Start / Stop KubeRaya Cluster
This page explains how to manage the lifecycle state of a KubeRaya cluster in CloudRaya, including starting and stopping the cluster, and how each state affects availability and billing.
Cluster state changes impact infrastructure-level components, including control plane and worker nodes.
Application workloads and Kubernetes configurations are preserved unless the cluster is deleted.
Start and Stop a KubeRaya Cluster
You can start or stop a KubeRaya cluster using two available methods.
Option 1: From the KubeRaya Cluster List
On the KubeRaya page, each cluster card provides quick actions for lifecycle management.
From the cluster list, you can:
- Start Cluster – Bring a stopped cluster back online
- Stop Cluster – Gracefully stop a running cluster
These actions allow quick operational control without opening the cluster detail page.
Option 2: From the KubeRaya Detail Page
You can also manage the cluster state from the cluster detail view.
To change the cluster state:
- Open KubeRaya from the CloudRaya dashboard
- Select the target cluster
- On the KubeRaya Detail page, use the lifecycle action buttons:
- Start – Bring a stopped cluster back online
- Stop – Gracefully stop all cluster nodes
These actions control the operational state of the Kubernetes control plane and worker nodes.
Cluster State and Billing Behavior
Billing for a KubeRaya cluster depends on the current cluster state.
Unlike Virtual Machines, current charges are not displayed directly on the KubeRaya detail page.
Billing When the Cluster Is Running
When a KubeRaya cluster is in the Running state, billing is applied per node, per hour, based on the selected cluster package.
Each node’s hourly package price already includes:
- CPU resources
- Memory allocation
- Root disk storage
Both control plane and worker nodes are billed while active.
No separate charges are applied for compute or root disk storage while the cluster is running, as these are bundled within the node package.
Billing When the Cluster Is Stopped
When a KubeRaya cluster is in the Stopped state:
- Compute resources (CPU and memory) are not billed
- Root disk storage for each node continues to be billed
- Cluster configuration and Kubernetes data remain intact
Storage charges are calculated based on disk size and node count.
Storage Billing Calculation (Stopped Cluster)
- Storage rate: IDR 4.62 per GB per hour
- Each node retains its root disk
- A KubeRaya cluster has a minimum of 2 nodes
Example:
- Root disk size per node: 50 GB
- Storage cost per node:
50 × 4.62 = IDR 231/hour - Total storage cost for a stopped cluster (2 nodes):
231 × 2 = IDR 462/hour
Billing When the Cluster Is Started Again
When a stopped cluster is started:
- The cluster resumes using the same configuration
- All node data and Kubernetes state are preserved
- Compute billing resumes automatically
- Storage billing continues without interruption
No redeployment or reconfiguration is required unless you choose to modify the cluster.
Summary
- KubeRaya clusters support start and stop actions from both the cluster list and detail page
- Cluster state changes affect availability and infrastructure billing
- Stopping a cluster removes compute charges while retaining storage charges
- Starting the cluster restores full operation using existing data