This document covers about how to set up Compage on your local with helm charts and Compage will be exposed through NodePort service to the outer world. When you deploy on the server, you have to expose the ui service using LoadBalancer service type or need to create an ingress.

Create a KinD cluster

Make sure you have access to Kubernetes cluster. You can create a KinD cluster as explained here, in a cluster creation section. Make sure that you are installing the nginx ingress controller as well.

We have configured the KinD cluster to be available at these ports:

• ui-32222
• app—31111

So, you can access the compage using http://localhost:32222

Create a namespace

Currently, the compage namespace is made hard-coded but in-future, it will be completely configurable.

kubectl create ns compage
kubectl config set-context --current --namespace=compage

Configure Cassandra

You need to get the details for Cassandra as compage/app uses Cassandra as a database. You need to update the values.yaml file with the Cassandra details.

app:  
  cassandra:
    contactPoints: "test-dc1-service.k8ssandra-operator.svc.cluster.local"
    localDataCenter: "dc1"
    username: "test-superuser"
    password: "Vr58zqslXCgQniIfHdYe"
    keyspace: "compage"

You can use any Cassandra cluster to store the data. If you want to start a local Cassandra cluster, you can follow the steps given in the compage/deploy/steps.md file.

Install the latest version from GitHub helm repository.

Fire below set of commands and install the compage on your KinD cluster running locally.

Before this, you will have to create a docker image for ui component. As this is a UI component and commands in Dockerfile use below CONFIG values

• REACT_APP_COMPAGE_APP_SERVER_URL

to create it, you will have to use your configurations and create a docker image using below commands (run them from base folder of compage)

TAG_NAME="{version you are installing so that it will be automatically taken.}"
UI_IMAGE="ghcr.io/intelops/compage/ui:$TAG_NAME"
# Assuming this is the name of your kind cluster
CLUSTER_NAME=compage
# create docker image for ui
docker build -t $UI_IMAGE  --network host ui/
kind load docker-image --name $CLUSTER_NAME $UI_IMAGE

Once you are done with above commands, kindly run below set of commands.

helm repo remove intelops
helm repo add "intelops" "https://raw.githubusercontent.com/intelops/compage/main/charts"
helm install compage intelops/compage --values charts/compage/values.yaml
kubectl get pods -n compage
kubectl wait --for=condition=ready pod -l app.kubernetes.io/name=compage-ui
kubectl wait --for=condition=ready pod -l app.kubernetes.io/name=compage-core
kubectl wait --for=condition=ready pod -l app.kubernetes.io/name=compage-app

Go to http://localhost:32222

Uninstall

Simply, delete the cluster created above using kind delete cluster --name compage. If that’s not possible, delete the namespace kubectl delete ns compage