I have 2 folders dist and api_middleware in '/var/www/public_html/website.com/'

dist folder contains my frontend production code. This folder is symblink-ed from /home/user/frontend/dist

api_middleware folder contains codeignitor code, that we use as middleware for frontend to communicate with our erp. This folder is symblink-ed from /home/user/api_middleware.

I want to host both code with nginx. And this is the code that I came up with.

server {
  listen 443;
    ssl on;
    ssl_certificate /etc/ssl/website.com.crt; 
    ssl_certificate_key /etc/ssl/website.com.key;
    server_name website.com;
    access_log /var/log/nginx/nginx.vhost.access.log;
    error_log /var/log/nginx/nginx.vhost.error.log;

    index index.php index.html index.htm index.nginx-debian.html;

    location / {
       root /var/www/public_html/website.com/dist;
       try_files $uri $uri/ /index.html;
    }

    location /api_middleware {
       root /var/www/public_html/website.com;
       try_files $uri $uri/ /api_middleware/index.php?/$request_uri;
       client_max_body_size 100M;
    }

    location ~ \.php$ {
    include snippets/fastcgi-php.conf;
    fastcgi_pass unix:/run/php/php7.2-fpm.sock;
    fastcgi_read_timeout 300;
    }
}

navigating to website.com works. But the api calls with website.com\api_middleware\<PARAMS> is returning a 404

What am I doing wrong?

PS. Using ubuntu 18.4.

CodePudding user response：

My Updated working code:

server {
  listen 443;
    ssl on;
    ssl_certificate /etc/ssl/website.com.crt;
    ssl_certificate_key /etc/ssl/website.com.key;
    server_name website.com;
    access_log /var/log/nginx/nginx.vhost.access.log;
    error_log /var/log/nginx/nginx.vhost.error.log;
    root /var/www/public_html/website.com/dist;
    index index.php index.html index.htm index.nginx-debian.html;

    location /api_middleware {
        alias /var/www/public_html/website.com/api_middleware/;
        try_files $uri $uri/ /api_middleware/api_middleware/index.php?/$request_uri;
        client_max_body_size 100M;
    }

    location ~ /api_middleware/. \.php$  {
        include snippets/fastcgi-php.conf;
        fastcgi_pass unix:/run/php/php7.2-fpm.sock;
        fastcgi_read_timeout 300;
    }
}

The issue was that using multiple root was messing with nginx to access the directory location my second codebase.

I used alias to change the location to search for my 2nd codebase, while using root to set location of my 1st codebase.

Take a look at the line try_files $uri $uri/ /api_middleware/api_middleware/index.php?/$request_uri;. I have added 'api_middleware' twice here, even though the location to index.php file is /var/www/public_html/website.com/api_middleware/index.php . This is needed because of a very old bug in nginx.

finally, to process the php files, I changed the URI where to look for the php files to ~ /api_middleware/. \.php$.

Reference:

1. NGINX try_files alias directives
2. https://serverfault.com/questions/1035733/nginx-difference-between-root-and-alias
3. https://serverfault.com/questions/684523/nginx-multiple-roots/684605