Mar 20, 2015 - Natman

Comments

I’ve added a new Dockerfile for to bring Natman to the game. Natman will allow us to automatically map a public port on our router to port 22 on the Raspberry Pi 2. This assumes that the router support either uPnP or NAT-PMP which many or most do. This will ensure that we have access to administer the device at its final location without having to manually configure the routers of our loved ones.

Unfortunately, I still need to figure out how to pass configuration into our containers, so in the mean time the natman Dockerfile need to be customized to include the public port to request for the port mapping and the local address of the router. I don’t want the project to standardize on a specific port, as having a known port number that gets forwarded to port 22 would be a security concern, if a vulnerablity in ssh is found and the local address of the router will vary from network to network.

Down the road, I intend to have the setup scripts pick a random port and report it back to the user as well as detect the local address of the router, but first we will dockerize and get BitTorrent Sync setup and running, so stay tuned…

Mar 15, 2015 - Integration

Comments

It’s all begining to come together! I’ve updated the setup.sh script to clone the repo and start building docker images which proves out the pattern for getting most of the way to v1. Next steps will be adding the rest of the docker image builds and then I can move on to providing a good pattern for supplying BitTorrent Sync with its config.

Mar 15, 2015 - Base Image Build

Comments

The first thing we need is a base docker image to build the dockerized applications on. Unfortunately, since the armhf archecture of the Raspberry Pi 2 isn’t officially supported by Docker or DockerHub, we are going to bulid our own.

A new script has been added to the repo based Dashamir Hoxha’s work on building a 32-bit docker base image. What it does is use debootstrap to create a base minimal debian install in a temp folder and packages it up to be imported into the local docker system. It take a while to run, but does a good job of listed out the tasks it is taking, so you know what is going on.

The docker images we will create to dockerize the applications for PDC will be based on this base image.