Docker vs Kubernetes – What's the Difference and Why It Matters

Key Takeaways

One of the most popular tools to containerize applications and run containers is docker. So here we have our application code. Then we have the requirements.txt file which has all the dependencies that are required for the application. In our case, it's just the flask dependency. And we now build a docker file to package the application with its dependencies into a docker container. So the first line here creates an image from the python base image and then sets the right working directory. then copies the requirements.txt file to the working directory and then installs the dependencies and this is where you could add any other dependency to it and then copies the application code into the image and finally defines the command to run the application using the cmd instruction. Now by running the docker build command we build an image and by running the docker run command we run one instance of our application. So that was a quick introduction to containers and docker. If you are new to Docker, I would recommend checking out our free Docker for beginners course on CodeCloud using the link given here. You'll learn with hands-on labs using our interactive learning environment by working on real systems that's exclusive to you. So what is Kubernetes? With Docker, you are able to run a single instance of an application using the Docker run command, which is great. Running an application has never been so easy before. With Kubernetes, using the Kubernetes CLI known as cubectl, you can run a thousand instance of the same application with a single command. Kubernetes can scale it up to 2,000 with another command. Kubernetes can even be configured to do these automatically so that instances and the infrastructure itself can scale up and down based on user load. Kubernetes can upgrade these 2,000 instances of application in a rolling fashion, one at a time, with a single command. And if something goes wrong, it can help you roll back these images with a single command. Kubernetes can help you test new features of your application by only upgrading a percentage of these instances through AB testing methods. Now, don't worry about the command line tool for now. We will take a closer look at it very soon. With Kubernetes, you're also able to define the expected state of your application. For example, you're able to define that your application consists of four different services. The web server must have three instances running. The payment service to have two. There should be a ready service with three instances running and a database service to which these services connect to. And you're able to define these in code. Kubernetes will ensure that the state that you have defined in these files for your application is maintained at all times.