Skip to main content

On-Premises Deployment Guide

This guide provides detailed instructions for deploying IOMETE on-premises within a Kubernetes environment, ensuring you have a seamless setup process.

Essential Requirements Before You Start​

Before initiating the deployment, ensure your system meets the following prerequisites:

  • Kubernetes Cluster: Your cluster should include at least one data node with the following specifications:
    • Minimum Specs for Data Node: 4 CPU cores and 32GB of RAM.
    • Resource Allocation:
      • IOMETE System Node: Requires 8 CPU cores and 16GB of RAM.
      • Spark Driver: Utilizes the remaining CPU cores and memory.b v
  • Node Configuration Tips:
    • Opt for uniform node sizes to simplify management.
    • Larger nodes or VMs provide improved resource efficiency.
  • Object Storage: Have one of the following ready: Minio, DELL ECS, IBM Cloud Object Storage, AWS S3, Azure Blob Storage, or Google Cloud Storage.
  • Necessary Tools:
    • kubectl for cluster interaction.
    • helm for package management.
    • aws cli for object storage connectivity.

Deployment Steps​

Ensure you're targeting the right Kubernetes cluster with kubectl and have the necessary repository cloned.

Create namespace for IOMETE​

A dedicated namespace for IOMETE is recommended for better organization. Create it using the following command:

kubectl create namespace iomete-system

# Label the namespace for IOMETE
kubectl label namespace iomete-system iomete.com/managed=true
tip

Technically, you can deploy IOMETE in any namespace. If you choose to deploy in a different namespace, ensure you use the correct namespace in the following steps.

Object Storage (MinIO)​

If you need an object storage system, consider deploying MinIO, object storage solution. Follow the instructions here.

Deploying Metadata Database (PostgreSQL)​

For metadata storage, you need a PostgreSQL database. Please follow the instructions here.

Add IOMETE Helm Repository​

Add the IOMETE helm repository for access to necessary charts:

helm repo add iomete https://chartmuseum.iomete.com
helm repo update

Install CRDs​

Required file: iomete-crds.yaml

Require Kubernetes Admin Access

Admin permission required to install the command below as it installs cluster-level objects.

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

kubectl apply -f iomete-crds.yaml

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 iomete-system -f service-account.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 iomete-system -s spark-operator-webhook -r spark-operator-webhook-certs

# `spark-operator-webhook.yaml` file will be generated by the script above
kubectl apply -n iomete-system -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.

Launching IOMETE Data Plane​

See the IOMETE Data Plane Enterprise Helm for more details about available configurations. See here for the values.yaml

Required file: example-data-plane-values.yaml

tip

example-data-plane-values.yaml is a sample configuration file. You can customize it according to your requirements.

wget https://raw.githubusercontent.com/iomete/iomete-deployment/main/on-prem/example-data-plane-values.yaml

# helm repo update iomete
helm upgrade --install -n iomete-system data-plane \
  iomete/iomete-data-plane-enterprise \
  -f example-data-plane-values.yaml

Configure ISTIO Ingress Gateway​

Please follow the Configure ISTIO Ingress Gateway to configure the Ingress Gateway for IOMETE Data Plane to be able to access the UI.

ON THIS PAGE