Helm to the rescue!

Hey all! In this blog, all I am going to do is talk about the basic concepts and need for Helm.

What is Helm?

According to the Dictionary,

a tiller or wheel for steering a ship or boat

In a similar fashion technically,

Helm is basically package manager that helps you manage Kubernetes applications.It uses a packaging format called charts which is a collection of files that describe a set of Kubernetes resources. A chart can be used to deploy a single resource or multiple resources.

Just like in the World of Linux, Package managers like APT or YUM have made life easier similarly in the World of Kubernetes Helm has made things a lot easier than complex!!

The need of Helm:

I assume some of you might be familiar with helm but those who are not, to deploy a Kubernetes pod and a service in your cluster you might have written a deployment.yaml and a service.yaml files and might deploy them using the below command.

kubectl apply -f <your-fileName>

In this case, you had two files so you executed the above command twice but,

What if you need to deploy 20 Kubernetes resources or may be more than that then you have to execute the above command 20 times for each resource file!!

However, if you have a helm chart containing all these 20 resource files, you only need to execute a single command to deploy all those files! Easy enough?

Okay! So Helm saves me a few command line calls… so what?

Helm offers more than that,

• It makes dependency management easier.
• It helps in enforcing standards.
• Updating an application is much simpler.
• Keeps track of how many times an application has been deployed and also have made rollback much simpler compared to doing with kubectl.

Structure of Helm:

Helm is made of two components: A server called Tiller, which runs inside the Kubernetes cluster and a client called helm i.e command-line client with which we can browse package repositories (containing published Charts) and deploy those charts in the Kubernetes cluster.

Playing with Helm Chart:

Before we start this, first of all, you need to have Helm installed if not then you may follow this: https://helm.sh/docs/intro/install/

A chart is organized as a collection of files inside a directory, to create a chart execute the following command:

helm create <your-ChartName>

After this, you need to go inside the Chart directory you just created and you will see a structure something like this:

├── charts
├── Chart.yaml
├── templates
│   ├── deployment.yaml
│   ├── _helpers.tpl
│   ├── ingress.yaml
│   ├── NOTES.txt
│   ├── service.yaml
│   └── tests
│       └── test-connection.yaml
└── values.yaml

1. Chart.yaml contains metadata about the Chart like its name, version, keywords.
2. Inside the templates folder, you can see files like deployment.yaml, service.yaml and ingress.yaml these will be deployed in your Kubernetes cluster if you install this helm chart in the cluster.
3. The most important file in this directory is the values.yaml file, because it stores variables for the template files inside the templates directory. If you look at the complete deployment.yaml file, you can clearly understand how the variables inside values.yaml file is referred to in the helm template files.

Similarly, if you need to deploy more resources to your cluster using helm, just create and insert those YAML files inside the template folder.

Installing a Helm chart:

Now, the final step once you are done with the creation of the helm chart, there is only just one command to install it in the cluster which will deploy all the YAML files you created in the template folder i.e.

helm install --name-of-your-deployment .

And yes we are done! It wasn't that tough, isn't it?

So, I would now like to bind up this. I hope you got the basic idea of working with helm. To learn in more depth, you can refer to https://helm.sh/

Thanks all and yes keep learning while enjoying quarantine! Stay safe!