy_buffers_size 256k; fastcgi_temp_file_write_size 512k; fastcgi_intercept_errors on; fastcgi_max_temp_file_size 0; fastcgi_connect_timeout 3s; fastcgi_send_timeout 5s; fastcgi_read_timeout 5s; include fastcgi.conf;
The difference between fastcgi.conf and fastcgi_params (during my installation) is only one line:
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
If this line is missing, php code cannot read $_SERVER['SCRIPT_FILENAME'] and (I think), it can break Wordpress code, which will lead to empty conclusions.
Finally, make sure php-fpm workflows have /usr/local/var/run/php-www.sock permissions
Usually the socket has the same owner: a group of workers.
The user and group of workers are installed in php-fpm.conf:
; Unix user/group of processes ; Note: The user is mandatory. If the group is not set, the default user group ; will be used. user = ...... group = ......