Skip to main content

Install Data-Plane to Kubernetes Namespace

This document will guide you through the process of installing and connecting a new Data-Plane to the IOMETE Control-Plane.

Namespace names

In this document, we use data-plane-ns and iomete-system as the namespace names for the Data-Plane and Control-Plane instances. You can replace it with your desired namespace name.
Note that the Control-Plane namespace should be the same as the one you used during the Control-Plane installation.

Create new namespace for Data-Plane instance​

# Change `new-data-plane` to your desired namespace name
kubectl create namespace data-plane-ns

# Label the namespace for IOMETE
kubectl label namespace data-plane-ns iomete.com/managed=true

Create Service Account and Role​

Required file: service-account.yaml

wget https://raw.githubusercontent.com/iomete/iomete-deployment/main/service-account.yaml

kubectl apply -n data-plane-ns -f service-account.yaml

wget https://raw.githubusercontent.com/iomete/iomete-deployment/main/role-binding-to-control-plane.yaml

export CP_NAMESPACE=iomete-system

sed -i "s/{{control-plane-namespace}}/$CP_NAMESPACE/g" role-binding-to-control-plane.yaml
#For macOS use the following command
# sed -i '' "s/{{control-plane-namespace}}/$CP_NAMESPACE/g" role-binding-to-control-plane.yaml

kubectl apply -n data-plane-ns -f role-binding-to-control-plane.yaml

Generating webhook certificates​

Required file: gencerts.sh

wget https://raw.githubusercontent.com/iomete/iomete-deployment/main/gencerts.sh
chmod +x gencerts.sh

./gencerts.sh -n data-plane-ns -s spark-operator-webhook -r spark-operator-webhook-certs

# `spark-operator-webhook.yaml` file will be generated by the script above
kubectl apply -n data-plane-ns -f spark-operator-webhook.yaml
Require Kubernetes Admin Access

The generated file spark-operator-webhook.yaml contains cluster-level objects. Admin permission required to apply the last command above.

Update IOMETE Control Plane​

In values.yaml file of the IOMETE Control Plane, add the following configuration to the namespaces section:

# Multi-Namespace Support: Spark resources can now be deployed to separate namespaces,
# allowing teams to manage their own CPU and memory resources independently.
# The data plane's namespace is automatically managed and doesn't need to be specified.
namespaces:
  - data-plane-ns
# helm repo update iomete
helm upgrade --install -n iomete-system data-plane iomete/iomete-data-plane-enterprise -f values.yaml

ON THIS PAGE