Image Docker Distribution

USE CASE 1 - Hello World

So rich man wants deploy a server with following characteristics:

Amazon EC2 - t2.2xlarge (expert reviewed)

• vCPU: 8
• MEM: 32GB

The application that he wants it is a variation of cowsay, called dockersay but he needs a change in this app name to trumpsay.

Step 1 - Run my first container

docker run --rm docker/whalesay cowsay hello world

Step 2 - My first container with volumes

docker run --rm \
  -v `pwd`/trump.cow:/cowsay/cows/trump.cow \
  docker/whalesay cowsay -f /cowsay/cows/trump.cow "BOOM"

Step 3 - Create an alias

alias trumpsay="docker run --rm \
  -v `pwd`/trump.cow:/cowsay/cows/trump.cow \
  docker/whalesay cowsay -f /cowsay/cows/trump.cow"

trumpsay BURN FU BURN, BOOM

Step 4 - Container debug

docker run -it --rm docker/whalesay bash

(pwd && ls -lh)
(cd cows && ls)

cowsay -f turkey.cow aaa

USE CASE 1.5 - Deploy meteor app

Now lets deploy a meteor app locally

Step 1 - Meteor project app

First we need a meteor project app, so lets use this zip file:

• todos.tar.gz

Step 2 - Pull mongo and node images

Our meteor project runs with node 4.8.1 and mongo 3:2 so we need the next images:

• NodeJS:4.8.1
• MongoDB:3.2

Step 3 - How to create a docker image

Lets see the Dockerfile for this example

• Dockerfile

Run this command to create an image:

docker build -t todos:1.0.0 .

Step 4 - How to deploy mongo

Lets run mongo image

docker run -d -p 27017:27017 --restart=always \
    --name mongo \
    -v `pwd`/database/:/data/db \
    mongo:3.2

Step 5 - How to deploy node (meteor app)

Lets run node image

docker run -d -p 3000:3000 --restart=always \
  --link mongo \
  -e MONGO_URL=mongodb://mongo:27017/todos \
  --name=app \
  todos:1.0.0

Step 6 - Loggin mongo and bash console

Mongo logs

docker logs mongo

Mongo console

docker exec -it mongo bash

USE CASE 2.0 - Happy north korean

So happy man wants to deploy ToDos App in North Korea, the server has the following specifications:

Unknown but same to “EC2 - t1.micro”

• vCPU: 1
• MEM: 0.6GB

They have not a Internet connection, but there is a USB port in the server.

USE CASE 2.1 - Resources & Limitations (RAM & CPU)

Step 1 - Create our stress dockerfile to verify that limiting resources is working

FROM fedora:latest
RUN yum -y install stress && yum clean all
ENTRYPOINT ["stress"]

Step 2 - Build our Dockerfile to hava an "stress" image

docker build -t stress .

Step 3 - Verify our image is available

docker images

Step 4 - Check if our stress image works

docker run -it --rm stress --cpu 4

Step 5 - Limit the cpu to use only the first core

docker run -it --rm --cpuset-cpus=0 stress --cpu 4

Step 6 - Limit the cpu to use only the first and third core

docker run -it --rm --cpuset-cpus=0,2 stress --cpu 4

Step 7 - As a test lets see how --cpu-period --cpu-quota behave

docker run -it --rm --cpu-period="100000" --cpu-quota="150000" --cpu-cpus=0 stress --cpu 4

Step 8 - Now let's limit the memory

docker run -it --rm -m 200m stress --vm 1 --vm-bytes 400M --vm-hang 0

Step 9 - Get our container name

docker ps

Step 10 - Verify memory limit with docker stats

docker stats [container_name]

USE CASE 2.2 - Share image as file

Usage

$ docker save [OPTIONS] IMAGE [IMAGE...]

Step 1 - Create a backup that can then be used with docker load.

docker save todos:1.0.0 > todos.tar
docker save --output todos_2.tar todos:1.0.0
docker save todos:1.0.0 | bzip2 > todos_3.tar

Step 2 - Verify image files

ls -sh todos.tar

Step 3 - Copy file

cp todos.tar /path/to/usb/todos.tar

Step 4 - Remove image from host

docker rmi todos:1.0.0

Step 5 - Load an image from a tar archive or STDIN.

docker load < todos.tar
docker load --input todos.tar

Step - Create container

docker run -d -p 3000:3000 --restart=always \
  --link mongo \
  -e MONGO_URL=mongodb://mongo:27017/todos \
  --name=app \
  todos:1.0.0

USE CASE 3.0 - Share image using registry

So sad man wants to deploy ToDos App in Russia, the server has the following specifications:

Amazon EC2 - t1.micro

• vCPU: 1
• MEM: 0.6GB

They have an good Internet connection, also they want periodical updates.

Step 1 - Deploy my first registry

docker run -d -p 5000:5000 --name registry registry:2

Step 2 - Tag a docker image

docker tag alpine localhost:5000/myfirstimage

Step 3 - Push my first image into local registry

docker push localhost:5000/myfirstimage

Step 4 - Check images

docker run -it --rm --net=host centos bash

curl http://localhost:5000/v2/_catalog

Step 5 - Stop & Remove registry

docker stop registry && docker rm -v registry

USE CASE 3.1 - Share image using registry

Step 1 - Deploy a registry

docker run -d -p 5000:5000 --restart=always \
      --name registry \
      -v `pwd`/data:/var/lib/registry \
      registry:2

Step 2 - Tag my image

docker tag todos:1.0.0 localhost:5000/todos:1.0.0

Step 3 - Push todos:1.0.0 into registry

docker push localhost:5000/todos:1.0.0

Step 4 - Remove image from my local cache

docker rmi todos:1.0.0

Step 5 - Pull todos:1.0.0 from local registry

docker pull localhost:5000/todos:1.0.0

Step 6 - Stop & Remove registry

docker stop registry && docker rm -v registry