Material guidance for Containers event in fCC Algarve

A bit of a word of advice about the following guide. I am an amateur with containers and only discovered them recently. So some information might need correction. It would help me tremendously if you communicate my mistakes so I can proceed to correct them in this guide.

We've finally built a new software on time. The software runs on our machine, a laptop with a SuperAlien OS and it's written in Hyeroglyphs++.

We want to show it to our CEO to impress him. But he's on the other side of the world. And his machine runs Banana OS because his computer sucks. We send him our project but now he has to install all this mumbo jumbo for the program to run on his machine.

And it might not even work at the first time, he's running out of patience and you're running out of confidence.

OR in the real world...

Imagine trying to install a new piece of software, for example an SQL server or a local SMTP email server to deliver emails. You'd have to follow a certain amount of steps to prepare your machine to have these install locally and you might even run into some errors or incompatibilities.

Containers are isolated pieces of software that use some resources from the host machine plus the host's kernel to deliver you the same services, environment and/or specs without the need to install dependencies, software like NPM or Composer and run into problems and precious time wasting.

Virtual machines require a guest operating system running alongside your own. They take a bit of time and a bit of resources to set up. Not the case with containers. You can mount a container or network of containers from the same setup and connect them to your local code to deliver an environment as equal as possible to the experience you'll get in a production environment. It requires less resources hence faster.

Images are single immmutable files that have all the dependencies and configurations container needs to run.

With Docker as the container software, we script our images using a file named Dockerfile where we pull images from the Docker Hub, execute commands, copy our projects into containers and run them in.

We can use yaml files called docker-compose.yml to make it easier to configure our containers. Then all we have to do is call a simple instruction docker-compose up

There's a desktop software called Docker Desktop that...

Right let's stop talking and get on with the show.

First, we'll decide which container software we are going to use. Some use DevContainers, others Docker itself. For the purpose of this event we're going to use Docker.

1. Docker Instalation

Windows and Mac users will probably need to install Docker Desktop right away as it contains the CLI and the desktop software to easily visualize running containers. We're going to need the Docker CLI and you'll probably need to create an account to fetch docker images from the hub.

a) Docker CLI or Desktop b) Create a Docker account or use existing account

https://docs.docker.com/get-docker/

2. Docker is Installed, let's test it.

docker version

This should output Client and Server information. If something's wrong you might need to reinstall docker.

3. List running containers

docker ps

It will display you Container IDS and the image they're coming from. Container IDs are important as they serve as reference in other docker commands like run, stop or kill.

Right, let's go to https://hub.docker.com/ and search for the container Hello world for instructions.

1. To run a container from an image:

docker run <image-name>

If you don't have the image into your machine, docker will fetch it from the docker hub and then create a container from it.

2. To just pull the image to our machine:

docker pull <image-name>

This will download the image without running it into a container.

We want to run the hello world image and test if docker daemon is running. If everything works ok, docker will output you a block of text and exit by itself.

But what exactly is inside a container? Let's search for a precise docker image that lets us execute a command.

docker run busybox ls

You'll see a filesystem root similar to linux that is not of your machine. It's a filesystem snapshot from the image.

We want to list all containers we have available in docker, not just the running ones.

docker ps --all

We can restart containers with the container ID provided.

docker start <container-id>

To delete all stopped containers that are just eating disk space

docker system prune

Now let's try running a redis server.

docker run redis

It will pull the redis image and start a container with a redis server. It will indicate you the port available to communicate.

One more important command if you want to delve into a container's filesystem in it's own terminal

docker exec -it <redis-container-id> sh

You'll be able to see redis in action and execute command lines inside the container. Run redis-cli inside the shell's container.

redis-cli

The redis cli prompt will come up and you can toy with it.

...:6379> set suncolor blue ...:6379> get suncolor

Now run exit to get out of the redis-cli and exit again to quit the container shell. Do a system prune if needed to clear our images.

Let's continue in the next phase, Dockerfiles.