Introduction

Software engineering has come a long way since its inception and we have used a lot of techniques to deploy our software for users over the last few decades. The latest buzzword that we often hear nowadays for deployment is Kubernetes. There are tons of articles on the internet that says Kubernetes is an orchestrating software that manages the deployment of different applications as containers. But is the definition enough for any beginner who wants to learn Kubernetes? This article not just helps you understand what Kubernetes does, but it also tells you why it is the best deployment tool developers have created to date.

The Containers

Over the years, we have been deploying our applications in machines made of bare metals or virtual machines as artifacts directly. The problem with this form of deployment is that all the necessary software runtimes and dependent libraries have to be installed before our app to make it work, and guess what this was completely error-prone and took a lot of time to get the setup up & running. To eradicate this problem, companies came up with a solution that would install the app along with its dependencies. They call these packages containers and these containers can be deployed in any machine, anywhere, and at any time.

Why Kubernetes?

All the digital apps that we use in our daily lives are a type of distributed system and these systems are expected to have a constant up time by their customers. Traditional deployment strategies like bringing the application down, updating the artifacts, testing the newly deployed artifact, and finally getting them back in production are no longer accepted, hence we need a unique style of deployment that would deploy the latest versions while the apps are up and running.

There are several ways through which Kubernetes can solve this problem and one way that immediately comes to our mind is to roll the changes gradually without any downtime and then make it available to all the users once the partial rollout is successful.

The Benefits:

There are numerous reasons why people love containers and container APIs like Kubernetes, but they all can be traced back to one of the benefits below.

• Velocity
• Immutability
• Declarative Configuration

Velocity

In software development, velocity might be defined as the number of features shipped during an hour or a day, but it is not just that. It is more than that. The ability to upgrade the services while still making the service available is the new definition of velocity. In this way, both containers and Kubernetes help us achieve these highly maintainable services without downtime.

Scaling

We all know that distributed systems are bound to be scaled, either up or down. By using Kubernetes, it is very easy to scale up or scale down a service. Most importantly, this can be done either manually or automatically. For Kubernetes, it is just creating new pods to destroying old ones depending on the traffic for a particular service.

Declarative Configurations

Here comes the best and my favorite feature of Kubernetes too. Kubernetes is built for developers, and developers love to code. Gone are those days when the deployment was down through an imperative approach i.e by running a series of commands to deploy a piece of software. Kubernetes follows a declarative style of deployment where developers present a YAML configuration having the desired state and Kubernetes set up a cluster that matches the desired state. Kubernetes not just helps in setting up the desired state, but also continuously takes action to match the current state with the desired state.

Conclusion

Given that containers have become the de facto standard for deployment, Kubernetes has become a real game changer. The amount of burden Kubernetes has taken off of developers is immeasurable and will continue to gain popularity over the coming years. In the upcoming series of articles, we will further discuss the architecture of Kubernetes and will also help you decide which type of Kubernetes solution will perfectly fit your organization.

Until then, Sayanora!