עבריתKubernetes has become the de-facto standard for container orchestration, making it easier to manage, scale, and deploy containerized applications. However, as Kubernetes clusters grow in complexity, managing deployments using YAML manifests alone can become cumbersome. Helm, Kubernetes' package manager, streamlines application deployment, management, and rollback by allowing developers to bundle Kubernetes manifests into versioned packages called charts. Custom Resources (CRs), on the other hand, extend Kubernetes' API, providing a way to define and manage new types of objects.
Kubernetes Deployment with Helm Charts and Custom Resources
Kubernetes has become the de-facto standard for container orchestration, making it easier to manage, scale, and deploy containerized applications. However, as Kubernetes clusters grow in complexity, managing deployments using YAML manifests alone can become cumbersome. Helm, Kubernetes' package manager, streamlines application deployment, management, and rollback by allowing developers to bundle Kubernetes manifests into versioned packages called charts. Custom Resources (CRs), on the other hand, extend Kubernetes' API, providing a way to define and manage new types of objects.
This article will explore how to deploy applications on Kubernetes using Helm charts and custom resources, highlighting the benefits of both methods. We'll cover the following topics:
- Introduction to Helm and Custom Resources
- Setting up a Kubernetes cluster
- Helm basics: Installing Helm and creating charts
- Managing Kubernetes applications with Helm charts
- Creating and managing custom resources in Kubernetes
- Combining Helm charts and custom resources for flexible deployments
What is Helm?
Helm is a Kubernetes package manager that simplifies the process of defining, installing, and upgrading Kubernetes applications. It uses pre-configured templates (known as charts) to deploy applications and manage their lifecycle, allowing developers to:
- Package Kubernetes resources into a single unit (a Helm chart).
- Install and upgrade applications easily.
- Roll back changes with ease.
- Reuse charts for different environments by modifying values.
Helm uses three main concepts:
- Charts: Collections of files that describe Kubernetes resources.
- Release: An instance of a Helm chart running in a Kubernetes cluster.
- Repositories: Locations where charts are stored and shared.
What are Kubernetes Custom Resources?
Kubernetes Custom Resources (CRs) are user-defined types of objects that extend Kubernetes' functionality. Custom Resources enable developers to define custom objects and use them as Kubernetes-native resources. For example, while Kubernetes provides built-in objects like Pods, Deployments, and Services, a Custom Resource can define anything from database configurations to monitoring rules.
Custom Resources are paired with Custom Resource Definitions (CRDs), which serve as the schema for the CR. By using CRDs, developers can create new Kubernetes API objects that work just like built-in ones. Custom Resources can be coupled with controllers that observe these resources and take actions, automating complex workflows.
Setting Up a Kubernetes Cluster
We need a Kubernetes environment before deploying applications using Helm and custom resources. The following steps show how to locally set up a Kubernetes cluster using Minikube.
Install Minikube
Minikube is a lightweight tool for running Kubernetes clusters locally. Install Minikube based on your operating system:
curl -LO https://storage.googleapis.com/minikube/releases/latest/minikube-linux-amd64
sudo install minikube-linux-amd64 /usr/local/bin/minikube
Start Minikube
Start Minikube and create a local Kubernetes cluster:
minikube start
Verify Kubernetes Installation
Check the status of the cluster using kubectl, the Kubernetes command-line tool:
kubectl cluster-info
kubectl get nodes
Install Helm
Helm requires a running Kubernetes cluster, so install Helm on your local machine and configure it to connect to your cluster:
curl https://raw.githubusercontent.com/helm/helm/master/scripts/get-helm-3 | bash
Helm Basics Installing Helm and Creating Charts
With the Kubernetes cluster up and running, we can dive into Helm's functionality by creating and deploying charts.
Helm Chart Structure
Helm charts follow a specific structure that organizes all Kubernetes manifest files in a reusable format.
Creating a Helm Chart
Run the following command to create a new Helm chart:
helm create my chart
This command generates the directory structure described above. Modify the templates/ directory to include your Kubernetes resource definitions.
Deploying an Application with Helm
To deploy an application using Helm, follow these steps:
-
Package the chart: helm package my chart
This command deploys the Helm chart and creates a new release (myrelease).
Customizing Deployments with Values
The values.yaml file allows for flexible configuration of your Helm chart. For example, modify resource limits, environment variables, or image versions by simply changing values in this file.
You can also override values.yaml at runtime by passing values through the command line:
helm install my release ./mychart set image.tag=2.0
Managing Kubernetes Applications with Helm Charts
Helm simplifies complex Kubernetes applications by templating and reusing them. Below, we explore some advanced Helm use cases.
Updating Applications
To update an application’s configuration, modify the chart or its values and use the following command:
helm upgrade my release ./mychart
Helm upgrades the application while keeping track of the new release.
Rolling Back Changes
If an upgrade causes issues, Helm can roll back to a previous version: helm rollback release 1
Using Helm Repositories
You can share Helm charts by pushing them to Helm repositories. To use a chart from a public repository:
helm repo add stable https://charts.helm.sh/stable
helm install stable/MySQL
Creating and Managing Custom Resources in Kubernetes
While Helm is excellent for managing traditional Kubernetes resources, custom resources provide flexibility for complex use cases. Here’s how to define, create, and manage custom resources in Kubernetes.
Creating Custom Resources
Once the CRD is in place, you can create instances of the custom resource:
apiVersion: example.com/v1
kind: Database
metadata:
name: my-database
spec:
engine: Postgres
version: 12.3
Managing Custom Resources
Custom resources behave like any other Kubernetes object. You can list, describe, and delete them using standard kubectl commands:
kubectl get databases
kubectl describes database my-database
kubectl delete database my-database
Combining Helm Charts and Custom Resources for Flexible Deployments
Helm charts and custom resources can be combined to create highly flexible and reusable deployments. A common use case is integrating CRDs into Helm charts for better application extensibility.
Creating Helm Templates for Custom Resources
You can also include custom resource definitions inside Helm chart templates. This allows you to use values and conditionally create custom resources based on configuration options:
apiVersion: Values. card.apiVersion
kind: Database
metadata:
name: Release.Name
spec:
engine: Values.database.engine
version: Values.database.version
Deploying with Helm
Finally, deploy the chart that includes both traditional Kubernetes resources and custom resources:
helm install myrelease ./mychartHelm manages the entire application stack, including the custom resource definitions and their instances.Helm charts and custom resources provide powerful tools for managing Kubernetes applications. Helm simplifies the process of packaging, deploying, and upgrading applications, while custom resources extend Kubernetes' capabilities, allowing you to define your APIs. By combining both tools, developers can create highly flexible, reusable.
