danbuchan / adv_web_dev_topic_10_docker

topic 10 content

Basic Container for Coursera Dockerfile

The following builds a basic docker container containing python3, nginx and visual Studio server. Supervisord is used to handle starting nginx and VS Server. Nginx acts as a proxy/firewall and handles mapping http requests from outside the container to applications running inside the container.

Make sure you're passingly familiar with docker. A small number of useful docker commands have been summarised in the useful_docker_commands.md markdown.

You should be familiar with linux and specifically ubuntu and the apt-get ecosystem as that will make the Dockerfile make more sense.

You must also be familar with the coursera docker build process. Anything you successfully build must also be run through their build tool to check for safety/security. https://github.com/coursera/coursera-labs. Roughly though, get your docker image working locally. Set up the coursera tool. Then use the run-lab/run.sh to build a version and if that works you can zip up your docker directory and upload your docker info to coursera labs.

To upload to coursera you need to provide a zip of the directory (and all its contents) that contains the Dockerfile, zip -r name.zip .. Note that you should make the zip file inside directory that contains the dockerfile, don't zip the containing directory as the coursera docker build process assumes the Dockerfile is in the vary root of the zip.

Things of note

build.sh

Somewhat over engineered way to call the docker build process

manifest.json

Used by the coursera docker build tool. Note that the name of the project in the build.sh and here is different. Make sure that's always the case to stop any namespace collisions in your docker images list/set when working locally

reverse-proxy.conf

configures nginx to map http requests in the browser to the applications running inside you container. So Visual Studio is running on port 8443 and is being mapped to any request for http at /

supervisord.conf

configures which services, daemons and whatnot are started at runtime when a container starts

Build

sh build.sh

Run and access

docker run -p 8888:8000 <name_image>

docker run -p 8888:8000 coursera_master:dev

Access the app at root of localhost:8888 for the Visual Studio interface. Add the path as defined in the