Unable to bind listening socket for address '127.0.0.1:9000': Address in use (98)

[cashel]

I've created and run the docker using the following settings:

sudo docker run \ --name=phlex \ --restart=always \ --detach=true \ --net=host \ -v /docker/containers/phlex/config:/config \ -e PGID=1000 -e PUID=1000 \ -e TZ="America/Los_Angeles" \ -p 5666:5666 -p 5667:5667 \ digitalhigh/phlex

The docker starts without issue, but loading the web UI results in a 502 Bad Gateway error. Docker logs report a stream of the following:

[06-May-2017 19:53:05] ERROR: unable to bind listening socket for address '127.0.0.1:9000': Address in use (98) [06-May-2017 19:53:05] ERROR: FPM initialization failed

It seems this issue is a known issue, but I'm afraid I'm not quite sure how to address it as stated in this bug report: https://github.com/abiosoft/caddy-docker/issues/6

[d8ahazard]

@cashel

Could you please try with the latest image? I've removed the port mapping requirement. If you create a container using the following command, what happens?

docker create --privileged \ --name=Phlex \ --net=host \ --restart=always \ -v /docker/containers/phlex/config:/config \ -e PGID=1000 -e PUID=1000 \ -e TZ="America/Los_Angeles" \ digitalhigh/phlex

[cashel]

I cleared everything out and pulled a new image, still same results:

[08-May-2017 23:52:25] ERROR: FPM initialization failed [08-May-2017 23:52:26] ERROR: unable to bind listening socket for address '127.0.0.1:9000': Address in use (98)

One thing that's come to mind - I do have a few other Dockers running UIs, and one specifically, Organizr, is mapped to port 80 it looks like:

lsiocommunity/organizr "/init" 2 days ago Up 22 hours 0.0.0.0:80->80/tcp, 443/tcp organizr

Would this be the issue? Workarounds?

[tboyce021]

I'm having issues with this as well. I tried pulling all the latest changes but it still never finds any Chromecasts without net=host. However, I have other things running on port 9000, so trying to run with net=host gives me the "Address in use" error.

[Steven-Harris]

Has anyone ever figured this out?

[emqMalte]

Port 9000 is being used for 'fastcgi_pass' which, is specified in the default file fastcgi_pass 127.0.0.1:9000;

and also in /etc/php7/php-fpm.d/www.conf inside the container changing both to a different port freed up port 9000 for me again

[tboyce021]

It would be nice if the port was configurable when creating the container, e.g. via an environment variable, so that it can work without going in and making manual modifications.

[d8ahazard]

@cashel @tboyce021 @Steven-Harris @EMQcarlos

https://forum.linuxserver.io/thread-495.html

I'm trying to get LSIO to make a proper image so this can become a non-issue. Pop over and show some support! :D

[tboyce021]

FYI this still isn't working completely with the environment variables. The port for fastcgi_pass gets set to FASTCGIPORT, but the one in /etc/php7/php-fpm.d/www.conf is still set to 9000.