Recently I installed the latest version of Nginx and looks like I'm having hard time running PHP with it.
Here is the configuration file I'm using for the domain:
server {
listen 80;
server_name localhost;
location / {
root /usr/share/nginx/html;
index index.php;
}
error_page 500 502 503 504 /50x.html;
location = /50x.html {
root /usr/share/nginx/html;
}
location ~ \.php$ {
fastcgi_pass 127.0.0.1:9000;
fastcgi_index index.php;
fastcgi_param SCRIPT_FILENAME /scripts$fastcgi_script_name;
include fastcgi_params;
}
}
Here is the error I'm getting on the error log file:
FastCGI sent in stderr: "Primary script unknown" while reading response header from upstream
Try another *fastcgi_param* something like
fastcgi_param SCRIPT_FILENAME /usr/share/nginx/html$fastcgi_script_name;
I had the "file not found" problem, so I moved the "root" definition up into the "server" bracket to provide a default value for all the locations. You can always override this by giving any location it's own root.
server {
root /usr/share/nginx/www;
location / {
#root /usr/share/nginx/www;
}
location ~ \.php$ {
try_files $uri =404;
fastcgi_split_path_info ^(.+\.php)(/.+)$;
fastcgi_pass 127.0.0.1:9000;
fastcgi_index index.php;
include fastcgi_params;
}
}
Alternatively, I could have defined root in both my locations.
Probably it's too late to answer but a couple things since this is a really annoying error. Following solution worked on Mac OS X Yosemite.
It's the best if you have
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
The include with fast cgi params should go above that line.
All your directories down to the PHP file you're executing (including that file too) should have a+x permissions, e.g.
sudo chmod a+x /Users/
sudo chmod a+x /Users/oleg/
sudo chmod a+x /Users/oleg/www/
sudo chmod a+x /Users/oleg/www/a.php
I had been having the same issues,
And during my tests, I have faced both problems:
1º: "File not found"
and
2º: 404 Error page
And I found out that, in my case:
I had to mount volumes for my public folders both on the Nginx volumes and the PHP volumes.
If it's mounted in Nginx and is not mounted in PHP, it will give: "File not found"
Examples (Will show "File not found error"):
services:
php-fpm:
build:
context: ./docker/php-fpm
nginx:
build:
context: ./docker/nginx
volumes:
#Nginx Global Configurations
- ./docker/nginx/nginx.conf:/etc/nginx/nginx.conf
- ./docker/nginx/conf.d/:/etc/nginx/conf.d
#Nginx Configurations for you Sites:
# - Nginx Server block
- ./sites/example.com/site.conf:/etc/nginx/sites-available/example.com.conf
# - Copy Public Folder:
- ./sites/example.com/root/public/:/var/www/example.com/public
ports:
- "80:80"
- "443:443"
depends_on:
- php-fpm
restart: always
If it's mounted in PHP and is not mounted in Nginx, it will give a 404 Page Not Found error.
Example (Will throw 404 Page Not Found Error):
version: '3'
services:
php-fpm:
build:
context: ./docker/php-fpm
volumes:
- ./sites/example.com/root/public/:/var/www/example.com/public
nginx:
build:
context: ./docker/nginx
volumes:
#Nginx Global Configurations
- ./docker/nginx/nginx.conf:/etc/nginx/nginx.conf
- ./docker/nginx/conf.d/:/etc/nginx/conf.d
#Nginx Configurations for you Sites:
# - Nginx Server block
- ./sites/example.com/site.conf:/etc/nginx/sites-available/example.com.conf
ports:
- "80:80"
- "443:443"
depends_on:
- php-fpm
restart: always
And this would work just fine (mounting on both sides) (Assuming everything else is well configured and you're facing the same problem as me):
version: '3'
services:
php-fpm:
build:
context: ./docker/php-fpm
volumes:
# Mount PHP for Public Folder
- ./sites/example.com/root/public/:/var/www/example.com/public
nginx:
build:
context: ./docker/nginx
volumes:
#Nginx Global Configurations
- ./docker/nginx/nginx.conf:/etc/nginx/nginx.conf
- ./docker/nginx/conf.d/:/etc/nginx/conf.d
#Nginx Configurations for you Sites:
# - Nginx Server block
- ./sites/example.com/site.conf:/etc/nginx/sites-available/example.com.conf
# - Copy Public Folder:
- ./sites/example.com/root/public/:/var/www/example.com/public
ports:
- "80:80"
- "443:443"
depends_on:
- php-fpm
restart: always
Also here's a Full working example project using Nginx/Php, for serving multiple sites:
https://github.com/Pablo-Camara/simple-multi-site-docker-compose-nginx-alpine-php-fpm-alpine-https-ssl-certificates
I hope this helps someone,
And if anyone knows more about this please let me know,
Thanks!
I just spent like 40 minutes trying to debug a non-working /status with:
$ SCRIPT_NAME=/status SCRIPT_FILENAME=/status QUERY_STRING= REQUEST_METHOD=GET cgi-fcgi -bind -connect /var/run/php5-fpm.sock
It just produced "File not found" error, while the actual scripts (that are found on the filesystem) worked just fine.
Turned out, I had a couple of orphaned processes of php5-fpm. After I killed everything and restarted php5-fpm cleanly, it just went back to normal.
Hope this helps.
The error message “Primary script unknown or in your case is file not found.” is almost always related to a wrongly set in line SCRIPT_FILENAME in the Nginx fastcgi_param directive (Quote from https://serverfault.com/a/517327/560171).
In my case, I use Nginx 1.17.10 and my configuration is:
location ~ \.php$ {
fastcgi_pass unix:/var/run/php/php7.2-fpm.sock;
fastcgi_index index.php;
fastcgi_param SCRIPT_FILENAME $realpath_root$fastcgi_script_name;
include fastcgi_params;
fastcgi_read_timeout 600;
}
You just change $document_root to $realpath_root, so whatever your root location, like /var/www/html/project/, you don't need write fastcgi_param SCRIPT_FILENAME /var/www/html/project$fastcgi_script_name; each time your root is changes. This configuration is more flexible. May this helps.
=================================================================================
For more information, if you got unix:/run/php/php7.2-fpm.sock failed (13: Permission denied) while connecting to upstream, just change in /etc/nginx/nginx.conf, from user nginx; to user www-data;.
Because the default user and group of PHP-FPM process is www-data as can be seen in /etc/php/7.2/fpm/pool.d/www.conf file (Qoute from https://www.linuxbabe.com/ubuntu/install-nginx-latest-version-ubuntu-18-04):
user = www-data
group = www-data
May this information gives a big help
In my case the PHP-script itself returned 404 code. Had nothing to do with nginx.
In my case, it was because the permissions on the root web directory were not set correctly. To do this, you need to be in the parent folder when you run this in terminal:
sudo chmod -R 755 htmlfoldername
This will chmod all files in your html folder, which is not recommended for production for security reasons, but should let you see the files in that folder, to be sure that isn't the issue while troubleshooting.
I had this error as well. In my case it was because there was another virtual host that was pointing to the same root directory.
After upgrading to PHP72, we had an issue where the php-fpm.d/www.conf lost the settings for user/group which was causing this error. Be sure to double check those if your setup involves php-fpm.
The website will show "File Not Found" error.
You should check this configure file: /etc/nginx/nginx.conf (error_log) to get the nginx webservice log location.
Then check nginx log: /var/log/nginx/error.log to get the root cause:
Keyword: FastCGI sent in stderr: "Primary script unknown"
Root cause: php-fpm service's account: apache doesn't have access permission to open webapps directory
Step 0. Find your root directory:
open /etc/nginx/nginx.conf file and find root keyword
Root directory: /var/lib/nginx/webapps/
Step 1. Make sure apache account can access the ROOT directory.
sudo -u apache ls -l /var/lib/nginx/webapps/
Step 2. chmod a+x permission for all ROOT folder
sudo chmod a+x /var/
sudo chmod a+x /var/lib/
sudo chmod a+x /var/lib/nginx/
sudo chmod a+x /var/lib/nginx/webapps/
For me, problem was Typo in location path.
Maybe first thing to check out for this kind of problem
Is path to project.
When getting "File not found", my problem was that there was no symlink in the folder where was pointing this line in ngix config:
root /var/www/claims/web;
in case it helps someone, my issue seems to be just because I was using a subfolder under my home directory, even though permissions seem correct and I don't have SELinux or anything like that.
changing it to be under /var/www/something/something made it work.
(if I ever found the real cause, and remember this answer, I'll update it)
try this command
sudo chmod 755 -R htdocs/
my case: I used relative dir
location ~* \.(css|js)\.php$ {
root ../dolibarr/htdocs;
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
fastcgi_param SCRIPT_NAME $fastcgi_script_name;
fastcgi_index index.php;
include /etc/nginx/fastcgi_params;
}
this line did not work too : fastcgi_param SCRIPT_FILENAME /vagrant/www/p1/../p2/htdocs/core/js/lib_head.js.php;
So I discovered fastcgi_param does not support relative path.
This works
fastcgi_param SCRIPT_FILENAME /vagrant/www/p2/htdocs/core/js/lib_head.js.php;
I have solved this issue in nginx version: nginx/1.21.3 PHP 7.4.23 macOS Catalina version 10.15.7
nano /usr/local/etc/nginx/nginx.conf
location ~ \.php$ {
root html;
include fastcgi.conf;
fastcgi_pass 127.0.0.1:9000;
fastcgi_index index.php;
#Comment bellow Line
#fastcgi_param SCRIPT_FILENAME /scripts$fastcgi_script_name;
#Add This line
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
include fastcgi_params;
}
For me, this File not found problem was because i started php-fpm as sudo. First, stop it as sudo and then start without sudo.
Related
I try to create a small webserver on my raspberry pi with docker.
This is my docker-compose file:
version: "3.6"
services:
web:
image: nginx:latest
ports:
- "8080:80"
volumes:
- /home/pi/testData/code:/var/www/html
- /home/pi/testData/site.conf:/etc/nginx/conf.d/site.conf
depends_on:
- php
php:
image: php:7.4-fpm
volumes:
- /home/pi/testData/code:/var/www/html
And this my site.conf
server {
index index.html index.php;
listen 80;
server_name localhost;
error_log /var/log/nginx/error.log;
access_log /var/log/nginx/access.log;
root /var/www/html;
location ~ \.php$ {
try_files $uri =404;
fastcgi_split_path_info ^(.+\.php)(/.+)$;
fastcgi_pass php:9000;
fastcgi_index index.php;
include fastcgi_params;
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
fastcgi_param PATH_INFO $fastcgi_path_info;
}
}
Inside of the folder /home/pi/testData/code is a file named index.php which just echos a Hello World.
So this file will be mapped into the Docker Container into the path /var/www/html/ and inside of the site.conf is the root mentioned also to /var/www/html.
I also checked if the files are really available inside of the nginx container - both are.
So as I understand it: If I put an index-file inside of the /var/www/html folder of my nginx docker container, then this file should be displayed if I call the IP of my Raspberry pi on Port 8080.
But unfortunately I only receive the Welcome to Nginx Page.
Did I miss something or did I something wrong?
Nginx will load configuration files found in /etc/nginx/conf.d/ and their names ends with .conf as suffix in alphabetical order.
Your configuration does have conflicting server name with the default configuration file of nginx. /etc/nginx/conf.d/default.conf
you can check that by exec into nginx container and run nginx -t command to check the configuration for warnings and errors.
you should see something like:
[warn] 502#502: conflicting server name "localhost" on 0.0.0.0:80, ignored
nginx: [warn] conflicting server name "localhost" on 0.0.0.0:80, ignored
you should rename the default file to a something like default.conf.old or you can simply overwrite its contents with your configuration.
This seems to be a common issue when people setup a Docker Container to serve a Laravel App using the container.
My setup is mostly this (I removed some of the other services because I do no think they apply here)
#docker-compose up --build
version: "2"
services:
nginx:
image: nginx
depends_on: [pacs-1,mysql_db]
restart: unless-stopped
ports: ["443:443"]
volumes:
- ./nginx-home:/nginx-home
- ./tls:/etc/nginx/tls
- ./nginx/default.conf:/etc/nginx/conf.d/default.conf
php-fpm:
build: php-fpm
volumes:
- ./nginx-home:/nginx-home
The default.conf for nginx for Laravel project is:
server {
# implement the TLS
listen 443 ssl;
server_name dockerlaravel.medical.ky;
root /nginx-home/PortalRads/public;
access_log /nginx-home/logs/PortalRads-access.log;
error_log /nginx-home/logs/PortalRads-error.log info;
index index.php index.html index.htm;
client_max_body_size 4000M;
client_body_buffer_size 4000M;
# skip favicon.ico
#
#location = /favicon.ico {
#access_log off;
return 444;
#
}
location / {
try_files $uri $uri/ /index.php?$query_string;
gzip_static on;
}
location ~ \.php$ {
try_files $uri =404;
fastcgi_split_path_info ^(.+\.php)(/.+)$;
fastcgi_pass php-fpm:9000;
fastcgi_index index.php;
include fastcgi_params;
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
fastcgi_param PATH_INFO $fastcgi_path_info;
error_page 404 /error_pages/404.php;
error_page 403 /error_pages/403.php;
error_page 401 /error_pages/401.php;
error_page 500 502 503 504 /errors/50x.html;
}
}
Everything actually seems to work OK once I get it all setup. Composer is installed globally on the host, and I use composer update on the host side to update the package. It does not seem like it is necessary to install composer on Docker with the setup that I have since the php container has php and the artisan binary is at the root of my Laravel app.
The problem that I am having is that the symlink from public/storage to the storage folder does not seem to be working. I have tried various methods to create the link. I actually have separate containers for nginx and php, so artisan will not run in the nginx container.
Since I am using nginx as a webserver, wondering if it might just be possible to have another location block or server block that will basically create that connection on the server side so that I do not have to fiddle around the the symlink created by artisan ?
location /storage/ {
root . . .
. . .
}
I can get into those containers from the CLI using:
sudo docker exec -it orthanc-docker-dev_ris_php-fpm_1 /bin/bash
sudo docker exec -it orthanc-docker-dev_ris_nginx_1 /bin/bash
and I've tried created the symlink from within the container(s) and that does not seem to work either.
As an example, I am using JetStream with Livewire in Laravel and the profile photos there map to:
https://dockerlaravel.medical.ky/storage/profile-photos/image.png
But that gives a 404 error because the symlink must not be setup correctly.
There are quite a few posts about that issue. Just wondering what the "best" solution is, preferably one that can be automated during the build of the containers so the user does not have to worry about that.
Thanks.
I'm trying to set up a simple Grav site workflow using git, Docker and two containers: one for nginx and one for PHP. The idea is to git clone into my Digital Ocean droplet and run docker-compose up -d --build to build and serve the website.
I'm getting permission issues whenever I try to access the sites, and even Grav's documentation about troubleshooting permission issues does not help.
Here's my docker-compose.yml:
version: '3'
services:
web:
build:
context: .
dockerfile: ./docker/nginx/Dockerfile
ports:
- "80:80"
volumes:
- ./src:/var/www/html
links:
- php
php:
build:
context: .
dockerfile: ./docker/php/Dockerfile
volumes:
- ./src:/var/www/html
And here's nginx's Dockerfile:
FROM nginx:stable-alpine
WORKDIR /var/www/html
COPY ./src .
COPY ./docker/nginx/default.conf /etc/nginx/conf.d/default.conf
If that's any use, here's the nginx configuration I'm using:
server {
listen 80;
index index.php index.html;
server_name www.gravtest.test gravtest.test;
error_log /var/log/nginx/gravtest.test.error.log;
access_log /var/log/nginx/gravtest.test.access.log;
root /var/www/html;
location / {
try_files $uri $uri/ /index.php?$query_string;
}
location ~ \.php$ {
try_files $uri =404;
fastcgi_split_path_info ^(.+\.php)(/.+)$;
fastcgi_pass php:9000;
fastcgi_index index.php;
include fastcgi_params;
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
fastcgi_param PATH_INFO $fastcgi_path_info;
}
}
The PHP Dockerfile is simple, it just spawns php:7.3-fpm and installs a few dependencies like opcache, gd, etc...
Whenever I try to access the site via localhost, I get this error:
Fatal error: Uncaught RuntimeException: Creating directory failed for /var/www/html/cache/compiled/files/40779d000b68629af00dd987148afc06.yaml.php in /var/www/html/vendor/rockettheme/toolbox/File/src/File.php:325 Stack trace:....
Files are copied from the host to the container with the nginx:nginx owner, so I should be good, but looks like I'm not. I've tried setting folders/files chmod using Grav's documentation but no dice.
Am I missing something?
Answering my own question:
It turns out the images php-fpm and nginx do not use the same user, so the permission problem came from that. I simply had to add a new user to both Dockerfile, and run that container from that user.
So for PHP, my Dockerfile is now:
FROM php:7.3-fpm
# Install a few dependencies here...
COPY ./src /var/www/html
RUN addgroup --gid 1000 mygroup
RUN adduser --system --no-create-home --disabled-password --disabled-login --uid 1000 --ingroup mygroup myuser
RUN chown -R myuser:mygroup /var/www
USER myuser
And for nginx:
FROM nginx:stable-alpine
RUN addgroup --gid 1000 mygroup
RUN adduser --system --no-create-home --disabled-password --disabled-login --uid 1000 --ingroup mygroup myuser
WORKDIR /var/www/html
RUN chown -R myuser:mygroup .
USER myuser
And now everything works fine! :)
I'm using docker compose to boot up a development workspace, consisting of php, nginx and mysql. Everything boots, static html get's served, but when trying to start a laravel app, i get the following error:
The stream or file "/home/html/storage/logs/laravel-2019-06-10.log" could not be opened: failed to open stream: Permission denied
I searched around and it looked like a permissions issue? Do note, that the docker with just the database and the build in php server does seem to work.
My docker-compose.yml
version: "3"
services:
db:
image: mysql
command: --default-authentication-plugin=mysql_native_password
restart: always
environment:
MYSQL_ROOT_PASSWORD: "root"
ports:
- 3306:3306
php-fpm:
image: php:7.3-fpm-alpine
links:
- db
volumes:
- "./:/home/html/"
nginx:
image: nginx:1-alpine
ports:
- "8080:80"
links:
- php-fpm
volumes:
- "./site.conf:/etc/nginx/conf.d/default.conf"
- "./:/home/html/"
My nginx config:
server {
index index.php index.html;
listen 80 default_server;
error_log /var/log/nginx/error.log;
access_log /var/log/nginx/access.log;
root /home/html/public;
location ~ \.php$ {
try_files $uri =404;
fastcgi_split_path_info ^(.+\.php)(/.+)$;
fastcgi_pass php-fpm:9000;
fastcgi_index index.php;
include fastcgi_params;
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
fastcgi_param PATH_INFO $fastcgi_path_info;
}
}
Kind regards :)
Enter the php-fpm container:
docker-compose -i -t exec php-fpm /bin/sh
Then change access rights of storage folder:
chmod -r 777 /home/html/storage
Cause it's local development environment, correct rights doesn't matter.
This question already has answers here:
How do I get PHP errors to display?
(27 answers)
Closed 10 months ago.
I create Laravel PHP application in Docker. First I setup Laravel app using
laravel new laravelDockerApp
it creates successfully.I verify it's setup by built-in server
php artisan serve
Then setup Local environment with Docker
docker-compose.yml
version: '2'
services:
web:
build:
context: ./
dockerfile: web.docker
volumes:
- ./:/var/www
ports:
- "8080:80"
links:
- app
app:
build:
context: ./
dockerfile: app.docker
volumes:
- ./:/var/www
app.docker
FROM php:7-fpm
RUN apt-get update && apt-get install -y libmcrypt-dev mysql-client \
&& docker-php-ext-install mcrypt pdo_mysql
WORKDIR /var/www
web.docker
FROM nginx:1.10
ADD ./vhost.conf /etc/nginx/conf.d/default.conf
WORKDIR /var/www
vhost.conf
server {
listen 80;
index index.php index.html;
root /var/www/public;
location / {
try_files $uri /index.php?$args;
}
location ~ \.php$ {
fastcgi_split_path_info ^(.+\.php)(/.+)$;
fastcgi_pass app:9000;
fastcgi_index index.php;
include fastcgi_params;
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
fastcgi_param PATH_INFO $fastcgi_path_info;
}
}
I run docker-compose up -d command. app & web containers up successfully.When I check app in Browser using
localhost:8080
I got
500(Internal Server Error)
Please, can you help to solve this? Thanks.
I found a solution. I set the permission on my Laravel app using:
sudo chmod -R 777 storage && sudo chmod -R 777 bootstrap/cache
This solution might not apply to you since you are using Nginx, but in my case I am using the php:7.0-apache as source image, so I made the Apache user the owner of my app's files.
In my Dockerfile I have:
...
USER www-data
WORKDIR /var/www/html
COPY --chown=www-data:www-data . .
...
This solved the problem, so it could be worth trying, either modifying your Dockerfile or maybe Docker Compose has some option for user permissions when mounting volumes.
I ran into this today, and the issue for me was that while I'd created my directory structure, I'd failed to copy the .env file from .env.example. Copying this and hitting the webpage gave me a page which had this in the top right corner:
Clicking "Generate App Key" resolved this issue for me, but it's probably worth giving the .env file a once-over to make sure it's not got some other unset variables you'll need!