You Need to Learn Docker

Welcome back to another insight!

I appreciate you being here.

DevOps has crept its way into many organizations worldwide. It's a cultural structure of collaboration and shared responsibility between development and operations teams. Through the DevOps approach, teams can achieve faster, more efficient development cycles and deployments. Deploying fast and developing fast makes for easier incremental changes to the software that the company creates. One tool changed the game. Docker.

In the evolving landscape of DevOps, Docker has risen to prominence and become nearly indispensable. Docker is more than just a buzzword in today's engineering community; it has fundamentally changed the way engineers develop, deploy, and scale applications.

Gone are the days of “It works on my machine!”. Docker allows developers to 'build once and run anywhere,' which makes it a powerful tool in an industry that increasingly prioritizes flexibility, scalability, and efficiency. As we continue to move towards distributed systems and microservices, Docker's importance in engineering, and specifically DevOps, cannot be overstated.

Enough said about how it’s revolutionizing the industry and let’s talk specifics of why.

‍Understanding Docker: A Shift from Traditional Virtualization

Docker is an open-source platform that automates the deployment, scaling, and management of applications. At its core, Docker utilizes a technology known as containerization. But what makes it stand apart, and how does it contrast with traditional virtualization?

Traditional virtualization relies on the process of creating multiple virtual machines (VMs) on a single physical host. Each VM includes a full copy of an operating system (OS), the application, necessary binaries, and libraries - taking up tens of GBs. VMs are isolated and secure, but they can be heavy on resources. Most of the time, this means maintaining Ansible configuration to handle all the necessary setup.

In contrast, Docker introduces a lightweight alternative: containers. Containers, unlike VMs, share the host system's OS kernel, making them incredibly lightweight and fast to start. A Docker container holds the application and its dependencies but leverages the underlying OS, allowing it to be far more efficient in terms of system resources. Docker packages the application and its dependencies in a virtual container that can run on any Linux, Windows, or Mac system. This simplifies deployment, especially in a complex, distributed environment.

The implications of this approach for DevOps are enormous. Docker ensures consistency across multiple development, staging, and production environments. By using Docker, developers can construct a complex application that includes various microservices, without worrying about conflicting dependencies or inconsistent environments. In essence, Docker containers drive efficiency, provide application isolation, and reduce conflicts in multi-tenant environments. This is crucial for DevOps success. Engineers trying to build pipelines against ever-changing system dependencies will end up crying themselves to sleep most nights.

‍

Exploring Images, Containers, and Dockerfiles

‍

Docker operates on a few key concepts that form the cornerstone of its functionality: images, containers, and Dockerfiles. Understanding these concepts is fundamental to using the platform.

Images are the building blocks of Docker. You can think of an image as a lightweight, stand-alone, executable software package that includes everything needed to run a piece of software. The image encompasses the code, a runtime, libraries, environment variables, and config files. Images are immutable. If you make any change to the Dockerfile (read below), you will get a new image with a new manifest and digest.

The Dockerfile is a text document that contains all the commands a user could call on the command line to assemble an image. By executing a docker build command, we can build the image from the Dockerfile. In short, Dockerfiles are the blueprint for your image.

Lastly, containers are the running instances of Docker images. They are lightweight and portable encapsulations of an environment in which to run applications. A single image can be used to create multiple containers, each of which can be scaled, removed, started, or stopped independently. Containers maintain the application's isolation from the rest of the system, ensuring it runs the same regardless of the environment.

As a DevOps team lead, in interview, I commonly ask what the difference is between the image, Dockerfile, and container. Here’s how I would answer:

“The docker image is the distributable version of the application. It’s the binary that gets published. It’s similar to how we have .exe files for windows, .jar/.war files for Java apps, or other types of binaries for other software languages. Dockerfiles are the blueprint for the image. Simple as that. Containers are the living version of the image. It’s the image at runtime.”

Together, these Docker concepts form a powerful framework for developing, deploying, and scaling applications efficiently and consistently.

‍

Cloudy with a Chance of Docker: Running Containers Across Various Cloud Environments

The beauty of Docker lies in its universality and flexibility. Docker containers can run just about anywhere - from your personal laptop to a corporate network, and, most interestingly, across various cloud environments. This flexibility has made Docker a must-have tool in the world of cloud computing.

Let's consider Amazon Web Services (AWS), a staple amongst developers for its extensive suite of cloud services. Docker meshes beautifully with AWS, and you can run Docker containers on several AWS services like EC2 instances, ECS, and AWS Fargate. EC2 instances offer you complete control and are a good option when you need to run your Dockerized applications on infrastructure that you manage. Fargate, on the other hand, allows you to run containers without managing the underlying infrastructure, giving you more time to focus on designing and building your applications.

Also, serverless computing. In AWS, you can use services like AWS Lambda, which allow you to run your Docker containers in a serverless environment, meaning you don't have to worry about provisioning or managing servers.

Beyond AWS, Docker plays nicely with other cloud platforms like Google Cloud Platform (GCP) and Microsoft Azure, each offering its own services for running Docker containers. I’ve found Azure’s products to be quite easy to use in comparison to AWS and GCP.

In a nutshell, Docker's ability to run seamlessly across these diverse cloud environments. It not only increases deployment options for your applications, but also makes it easier to move applications across environments. It truly embodies the promise of "build once, run anywhere."

Docker will be a key tool that I write about in the future. It’s a crucial tool in the toolbelt of any DevOps engineer, and developers in general. I personally deploy nearly all of the applications using Docker and find it easy to use after gaining some experience. It makes the setup process significantly easier for any environment.

If you know someone who needs to see this content, please share the newsletter to them!

‍