In this exercise, you will install Korifi locally on kind using a locally deployed container registry. To install Korifi on another Kubernetes provider, see the installation documentation in GitHub.

Prerequisites

You will need the following prerequisites to use Korifi locally. If you are a Kubernetes user, you likely already have most of these tools installed.

Follow the instructions below to install the prerequisites.

• Docker Desktop: Kind uses Docker to run Kubernetes container nodes. Install Docker Desktop according to the instructions.

• kind: kind is a tool for running local Kubernetes clusters using Docker container “nodes”. Install kind by following the directions. You only need to install kind; you do not need to create a cluster yet.

• Helm: Helm is a package manager for Kubernetes. Helm is used to install Korifi on a Kubernetes cluster. Install Helm according to the instructions.

• kubectl: kubectl is the the Kubernetes command-line tool, allowing you to run commands against Kubernetes clusters. Install kubectl according to the instructions.

• cf: cf is the Cloud Foundry command-line tool. Install the latest version of the Cloud Foundry CLI according to the instructions.

• kbld: kbld builds and copies the required docker images to the local registry. Install the latest version of kbld according to the instructions.

Installation Overview

The Korifi development team maintains an installation script to install Korifi locally. It installs the Kubernetes ecosystem dependencies outlined below and a local container registry. We will use this script to install Korifi. If you prefer, you can follow the installation instructions on GitHub to install Korifi manually.

The install script does the following before installing Korifi:

• Creates a kind cluster with the correct port mappings for Korifi and other components.
• Deploys a local Docker registry using the twuni helm chart.
• Creates an admin user for Cloud Foundry.
• Installs cert-manager. cert-manager is used to create and manage internal certificates within the cluster.
• Installs kpack. kpack is used to build runnable applications from source code using Cloud Native Buildpacks.
• Installs contour. Contour is the ingress controller for Korifi.
• Installs the service binding runtime which is an implementation of the service binding spec.
• Install the metrics server

Installing Korifi

While you can install all of the above manually, it is far easier to use the deploy-on-kind.sh script in the Korifi repository.

Please note the Korifi development team uses this script for local development. It is not intended for production use. At times the script may change or break.

To install Korifi:

• Clone the Korifi repository locally: git clone https://github.com/cloudfoundry/korifi

• Change to the scripts directory: cd korifi/scripts

• Install Korifi using the deploy-on-kind.sh script: ./deploy-on-kind.sh korifi

  Note, the name ‘korifi’ above is the name of the cluster created in kind. You can use any name you would like.

Deploying an Application

We have provided a sample application you can deploy to your Korifi instance. However, before deploying, we need to set up our Cloud Foundry instance a bit by doing the following:

• Target the Cloud Foundry instance: cf api localhost --skip-ssl-validation
• Authenticate: cf auth cf-admin
• Create an Org: cf create-org tutorial
• Create a Space: cf create-space -o tutorial dev
• Target the Org and Space you created: cf target -o tutorial -s dev

Now you are ready to deploy the application:

• Clone the sample application: git clone https://github.com/cloudfoundry-tutorials/korifi-sample-app.git
• Change the application directory: cd korifi-sample-app
• Deploy the application: cf push

At a high level, the following happens:

• The cf command line interface uploads the application bits
• Korifi uses kpack and Paketo buildpacks to build a container image for the application
• The container image is stored in the local Docker registry
• The container image is deployed to Kubernetes
• Ingress traffic will be mapped to the application using Contour

You will see an output that ends with something similar to the following. If you copy the route, you should be able to open the application in a browser.

requested state:   started
routes:            sample-app.apps-127-0-0-1.nip.io
last uploaded:     Thu 09 Feb 11:02:52 MST 2023
stack:             io.buildpacks.stacks.bionic
buildpacks:        

type:            web
sidecars:        
instances:       1/1
memory usage:    32M
start command:   sh "/workspace/start.sh"
     state     since                  cpu    memory   disk     logging      details
#0   running   2023-02-09T18:03:43Z   0.0%   0 of 0   0 of 0   0/s of 0/s   

In the next section, we will look at Korifi and the deployed application using kubectl.