Serverless Kubernetes with no nodes to manage
A Cloider Kubernetes cluster is a fully managed, serverless Kubernetes cluster. You get your own Kubernetes API, a kubeconfig and cluster admin rights, while Cloider runs the control plane and the nodes. There are no node pools to size, patch or scale, and you pay for the CPU and memory your pods request or use, plus the volumes and load balancers they claim.
- Standard Kubernetes API
- kubectl, Helm and Lens
- No charge for idle nodes
- Pause and resume
Kubernetes without the infrastructure
No nodes to manage
No node pools, instance types or node counts. You declare pods with CPU and memory requests, and Cloider schedules them on nodes it operates, patches and scales.
Billed on pods, not servers
Each running container is billed for the larger of its request and its actual use. Pending, completed and failed pods cost nothing, and spare room on a node is never billed.
Standard Kubernetes
kubectl, Helm, Kustomize, Lens, k9s and Terraform's kubernetes and helm providers all work. The metrics server is installed, so kubectl top and the HorizontalPodAutoscaler work too.
Your own isolated cluster
Each cluster has its own API server, namespaces, RBAC rules, Secrets and objects, isolated from every other cluster. Your kubeconfig is cluster admin.
Domains with automatic HTTPS
Route a domain, or a path of it, to a Service with a target, and Cloider serves it with an HTTPS certificate. No ingress controller to install.
Load balancers
A Service of type: LoadBalancer gets a public IP for its TCP and UDP ports, billed per load balancer for the time it is active.
Persistent volumes
Standard PersistentVolumeClaims for stateful workloads, billed per GiB of provisioned capacity.
Pause and resume
Pause a cluster you don't need and compute stops while its objects and volumes are kept. Resume it from the dashboard or with cloider cluster resume.
Browser terminal
A terminal with kubectl and helm already pointed at the cluster, in the dashboard or with cloider cluster shell.
From zero to running pods in minutes
Create a cluster
In the dashboard or with
cloider cluster create. Cloider provisions a dedicated control plane with its own API endpoint.Connect kubectl
cloider cluster kubeconfig --merge --useadds the cluster to your kubeconfig, or download the file from the dashboard.Deploy your workloads
Apply manifests or install Helm charts as on any Kubernetes cluster, with CPU and memory requests that match what your pods use.
Publish them
Route a domain to a Service for HTTPS, or give it a
LoadBalancerfor a public IP.
# create a cluster and point kubectl at it
cloider cluster create staging --wait
cloider cluster kubeconfig staging --merge --use
# deploy as on any Kubernetes cluster
kubectl create deployment web --image=nginx
kubectl get pods
Pay for pods, volumes and load balancers
Three meters, all prorated to the second. There are no node charges and no fee per cluster.
| Compute | $35 per compute unit per month |
|---|---|
| Compute unit | 1 vCPU or 3 GB of memory, whichever is larger. Each container counts the larger of its request and its use |
| Persistent volumes | $0.50 per GiB per month |
| Load balancers | $10 per month each |
| Nodes, cluster fee, egress | Free |
Pods that request 1 vCPU and 2 GB in total, and use no more, are 1 compute unit: $35.00 a month. The control plane's own small CPU and memory use counts as compute, and a paused cluster pays only for its volumes. Every price is on the pricing page.
Kubernetes questions
What is serverless Kubernetes?
A serverless Kubernetes cluster is one where the provider runs all the infrastructure underneath and bills you for what your pods request or use, not for servers. On Cloider, each cluster has its own Kubernetes API server, namespaces, RBAC and objects, and your pods run on nodes that Cloider operates, patches and scales.
How is Cloider Kubernetes different from EKS, GKE or AKS?
On Amazon EKS, Google GKE or Azure AKS you normally choose node pools and machine sizes and pay for every node, busy or idle. A Cloider cluster has no node pools: you declare pods with requests, and each container is billed for the larger of its request and its actual use. The Kubernetes API is the standard one, so kubectl, Helm and your manifests work the same way.
How is a Cloider Kubernetes cluster billed?
On three meters, all prorated to the second. Compute is billed in compute units of 1 vCPU or 3 GB of memory, whichever is larger, for the larger of each running container's request and its use, at $35 per unit per month. Volumes are billed per GiB provisioned ($0.50 per GiB per month), and each LoadBalancer Service with a public IP per load balancer ($10 per month). There is no fee per cluster, but the control plane's own small CPU and memory use counts as compute.
Does a paused cluster cost anything?
A paused cluster runs no pods, so it accrues no compute charge. Its persistent volumes are kept and still billed per GiB, and any LoadBalancer Service that still holds a public IP is still metered.
Does Ingress work on Cloider Kubernetes?
Ingress objects created inside a Cloider cluster are not served. To publish an HTTP service on a domain, create a target that routes the domain, or a path of it, to your Service. Cloider creates the ingress and the HTTPS certificate for you.
Can I manage clusters with Terraform?
Yes. The cloider_cluster resource of the Cloider Terraform provider creates clusters, and Terraform's kubernetes and helm providers deploy into them. The provider also works with OpenTofu.
More answers in the documentation, or contact us.