Kat Passen

 Open source platform to manage deployment  Acts as a layer between an OS and an application  Images not dependent on system – similar to the JVM  Provides a “global standard” to code against

 VMs require apps to be containered within the guest OS  Potentially 1:1 on VM:application  Docker sits between the app and the host OS  Takes place of the hypervisor  Runs images as an isolated process  Only the app and its libraries are compartmentalized instead of requiring the guest OS share that space as well

 Installation:  Available on a variety of platforms: ▪ Your favorite *nix ▪ Windows ▪ Rackspace, EC2, G-Cloud, etc

 Requirements:  iptables  git  procps or similar  XZ Utils  A cfgroups hierarchy  Kernel requirements vary by distribution, general recommendation is v3.8+

 Important: Docker runs as root Giving non-root access The docker daemon always runs as the root user, and the docker daemon binds to a Unix socket instead of a TCP port. By default that Unix socket is owned by the user root, and so, by default, you can access it with sudo. If you (or your Docker installer) create a Unix group called docker and add users to it, then the docker daemon will make the ownership of the Unix socket read/writable by the docker group when the daemon starts. The docker daemon must always run as the root user, but if you run the docker client as a user in the docker group then you don't need to add sudo to all the client commands.

 Docker Engine  Client-server application  Handles running images, registries and containers  Docker Hub  Public Docker registry  Holds a collection of existing images to be used publicly  Image examples: Ubuntu, Node.js, MongoDB, WordPress

 Integrates with Github and Bitbucket  Operates similar to them as well  Hosts images that can be used and contributed to by the Docker community  Location that Docker pulls and pushes to

 Tutorial:  Operates similar to any other *nix command line application

 Interactive containers:  $ sudo docker run -t -i ubuntu:14.04 /bin/bash ▪ -t creates a pseudo-terminal ▪ -i captures STDIN  Results (example):  pwd / ls bin boot dev etc home lib lib64 media mnt opt proc root run sbin srv sys tmp usr var

 Hello world daemon  $ sudo docker run -d ubuntu:14.04 /bin/sh -c "while true; do echo hello world; sleep 1; done“ ▪ -d runs containers in the background (daemonizes them)  This example returns a container ID, which you can use to interact with the container running that command  To stop a container, use docker stop $container_name

 Docker’s example runs a Python Flask app:  $ sudo docker run -d -P training/webapp python app.py ▪ -P: map required network ports inside the container to the host  Check results with sudo docker ps –l : ▪ CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES bc533791f3f5 training/webapp:latest python app.py 5 seconds ago Up 2 seconds :49155->5000/tcp nostalgic_morse ▪ -l: tells docker ps to return the last container started

 Ports: super malleable  -P is a shortcut for –p 5000  Docker generally picks a high external port to map to  Can run 1:1 on your port mappings, but this way you can have multiple apps thinking they’re on the same port  Lookup: use docker port $container_name $port

 View logs: docker log $container_name  Add -f and it behaves the same as tail –f  Other neat things:  docker top – same as standard top  docker inspect – returns a JSON hash of config and status info  docker tag – set a tag on an image (e.g. ‘devel’)

 Start/stop/restart  docker start $container_name  docker stop $container_name  docker restart $container_name  Removing things  Containers: ▪ docker rm $container_name ▪ Container in question has to be stopped first  Images: ▪ docker rmi $image_name  All deletions are final

 Updating and committing to an existing image:  Create a container from the image you want to update  Add what you want to the container, exit  Commit changes: ▪ $ sudo docker commit -m="Added json gem" -a="Kate Smith" 0b2616b0e5a8 ouruser/sinatra:v2 ▪ Params after flags: source container ID and user/target_name:tag

 Build an image from a Dockerfile  Dockerfile: set of instructions telling Docker how to create an image  Takes the format of INSTRUCTION statement ▪ # This is a comment FROM ubuntu:14.04 MAINTAINER Kate Smith RUN apt-get update && apt-get install -y ruby ruby-dev RUN gem install sinatra

 Build an image from a Dockerfile (cont)  In the same directory, run docker build – t=“$image_name” $path ▪ $ sudo docker build - t="ouruser/sinatra:v2".  Now you can create a container from the new image!  Size caveat: images have a maximum of 127 layers



 In no particular order:  Dist::Zilla  Rocket  Bamboo  Jenkins  Chef  Puppet  Rex