⎈
Kubernetes Clusters
CloudPe Knowledge Base · Compute
Last updated: April 2026
Kubernetes Clusters on CloudPe
Deploy and manage fully managed Kubernetes clusters with automated control-plane provisioning,
scalable worker nodes, and seamless integration with CloudPe networking and storage.
This guide walks you through every step of creating a Kubernetes cluster.
🕐 ~15 min setup
⚙️ Managed Control Plane
📈 Scalable Workers
🔒 HA Mode Available
✓
Prerequisites
Ensure these are in place before creating your cluster
☁️ Active Project
A CloudPe project must be selected. You can create one under Billing & Account → Projects.
🔑 SSH Key
An SSH key must exist in Billing & Account → SSH Keys to authenticate access to nodes.
💳 Billing Enabled
Ensure your account has sufficient credit or an active payment method for cluster node costs.
🌐 Region Selection
Know which CloudPe region to deploy in. Cluster resources cannot be migrated across regions after creation.
1
Navigate to Kubernetes
Access the Kubernetes section from the left sidebar
From the CloudPe dashboard, expand the Compute section in the left navigation sidebar and click Kubernetes. You will see the Kubernetes Clusters list page showing all existing clusters. Click the + Create Cluster button in the top-right corner to begin.
ℹ️ Empty State
If this is your first cluster, you will see an empty state with a “Create Cluster” button in the centre of the page.
2
Cluster Name & Description
Set a unique name and optional description for your cluster
The Create Kubernetes Cluster form opens as a single scrollable page. At the top, fill in the cluster identity fields.
1
Cluster Name
Enter a meaningful name such as
prod-cluster or staging-k8s. The name must be unique within your project.2
Description (Optional)
Add a short description to help identify the cluster’s purpose — e.g. “Production workload cluster for microservices”.
3
Select Project & Region
Choose the CloudPe project this cluster belongs to, then select the target region (e.g. Mumbai DC2 Zone A — IN-WEST2). The cluster’s resources are bound to this region.
💡 Naming Convention
Use lowercase letters, numbers, and hyphens only. Descriptive names like
prod-west-cluster make cluster management easier at scale.3
Select Kubernetes Version
Choose a version template for your cluster
CloudPe provides pre-configured Kubernetes version templates as selectable cards. Each template represents a specific Kubernetes version configuration optimised for different use cases.
Kub-1_template
Kubernetes version template
Test-bill_template
Kubernetes version template
kube-cluster_template
Kubernetes version template
uat-cluster_template
Kubernetes version template
ℹ️ Template Selection
Click a template card to select it. The selected card gets a purple border highlight. Choose the template that matches your target Kubernetes version and workload requirements.
4
Configure Control Plane (Masters)
Set up your cluster’s master nodes and high-availability options
The Control Plane section lets you configure the master nodes that manage your cluster’s state, scheduling, and API server. These are the brains of your Kubernetes cluster.
1
High Availability (HA)
Check High Availability (HA) to deploy 3 master nodes behind a load balancer. This is recommended for production environments to eliminate the single point of failure on the control plane. When unchecked, only 1 master node is deployed (suitable for development/testing).
2
Master Nodes
Select the number of master nodes from the dropdown. Options include 1 Master (Development) for dev/test workloads and higher counts for production with HA enabled.
3
Master Flavor
Choose the compute flavor for master nodes. The default is a.cpu1.4g (2 vCPU, 4 GB RAM). CloudPe recommends 2+ vCPUs and 4+ GB RAM for the control plane.
⚠️ Production Warning
For production clusters, always enable High Availability (HA). A single-master setup means your entire cluster becomes unreachable if the master node fails.
Master Flavor Reference
a.cpu1.4g2 vCPU · 4 GB RAM — Minimum recommended
a.cpu1.8g4 vCPU · 8 GB RAM — Recommended for medium clusters
a.cpu1.16g8 vCPU · 16 GB RAM — For large production clusters
5
Configure Worker Nodes
Define the compute capacity for running your workloads
Worker nodes are where your application pods, services, and workloads actually run. Configure the count, flavor, and storage for each worker node.
1
Worker Node Count
Set the initial number of worker nodes. You can scale this later from the cluster detail page. The displayed note says: “Initial worker node count (can be scaled later)”.
2
Worker Flavor
Select the compute size for each worker node. For general workloads, the same CPU-optimised flavors are available as for master nodes. Choose based on your expected pod density and resource requirements.
3
Docker Volume Size (GB)
Set the storage volume size per worker node for container images and runtime data. Default is 50 GB. This is the local disk allocated to Docker/container storage on each node.
💡 Scaling Later
You don’t need to provision all capacity upfront. Start with a smaller worker count and scale up from the cluster detail page as your workloads grow.
6
SSH Access
Configure SSH keys for secure node access
CloudPe injects an SSH key into all cluster nodes (master and workers) so you can SSH into them for debugging and maintenance. Select an existing key or generate a new one.
1
Select SSH Key
Choose from your existing SSH keys in the dropdown (e.g. kube (Default) — 8a:a9:8e:14:e9:4…). The key fingerprint is shown for verification.
2
Generate New Key
Click the Generate button next to the dropdown to create a new key pair. The private key will be available for download immediately after generation.
ℹ️ Advanced Options
The Advanced Options section (collapsed by default) contains additional networking and configuration parameters. For most use cases, the defaults are appropriate.
7
Review Pricing Estimate & Summary
Understand costs before creating your cluster
Before clicking Create, review the Pricing Estimate and Summary sections at the bottom of the form. These give you a full cost breakdown and configuration confirmation.
💰 Pricing Breakdown
Cluster management fee: ₹0.00/hr
Master nodes (1×): ₹0.75/hr
Worker nodes (2×): ₹1.49/hr
Total Hourly: ~₹3.79/hr
Est. Monthly: ~₹2,727/mo
Master nodes (1×): ₹0.75/hr
Worker nodes (2×): ₹1.49/hr
Total Hourly: ~₹3.79/hr
Est. Monthly: ~₹2,727/mo
📋 Configuration Summary
Kubernetes Version: Selected template
Control Plane: 1 Master — 2 vCPU, 4 GB
Worker Nodes: 2 nodes — 2 vCPU, 4 GB each
Storage per Node: 50 GB
Control Plane: 1 Master — 2 vCPU, 4 GB
Worker Nodes: 2 nodes — 2 vCPU, 4 GB each
Storage per Node: 50 GB
✅ Billing Note
Cluster management is free. You are only billed for the underlying compute resources (master and worker node VMs) at the standard hourly rate.
8
Create the Kubernetes Cluster
Submit the form to provision your cluster
Once you have reviewed all settings, click the Create Kubernetes Cluster button at the bottom of the form. CloudPe will immediately begin provisioning your cluster infrastructure.
⎈ Create Kubernetes Cluster
After clicking Create, CloudPe will:
- Provision the master node(s) and configure the Kubernetes API server
- Bootstrap etcd and the control-plane components (scheduler, controller-manager)
- Provision worker nodes and join them to the cluster
- Configure networking (CNI plugin) and storage integrations
- Make the cluster available via the cluster detail page
💡 Provisioning Time
Cluster creation typically takes 10–15 minutes. You can monitor progress on the cluster detail page. Do not navigate away immediately — bookmark the detail page URL for tracking.
✓
Post-Creation: Managing Your Cluster
What to do after your cluster is running
📥 Download kubeconfig
From the cluster detail page, download the kubeconfig file to connect
kubectl to your cluster.📈 Scale Worker Nodes
Use the cluster detail page to add or remove worker nodes dynamically based on workload demands.
🗑️ Delete Cluster
Delete the cluster from the detail page when no longer needed. All associated compute resources will be terminated.
🔐 SSH into Nodes
Use the SSH key configured at creation to directly access master or worker nodes for debugging.
✓
Frequently Asked Questions
Common questions about Kubernetes on CloudPe
Q: How long does cluster creation take?
A: Cluster provisioning typically takes 10–15 minutes. You can track progress on the cluster detail page. If it takes longer than 20 minutes, contact support.
Q: Can I upgrade the Kubernetes version after creation?
A: Kubernetes version upgrades depend on the template and platform updates. Check the cluster detail page for available upgrade options or contact CloudPe support for assistance.
Q: What is the difference between HA and non-HA mode?
A: High Availability (HA) mode deploys 3 master nodes with a load balancer, eliminating single points of failure on the control plane. This is recommended for production. Non-HA (single master) is suitable for development and testing only.
Q: Can I scale worker nodes after the cluster is created?
A: Yes. Worker nodes can be added or removed from the cluster detail page after creation without downtime to your running workloads (pods are rescheduled automatically).
Q: Is the cluster management fee always free?
A: Yes. CloudPe charges ₹0.00/hr for cluster management. You only pay for the underlying compute resources (VMs) for master and worker nodes at standard hourly rates.
Q: Can I use my own container registry with CloudPe Kubernetes?
A: Yes. Once your cluster is running and you have the kubeconfig, you can configure any container registry (Docker Hub, private registry, etc.) using standard Kubernetes imagePullSecrets.
CloudPe Knowledge Base
Kubernetes Clusters · Compute · For support visit
app.cloudpe.com/dashboard/support