Containerizing With Docker Engine

2 hours

Docker containers are self-contained units that include everything an application needs to run, from its code to its libraries and dependencies.

Simply put, it solves the "It worked on my machine" problem.

Docker Concepts

Docker Engine vs Docker Desktop

Docker Engine is a lower-level tool, geared towards system administrators. On the other hand, Docker Desktop provides developers with a user-friendly interface and additional tools.

Installing Docker Engine

The easiest way to install Docker is with a convenience script they provide.

curl -fsSL https://get.docker.com -o get-docker.sh
sudo sh get-docker. sh

Pulling a Docker Image

Pulling an image from a public repository like Docker Hub allows you to quickly deploy a pre-built application without manually installing all the dependencies. Let's draw the hello-world image.

sudo docker pull hello-world

Checking the images

sudo docker images

Running the image

sudo docker run hello-world

Running an Interactive Image in the Background

In most cases, you'll likely use the -d or -its flags along with the run` command.

1. Use -d for background processes or services.
2. Use -it to start a container in the foreground and interact with it like a regular terminal session.

Docker --Help

Use sudo docker --help to see all commands and sudo docker <command> --help to see all available flags(options) for the command.

Usefull Docker Commands

Installing a MYSQL container

Running a MYSQL database in a container prevents conflicts with other software or system-wide configurations.

docker pull MySQL

Running a MySQL container

docker run -e MYSQL_ROOT_PASWORD=password --container name -d -p 3306:3306 image-name

1. -e Sets environmental variables.
2. -d Executes the database in the background.
3. -p Opens a port to communicate with the database.

Accesing MYSQL to create a Database

Executing bash inside the MySQL container to manage the database.

docker exec -it container-name bash 

MYSQL client

Another way to interact with the database is through the Mysql client.

apt -y install mysql-client

Login to MYSQL to create a new database

After accessing bash we must login into MySQL to be able to manage our databases via SQL. After inserting the command below we will be prompted to insert the root user password.

MySQL -u root -p --protocol=tcp

Troubleshooting connection issues

Docker assigns an IP to every container to be able to connect with other distributed systems independently. To troubleshoot any connection issues we must use the inspect command.

docker inspect container-name

Data storage in containers

Containers are ephemeral, meaning their data is lost when the container stops. Mounts allow you to persist data beyond the container's lifecycle. A mount is a way to share data or files between the host system and the container. It essentially creates a link between a directory or file on the host machine and a directory within the container.

docker run -e MYSQL_PASSWORD=your-password --name container-name -d -p 3306:3306 --volume=/host-path -A:/container-path image-name

Mount Types

1. Bind: Binds a directory from the container to a directory in the host (example above)

docker run -v /hostdir:/containerdir image-name

2. Named: Manually created volumes inside a standard directory.

docker volume create volume-name # Creates volumes
docker run -v volume-name:/container-directory image name # Refeerences the created volume

Pulling an Apache Container

Apache, also called Apache HTTP Server, is free and open-source software that powers many websites. When you type a web address into your browser, Apache receives the request and fetches the relevant files (like HTML, CSS, and images) that make up the webpage.

It then sends this information back to your browser, which interprets it and displays the webpage.

docker pull httpd # Pulls the official Apache container from the docker hub

Interesting Fact

Apache is a core component of the LAMP stack, a popular combination of open-source software for building websites. LAMP stands for Linux (operating system), Apache (webserver), MySQL (database), and PHP (programming language).

Bindind a local volume to an Apache Web Server Container

By binding a local folder on your machine (where you're developing your website) to the Apache container's document root (the folder where it looks for website files), you can make changes to your code and see them reflected immediately in your web browser without needing to rebuild the container image or copy files back and forth. This allows for a much faster development workflow.

docker run --name container-name -d -p 80:80 --volume=host-volume-path -A:/usr/local/apache2/htdocs httpd

1. host-volume-path: Where the website files are
2. -d: Runs the container in the background
3. /usr/local/apache2/htdocs: The standard directory for file storage in the Apache container.
4. httpd: The image name

CPU and Memory Optimization

Resource optimization is important for containerized apps since an over-expanding container could hinder the performance of the whole system.

docker update container-name -m 128M --cpus 0.2
docker stats container-name # Provide container stats

1. -m 128M: Limit the memory to 128mb
2. --cpus 0.2: Limit the cpu usage to 20%

Stress Testing

Stress testing helps us understand how a containerized application behaves under heavy load. This can reveal bottlenecks in CPU, memory, or disk usage. The command stress is a stress test generator that simulates heavy workloads to see how your system responds.

apt install -y stress
stress --cpu 1 --vm-bytes 50m --vm 1 --vm-bytes 50m  

1. --cpu 1: Cpus to stress
2. --vm-bytes: Bytes volume to stress
3. --vm 1: Memory stress

Docker Info, Logs, Top

Here are some useful docker commands for managing containers.

docker info # Shows server info
docker logs container-name # Shows container logs
docker top container name # Shows container processes

Networking

In containerized environments, like Docker, multiple containers are often used to build applications. Each container typically runs a single service, but these services often need to communicate with each other to function properly. This is where container networking comes in.

docker network -ls # List networks
docker inspect network-name # List containers within the network
docker network create new-network # Create a new network

Container isolation: Network isolation allows you to define which containers can communicate, and how. For example; a container running a database might not need access to the internet. Also, if a container gets compromised by malware, network isolation can limit the attacker's ability to spread to other containers on the system.

docker run -dti -name container-name --network network-name image-name # Creates a container inside a defined network

Dockerfile

A Dockerfile is a text document that contains instructions for building a Docker image. It essentially acts like a recipe that Docker follows to create a customized environment for your application to run in.

Common instructions involve:

1. FROM: Specifying the base image to start from (like Ubuntu or a pre-built image).

2. COPY: Copying files from your local machine into the image.

3. RUN: Executing commands within the image to install dependencies, configure the environment, etc.

4. EXPOSE: Defining ports that the application running in the container will listen on.

5. CMD or ENTRYPOINT: Specifying the command to run when the container starts (like launching your application).

6. Creating a Dockerfile

nano dockerfile # Creates a dockerfle

2. Editing a Dockerfile example that creates an Ubuntu and Python image.

FROM ubuntu
RUN apt update && apt install -y python3 && apt clean
COPY app.py /apt/app.py
CMD phyton3 apt/app.py

3. Building a Dockerfile

The basic syntax for docker build is:

docker build [OPTIONS] <context>
# Example
docker build. -t ubuntu-python

1. <context> is typically the path to the directory containing your Dockerfile. By default, the current directory is used if not specified. There are several options available to customize the build process, such as:
2. -t: Tag the resulting image with a name and version.
3. -f: Specify an alternative Dockerfile location.

Docker Compose

A tool used to define and run multi-container applications. Employs a YAML file to configure "services" (essentially, containers) and their relationships. Docker Compose revolves around three main components that work together to define and run your multi-container application:

1. Services: These are the fundamental building blocks of your application, representing individual containers.
2. Networks: Docker Compose can create virtual networks for your containers to communicate with each other securely. This eliminates the need for complex manual network configuration.
3. Volumes: As mentioned earlier, volumes allow you to persist data outside the container itself. You map a directory on the host machine to a directory within the container. This ensures data isn't lost when the container restarts.

Dockerfile vs Docker Compose

In essence, Dockerfile provides the building blocks (images), and Docker Compose orchestrates them to run a multi-container application efficiently.

Installing Docker Compose

Before executing the yaml file it's important to install docker compose. This task is achieved by coding the following command into the terminal.

apt install docker-compose

Create the Compose file

After installing docker-compose, we must create the YAML file containing the infrastructure we desire to create. The script below shows an example of an Apache web server container deployment.

version: '3.9'
services:
  Apache:
    image: httpd:latest
    container_name: my-apache-app
    ports:
    - '8081:80'
    volumes:
    - ./website:/usr/local/apache2/htdocs

Executing the script

Finally, we must execute the script so our infrastructure( in this case our Apache web server) will be created. We can do that by typing the following command into the console.

docker-compose up -d

Conclusion

Docker offers a standardized approach to packaging, deploying, and running applications. This has revolutionized how software is built and delivered, making it a vital tool for developers and operations teams today.

Next

We will see how to perform container monitoring and optimization.