Nignix container doesn't start, and pb between phpfpm and nginx

[mathieutu]

Hi,
I'm new to docker, and I would have one nginx docker, and several others, especially php-fpm and node.js.

I want to use your images of nginx and php-fpm for that.

You built both to work together, but the volumes and names are different :
the "php" container have an image with the volume "/data/www" and the "phpfmp" one that nginx is expecting have the volume '/data/nginx/www'.

Even in the your Nginx-docker's readme, you have the both : /data/www (in the nginx conf) and /data/nginx/www in the run command.

Is there an error, or do I have missed anything ?
Should I re build the phpfpm image to use the same volume that nginx or just run the container with other volume arguments ?

I've try some things, but for now, the nginx container never started. No message or error, just started and stopped in the same time.

Thank you for your help,
Matt'

[alond-winner]

Try to run:

docker logs nginx

and you'll see any errors that the nginx container had.

[phpguru]

In my case, I had to create config/logs dir and config/logs/error.log (empty text file) before the nginx container would run. There is a small error in the docker-phpfpm README.md though, that is, the run command should have --name phpfpm instead of --name php for the --link command to work in the nginx container's run command. (Or, update the nginx container run command to --link php.)

Oddly enough, I have both containers running (on Windows Docker no less) yet, nginx refuses to serve any files, and there's nothing in the logs about it, either.

One other confusing thing is, I tried enabling the services via systemd and found these containers to be not systemd-based, so systemctl doesn't exist.

Otherwise, really great work. I was thinking about forking and submitting a PR for the documentation updates (and including my run command for Windows-based Docker hosts.)