morninj / django-docker Goto Github PK

• store secrets in secrets.py (and add to gitignore); this avoids storing them in the docker image
• use fabric to:
  • build image
  • push to docker hub
  • ssh to production server
  • pull image from docker hub
  • do docker run and set secrets as environment variables

In the README section on the development setup I followed your instructions but could not get the local files mounted in the container. So the container never ran gunicorn.

After a good deal of research I found that boot2docker, the system running in the docker-machine VM, looks for a share folder named /Users on the host and only allows you to bind-mount files under /User. So to make this work you have to define a share folder in VirtualBox called /Users then the -v argument would have to always specify the local path as being starting with /Users.

The README makes no mention of this. How did you ever get that -v bind-mount option to work? Your host was running Linux, right?

Thanks for the code examples. What license does the project use?

https://docs.docker.com/articles/dockerfile_best-practices/

System check identified some issues:

WARNINGS:
?: (1_8.W001) The standalone TEMPLATE_* settings were deprecated in Django 1.8 and the TEMPLATES dictionary takes precedence. You must put the values of the following settings int

IF DJANGO_PRODUCTION=true, run python manage.py check --deploy

I did all the configuration and then

docker run -d -p 80:80 -v $(pwd):/code --env DJANGO_PRODUCTION=false /django-docker

On running docker, the container is spawned, but supervisor cannot start the service. I am getting following error:

2016-05-01T22:05:58.420603194Z 2016-05-01 22:05:58,420 INFO exited: initialize (exit status 127; not expected)
2016-05-01T22:05:59.423536295Z 2016-05-01 22:05:59,423 INFO spawned: 'initialize' with pid 102
2016-05-01T22:05:59.427550188Z 2016-05-01 22:05:59,427 INFO success: initialize entered RUNNING state, process has stayed up for

than 0 seconds (startsecs)
2016-05-01T22:05:59.428110966Z 2016-05-01 22:05:59,428 INFO exited: initialize (exit status 127; not expected)
2016-05-01T22:06:00.431900979Z 2016-05-01 22:06:00,430 INFO spawned: 'initialize' with pid 103
2016-05-01T22:06:00.435845297Z 2016-05-01 22:06:00,435 INFO success: initialize entered RUNNING state, process has stayed up for
than 0 seconds (startsecs)
2016-05-01T22:06:00.436428078Z 2016-05-01 22:06:00,436 INFO exited: initialize (exit status 127; not expected)
2016-05-01T22:06:01.439192191Z 2016-05-01 22:06:01,438 INFO spawned: 'initialize' with pid 104
2016-05-01T22:06:01.443373806Z 2016-05-01 22:06:01,443 INFO success: initialize entered RUNNING state, process has stayed up for
than 0 seconds (startsecs)

I looked into initialize.sh script but couldn't figure out the issue. Can you please help me debug this?

hi
I want to use this with ssl on my project. how can I change nginx.conf and docker-compose.yml to use https

The requirements.txt file doesn't specify versions so running this installs django 1.10. The urls.py file imports patterns from django.conf.urls, but patterns has been deprecated. Explicitly setting the version of django in the requirements.txt file (to a version lower than 1.10) fixed this issue for me.

Hey,

First of all, thanks for this repo. Looks really good. One thing that I can't completely understand about your development and deployment workflow is that what would happen when I have a container deployed on the production server and then I have changed some models and I want to deploy them.

Obviously, current Dockerfile is going to install a new mysql instance which doesn't have the old data on the previous container so makemigrations won't be useful in this case.

Could you please explain what is the right workflow in this scenario?

• in supervisord.conf
• don't start if DJANGO_PRODUCTION is true

See Django deployment checklist.