Configuration for lithium on nginx - nginx

I would like to deploy lithium on nginx server, however there are configurations provided only for Apache and IIS.
I've successfully written several nginx server configurations for various applications in past, but I'm struggling with this one.
Already asked this question on nginx and lithium forums, no luck.
This is best of what I've made so far.
root /var/servers/my_app/app/webroot;
location / {
index index.php;
try_files $uri $uri/ index.php;
}
location ~ \.php {
fastcgi_pass unix:/tmp/php.socket;
fastcgi_index index.php;
include fastcgi_params;
fastcgi_param SCRIPT_FILENAME /var/servers/my_app/$fastcgi_script_name;
fastcgi_param PATH_INFO $fastcgi_script_name;
}
Problem is on / (root page) every link gets index.php prepended, e.g. instead of
www.example.com/something
I get
www.example.com/index.php/something
Not sure if this even is nginx configuration related or rather something, that lithium does when it cannot detect Apache/IIS environment. Either way I cannot solve it.
Another thing, that when I access "www.example.com/test" (via direct URL input), the page renders correctly, however "www.example.com/test/" (with trailing slash) and "www.example.com/test/anything_here" is broken - all links gets appended to current URL e.g. pressing the same link creates this:
www.example.com/test/
www.example.com/test/test
www.example.com/test/test/test
EDIT: Updated configuration
(Sorry for much delayed edit, but I'm still stuck and recently restarted solving this)
root /var/server/my_app/app/webroot/;
index index.php index.html;
try_files $uri $uri/ /index.php?$args;
location ~ \.php {
fastcgi_pass unix:/tmp/php.socket;
fastcgi_index index.php;
include fastcgi_params;
fastcgi_param SCRIPT_FILENAME /var/servers/my_app/$fastcgi_script_name;
fastcgi_param PATH_INFO $fastcgi_script_name;
}
location ~/\.ht {
deny all;
}
}
As I mentioned in comments this now causes all links to have index.php included, it looks like:
www.example.com/index.php/something
www.example.com/index.php/stylesheet.css

I think your problem is that the try_files shouldn't be inside a location block.
Try the configuration shown here: http://li3.me/docs/manual/configuration/servers/nginx.wiki
I helped define it and have been using it locally and in production. It shouldn't cause any of the issues you're reporting.
Copying it below:
server {
listen IP_ADDRESS_HERE:80;
server_name DOMAIN.COM;
root /var/www/DOMAIN.COM/webroot/;
access_log /var/log/DOMAIN.com/access.log;
error_log /var/log/DOMAIN.com/error.log warn;
index index.php index.html;
try_files $uri $uri/ /index.php?$args;
location ~ \.php$
{
fastcgi_pass 127.0.0.1:9000;
fastcgi_index index.php;
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
include /etc/nginx/fastcgi_params;
}
location ~ /\.ht {
deny all;
}
}

As I suspected the problem was that Lithium relies on some environment variables, in this case - link generation, it uses PHP_SELF, which happened to be incorrect.
Solution:
fastcgi_param PATH_INFO $fastcgi_path_info;
Instead of previously incorrect:
fastcgi_param PATH_INFO $fastcgi_script_name;
So final configuration:
root /var/server/my_app/app/webroot/;
index index.php index.html;
try_files $uri $uri/ /index.php?$args;
location ~ \.php {
fastcgi_pass unix:/tmp/php.socket;
fastcgi_index index.php;
include fastcgi_params;
fastcgi_param SCRIPT_FILENAME /var/servers/my_app$fastcgi_script_name;
fastcgi_param PATH_INFO $fastcgi_path_info;
}
location ~/\.ht {
deny all;
}
Thanks to rmarscher and mehlah # lithum forums.

Related

NGINX renders PHP files when given path, serves index as download without a path

My current NGINX & PHP-FPM setup works when given a working filepath like example.com/index.php, but still serves the index.php as a download when visiting example.com without a file.
I've tried adding a location / redirect and try_files, but it still doesn't redirect properly.
I know how it's done using .htaccess files, but I'd like to do this in NGINX to avoid changing every repository.
Here's my current conf file.
server {
server_name example.com;
error_log /var/log/nginx/error.log;
access_log /var/log/nginx/access.log;
root /var/www/html;
location / {
try_files $uri $uri/ =404;
}
location ~ \.php$ {
try_files $uri $uri/ =404;
include fastcgi_params;
fastcgi_pass php:9000;
fastcgi_index index.php;
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
fastcgi_split_path_info ^(.+\.php)(/.+)$;
fastcgi_param PATH_INFO $fastcgi_path_info;
}
}
Remove try_files $uri $uri/ =404; from location ~ \.php$ {.

install postfixadmin in a subfolder of a domain in nginx

I need to install postfixadmin as a subfolder of a domain hosted in nginx.
In other words, I'm trying to access postfixadmin using http://example.com/postfixadmin
Physically, the contents of the sites are stored this way:
example.com site in /var/www/example
Postfixadmin files in /var/www/postfixadmin
I've tried adding this within the server section of example.com:
location ~ /posfixadmin/ {
root /var/www;
index index.php;
include snippets/fastcgi-php.conf;
fastcgi_pass unix:/var/run/php5-fpm.sock;
}
The above works partially, php scripts are executed correctly, but css and image files found under /var/www/postfixadmin/css and /var/www/postfixadmin/images are not loaded.
I've checked at the generated html code and the links to css and image files in postfixadmin are called using relative paths, like this:
href="css/default.css"
I think nginx tries to get the css files from http://example.com/css instead of http://example.com/postfixadmin/css, that's why it's failing, I've tried something like this:
location /postfixadmin/css/ {
root /var/www/postfixadmin;
}
but the above doesn't work.
Any idea about how to fix this? Thanks in advance!
I know it's an old topic, but still: do not use "root" in a "location" block. Source: https://www.nginx.com/resources/wiki/start/topics/tutorials/config_pitfalls/
This works for me at the moment with the current PostfixAdmin 3.2 (which moved all the public facing stuff into the "public" subdirectory). Mind that I have defined the fastcgi_pass elsewhere, so this bit is not directly applicable.
location /postfixadmin {
alias /usr/local/www/postfixadmin/public;
location ~ ^(.+\.php)(.*)$ {
fastcgi_pass php;
fastcgi_index index.php;
fastcgi_param SCRIPT_FILENAME $request_filename;
include fastcgi_params;
fastcgi_split_path_info ^(.+\.php)(.*)$;
fastcgi_param PATH_INFO $fastcgi_path_info;
fastcgi_param PATH_TRANSLATED $document_root$fastcgi_path_info;
fastcgi_read_timeout 180;
fastcgi_buffers 4 256k;
fastcgi_buffer_size 128k;
}
location ~ \.php {
include /usr/local/etc/nginx/fastcgi_params;
fastcgi_param SCRIPT_FILENAME $request_filename;
fastcgi_pass php;
fastcgi_index index.php;
}
# deny access to .htaccess files, if Apache's document root
# concurs with nginx's one
#
location ~ /\.ht {
deny all;
}
}

Nginx rewrite to make a "personal URL"

I'd like to be able to make "personal URL" for our users (Facebook like), which is of course a dynamic strings. it needs to be in the root of the site, and that is why I'm having a big headache with it.
The requirements that I have are:
1. I need
www.example.com/John.Doe (it can be a-zA-Z0-9_-.)
and rewrite it to:
www.example.com/profile?id=John.Doe
2. I also need the site scripts to be extension less like (which I was able to do, with the great people here, using "$uri.php$is_args$query_string;"):
so
www.example.com/login
will go to:
www.example.com/login.php
I tried a lot of things, but I just can't get the right formula to make it work.
This is my configuration, right now:
location / {
try_files $uri $uri/ $uri.php$is_args$query_string;
}
location ~ \.php$ {
if ($request_uri ~ ^/([^?]*)\.php(\?.*)?$) {
return 301 /$1$2;
}
try_files $uri =404;
include /etc/nginx/fastcgi_params;
fastcgi_pass 127.0.0.1:9001;
fastcgi_index index.php;
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
}
You have an overlapping namespace for your extension-less scripts and your personal URLs, so you need to test for the presence of the $uri.php file before rewriting it to profile.php.
So rather than rewrite the URI in the first try_files directive (as you have it now), it may be better to use a named location block to process the PHP file without having to rewrite it first.
Like this:
location / {
try_files $uri $uri/ #php;
}
location #php {
try_files $uri.php #rewrite;
include /etc/nginx/fastcgi_params;
fastcgi_pass 127.0.0.1:9001;
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
}
location #rewrite {
rewrite ^/([a-zA-Z0-9_.-]+)$ /profile.php?id=$1 last;
}
location ~ \.php$ {
if ($request_uri ...) { ... }
try_files $uri =404;
include /etc/nginx/fastcgi_params;
fastcgi_pass 127.0.0.1:9001;
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
}
The first block serves static files. The second block processes extension-less PHP files (if and only if they exists). The third block performs a rewrite to profile.php (which does not need to be extension-less, as it is not exposed to the client). The fourth block processes normal .php URIs and includes your $request_uri fix.
Note that fastcgi_index is redundant in both this and the original configuration.
For more details, here are some links to the nginx documentation for the location, try_files and rewrite directives.

fastcgi - passing information to scripts

I`m new with nginx and I dont know even what should I look for. My task is to force (I think it is) fastcgi to pass information (PATH_INFO) to the scripts. I have been told that I should add something in block after location ~ .php$ . Adding before fastcgi_pass xxx;
fastcgi_param PATH_INFO $fastcgi_path_info;
would made that happened ?
My location part config file looks like that:
location ~ ^/backend\.php/(.*)$ {
try_files $uri /backend.php?$1;
}
location ~ \.php$ {
# Zero-day exploit defense.
# http://forum.nginx.org/read.php?2,88845,page=3
# Won't work properly (404 error) if the file is not stored on this server, which is entirely possible with php-fpm/php-fcgi.
# Comment the 'try_files' line out if you set up php-fpm/php-fcgi on another machine. And then cross your fingers that you won't get hacked.
try_files $uri =404;
fastcgi_split_path_info ^(.+\.php)(/.+)$;
include fastcgi_params;
fastcgi_index index.php;
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
fastcgi_pass xxx;
}

lemonstand in subdirectory nginx

I'm running Lemonstand under the subdirectory www.mysite.com/shop/
Here is my location rule for lemonstand:
# Lemonstand
location /shop {
root /home/sites/mysite.com/public_html/shop/;
index index.php;
include fastcgi_params;
fastcgi_split_path_info ^(.+\.php)(/.+)$;
fastcgi_param SCRIPT_FILENAME $document_root/index.php;
fastcgi_param SCRIPT_NAME index.php;
fastcgi_param QUERY_STRING url=$uri&$args;
fastcgi_pass 127.0.0.1:9000;
fastcgi_index index.php;
fastcgi_buffer_size 32k;
fastcgi_buffers 4 32k;
fastcgi_busy_buffers_size 64k;
}
I can access the page at mysite.com/shop.
All the URLS for the shop should be like this:
mysite.com/shop/category/freight
mysite.com/shop/products/dog-toy
When in fact on the page they are structured like this:
mysite.com/category/freight
mysite.com/products/dog-toy
And the strange thing is, even if I paste in the correct URL to the browser, it only ever shows my base /shop/ page, as if the other pages don't exist. Can anyone help?
You need to serve php from a different location block.
[EDIT] to avoid conflicting with other php scripts, you may use named location.
location ^~/shop/ {
root /home/sites/mysite/public_html;
try_files $uri $uri/ #anotherPhpSite
}
location #anotherPhpSite {
include fastcgi_params;
# Defend against arbitrary PHP code execution
# NOTE: You should have "cgi.fix_pathinfo = 0;" in php.ini
# More info:
# https://nealpoole.com/blog/2011/04/setting-up-php-fastcgi-and-nginx-dont-trust-the-tutorials-check-your-configuration/
fastcgi_split_path_info ^(.+\.php)(/.+)$;
fastcgi_param SCRIPT_FILENAME $document_root/shop/index.php;
fastcgi_param SCRIPT_NAME /shop/index.php;
fastcgi_param QUERY_STRING url=$uri&$args;
fastcgi_pass 127.0.0.1:9000;
fastcgi_index index.php;
fastcgi_buffer_size 32k;
fastcgi_buffers 4 32k;
fastcgi_busy_buffers_size 64k;
}
My main problems were expression syntax and config structure.
OK. So here's the working config file. best practice was to split up the two locations, and utilize one php block for them both.
First the main controller block, which uses silverstripe CMS, here i'm passing the data straight to the main controller in sapphire.
# Main silverstripe declaration
location / {
index /sapphire/main.php;
try_files $uri $uri/ /sapphire/main.php?url=$uri&$args;
}
Second was my shop under a sub-directory which ran lemonstand. It was mainly a case of getting the expressions correct, then passing the requests to the relevant path. I didn't set another root or alias here, as the rule was inherited from the re-write.
# Shop lemonstand files
location /shop/ {
index index.php;
rewrite ^/shop(.*)$ /shop/index.php?q=$1 last;
try_files $uri $uri/ /shop/index.php?q=$1;
}
And finally I used a generic PHP block so I didn't have to duplicate my code in the first two. So any thing that goes through the first two block will pass into here.
# PHP controller
location ~ ^(.*?\.php)(/.*)?$ {
fastcgi_split_path_info ^(.+?\.php)(/.+)$;
include fastcgi_params;
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
fastcgi_pass 127.0.0.1:9000;
fastcgi_index index.php;
fastcgi_buffer_size 32k;
fastcgi_buffers 4 32k;
fastcgi_busy_buffers_size 64k;
}

Resources