nginx+PHP-FPM about docker-webserver

[nemozar]

problem in line fastcgi_param

with fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name; work good.

in default config after pull line fastcgi_param SCRIPT_FILENAME /scripts$fastcgi_script_name;

[nazar-pc]

phpMyAdmin is self-container, so despite it uses Apache, it is irrelevant here.

From the log it seems like you don't have 1.php file in there. I'm wondering why did you put root /data/nginx/www; in each location instead of directly under server.

You can try to see if the file is there by entering Nginx container and examining contents:

docker-compose exec nginx ls -al /data/nginx/www

[nemozar]

docker-compose exec nginx ls -al /data/nginx/www

drwxr-xr-x 2 git git 4096 Jun 7 06:56 .
drwxr-xr-x 5 git git 4096 Jun 7 06:00 ..
-rw-r--r-- 1 git git 5 Jun 7 06:02 1.html
-rw-r--r-- 1 git git 12 Jun 7 06:37 2.php
-rw-r--r-- 1 root root 90 Jun 7 06:00 index.html

in default config root set in each location.

1.html is well opened

[nazar-pc]

Since files are in place and permissions are correctly it is purely Nginx configuration issue.
Did you restart Nginx after changing its configuration?

[nazar-pc]

Wait, you have 1.html, but not 1.php

[nemozar]

Did you restart Nginx after changing its configuration?

i will stop docker-compose and docker-compose up after. if i comment location ~ \.php$ { then php file open for download.

Wait, you have 1.html, but not 1.php it is test. for all *.php files i get error. How for exists or not exists in server.

[nemozar]

server {
    listen       80;
    server_name  localhost;
    root   /data/nginx/www;
    location / {
        index  index.html index.htm;
    }
    location ~ \.php$ {
        fastcgi_pass   php:9000;
        fastcgi_index  index.php;
        fastcgi_param  SCRIPT_FILENAME  /scripts$fastcgi_script_name;
        include        fastcgi_params;
    }
}

current config

[nazar-pc]

Can you post what is in logs when you try existing file, say, 2.php?
Because php_1 | 172.19.0.7 - 07/Jun/2017:07:39:58 +0000 "GET /1.php" 404 is actually correct answer for file that indeed doesn't exist.

You don't necessarily need to restart everything, you can just docker-compose restart nginx, since some other services like database might take longer and you shouldn't really wait for them as configuration is the same.

[nemozar]

drwxr-xr-x 2 git  git  4096 Jun  7 10:39 .
drwxr-xr-x 5 git  git  4096 Jun  7 06:00 ..
-rw-r--r-- 1 git  git     5 Jun  7 06:02 1.html
-rw-r--r-- 1 git  git     6 Jun  7 10:39 2.php
-rw-r--r-- 1 root root   90 Jun  7 06:00 index.html
-rw-r--r-- 1 git  git    12 Jun  7 06:37 index.php

php_1         | 172.19.0.7 -  07/Jun/2017:11:23:43 +0000 "GET /1.php" 404
php_1         | 172.19.0.7 -  07/Jun/2017:11:23:44 +0000 "GET /1.php" 404
php_1         | 172.19.0.7 -  07/Jun/2017:11:23:44 +0000 "GET /1.php" 404
php_1         | 172.19.0.7 -  07/Jun/2017:11:24:04 +0000 "GET /2.php" 404

when run docker-compose restart nginx

nginx_1       | 2017/06/07 06:31:48 [emerg] 1#1: unknown directive "server" in /data/nginx/config/conf.d/default.conf:1
nginx_1       | nginx: [emerg] unknown directive "server" in /data/nginx/config/conf.d/default.conf:1

config

server {
    listen       80;
    server_name  localhost;
    root   /data/nginx/www;
    location / {
        index  index.html index.htm;
    }
    location ~ \.php$ {
        fastcgi_pass   php:9000;
        fastcgi_index  index.php;
        fastcgi_param  SCRIPT_FILENAME  /scripts$fastcgi_script_name;
        include        fastcgi_params;
    }
}

[nazar-pc]

Did you change anything in /data/nginx/config/nginx.conf? [emerg] unknown directive "server" doesn't make any sense.

[nemozar]

not problem with EOL. if i comment block with php and restart nginx file *.php dowloaded (config file is well)

in file /data/nginx/config/nginx.conf i set new config

server {
	listen       80;
	server_name  localhost;
	root   /data/nginx/www;
	location / {
		index  index.html index.htm;
	}
	location ~ \.php$ {
		fastcgi_pass   php:9000;
		fastcgi_index  index.php;
		fastcgi_param  SCRIPT_FILENAME  /scripts$fastcgi_script_name;
		include        fastcgi_params;
	}
}

in error.log nginx

2017/06/07 12:02:11 [error] 10#10: *1 FastCGI sent in stderr: "Primary script unknown" while reading response header from upstream, client: 172.19.0.1, server: localhost, request: "GET /2.php HTTP/1.1", upstream: "fastcgi://172.19.0.6:9000", host: "localhost:81"

[nazar-pc]

This is not a correct config if you put virtual host right into /data/nginx/config/nginx.conf (see https://stackoverflow.com/questions/10934705/unknown-directive-server-in-etc-nginx-nginx-conf4)

What I'd suggest you to do is to restore original /data/nginx/config/nginx.conf from /etc/nginx_dist/nginx.conf and put your virtual host into /data/nginx/config/conf.d/default.conf. If should work this way.

[nazar-pc]

I do not change that particular line, it comes from base image. I typically replace the whole virtual host file with my own.

[nemozar]

Help me with php extension. How i understand me need instal in nazarpc/php:fpm container? but when i open him with exec bash and write apt-get install php-pgsql get error "Unable to locate package php-pgsql"

[nazar-pc]

This is a separate question and you should better ask it in separate issue so that it would be possible to to find it for other people.
The question about installing additional extensions was already asked here: #20