#Galaxy on Docker

##Why?

1. Independence from host OS, packages and libraries -> portability (PC, cluster, cloud)
2. Lightweight and transparent (dockerfile) compared to VMs
3. For reproducible distribution of a workflow (including datasets)

##How?

there are many ways to build a Galaxy docker image:

###a) roll your own (without dockerfile) Ideally your image would be built using a dockerfile, we experiment here with the shell to understand what would be needed

Step 0. Connect to VM:

Step 1. Install docker (Ubuntu example):

sudo apt-get update
sudo wget -qO- https://get.docker.com/ | sh
#alternatively you can install from package but that tends to be an older version 
#sudo apt-get install docker.io

#add user to docker group
sudo usermod -aG docker ubuntu
#at this point you might want to reconnect to the instance
#exit
#reconnect

Step 2. Create new docker image

sudo docker run -i -t --name=milky_whale ubuntu:14.04 /bin/bash

Step 3. Install Galaxy

sudo apt-get install git
sudo apt-get install python
#lynx for testing web interface
sudo apt-get install lynx

git clone https://github.com/galaxyproject/galaxy/

Step 4. Run Galaxy

cd galaxy/
sh run.sh

Step 5. Test Galaxy

on a different shell from host

sudo docker ps
sudo docker exec my_container_id lynx 127.0.0.1:8080

Step 6. Commit image

sudo docker commit -m "installed Galaxy" milky_whale milky_whale:v2
sudo docker stop milky_whale

Step 7. Test new instance run committed image

sudo docker run -i -t -p 8080 --name=milky2 milky_whale /bin/sh /galaxy/run.sh

on a different shell

sudo apt-get lynx
lynx 127.0.0.1:8080

###b) Roll your own (with Dockerfile)

here we will do something similar as in a) in an automated manner

Create a new directory my_galaxy

mkdir my_galaxy

Download Galaxy (will be added to the docker image)

git clone https://github.com/galaxyproject/galaxy/

Create Dockerfile

nano Dockerfile

Step 1. Define base image

Step 2. Update packages

Step 3: Add the package verification key

RUN apt-key adv --keyserver hkp://keyserver.ubuntu.com:80 --recv-keys 36A1D7869245C8950F966E92D8576A8BA88D21E9

Step 4: Install packages git, python and lynx (Hint: use the apt-get -y flag for auto)

A possible solution: Dockerfile

build docker image from dockerfile

sudo docker build -t my_galaxy .

run image (intaractively)

sudo docker run -it my_galaxy

connect to image from browser through URL: http://127.0.0.1:8080 (locally) or http://host.ip.add.ress:8080 (if you have a routable ip address) from a command-line based browser (just to check if it is running)

sudo apt-get install lynx
lynx 127.0.0.1:8080

###c) start from a pre-canned one Björn A. Grüning (bjoern.gruening@gmail.com) has kindly assembled a wide variety of Galaxy docker images available at: https://github.com/bgruening/docker-galaxy-stable and https://github.com/bgruening/docker-recipes/

to run:

docker run -d -p 8080:80 -p 8021:21 bgruening/galaxy-stable

A good description, and a tutorial could be found on the GitHub page. Also worth looking into the Dockerfile on how the docker image was built.

##More info: ###Cheat sheet https://github.com/wsargent/docker-cheat-sheet