Node-level autoscaling, load balancing and storage classes come automatically
Regions: tr-ist-01 (İstanbul) · tr-ank-01 (Ankara).
The managed control plane holds cluster state; the scheduler looks for a place for every pod.
Pods are placed on worker nodes by their resource requests; web and api pods can share a node.
A pod that finds no room waits; the autoscaler turns that pending demand into a capacity decision.
A new worker node joins the cluster and the pending pod lands there; when load drops the loop runs in reverse.
Answer
VKE is vMind’s open-source Kubernetes engine. VKE runs the cluster infrastructure; you run the application.
Cost depends on worker node count and profile, persistent storage and load balancing, traffic, backup and runtime. Review the current estimate in the cost calculator.Calculate Cost
Live capacity and quota depend on the selected resource; service level terms are explained on the SLA page.
Features
How it works
In a self-managed Kubernetes model, your team installs the control plane and handles version upgrades, high availability and core cluster operations. With VKE, the control plane is part of the managed service. You select the Kubernetes version when you create the cluster. Upgrade options are in the console and in the VKE documentation.
VKE operates the cluster resource lifecycle. Application images, secrets, access and network policy, data protection and workload observability stay in your layer.
Use a public Load Balancer when the service must be reachable from the internet, and an internal Load Balancer when it should stay on the project network. A persistent volume attaches to the pod through a storage class. For stateful workloads, plan protection with Block Storage and Backup.
You set the worker profile and the scale limit during setup. After that, the node count rises and falls with the workload. H100, H200 and L40S servers are on Compute. Managed Kubernetes on its own does not mean a GPU node. Charges follow the control plane, the worker profile, node count, persistent storage, load balancing, public IPs, data transfer, backup and runtime. Current prices are in the cost calculator.
You recreate the same cluster definition with Terraform or the API. The definition covers the version, key pair, master and worker profiles, networking and API access. Project access is opened with an Application Credential.
Scenarios
Decision criteria
Hover or press play; the steps light up in time with the recording.
FAQ
VKE provides managed Kubernetes: node-level autoscaling, Load Balancer and storage classes come automatically. Applications, images, secrets, network policies, data protection and observability stay in your layer. The responsibility split is set out in the product documentation and service terms.
VKE offers selectable Kubernetes versions on the cluster creation screen. Version and upgrade options are in the console and the VKE documentation.
VKE automatically provides load balancers for Kubernetes services and storage classes for persistent data. Protect stateful workloads with Block Storage and Backup.
GPU servers are offered on Compute, with H100, H200 and L40S options. VKE provides Kubernetes clusters with node-level autoscaling, Load Balancer and storage classes.
Cost is set by the control-plane model, worker-node count and CPU/RAM profiles, persistent storage, Load Balancer, public IPs, data transfer, backup and runtime. The line items are in calculator.portvmind.com and the console resource creation screen.
Next step
Create the resource in the console. For a migration or scaling plan, continue together in an architecture consultation.
Model your scenario in the cost calculator.Calculate Cost
