Managed HA control plane
Loopback operates the Kubernetes control plane in high availability. We take care of operations and updates, you take care of your workloads.
Kubernetes
Kubernetes without the day-to-day operational burden.
Create and operate clusters through one platform: HA control plane, private network, scaling groups, and one-click upgrades. You choose provider and location. The platform stays the same.
What Loopback takes on for the cluster.
HA control plane, private networking, and scaling groups.
- Managed HA control plane: Loopback operates the Kubernetes control plane in high availability. We take care of operations and updates, you take care of your workloads.
- Private networking: Loopback provides your cluster's private network, the same with every provider. Your nodes talk to each other directly over it.
- Cilium networking: Cilium is the CNI of the cluster. Loopback operates the managed cluster's private networking with it.
- NGINX Ingress: NGINX Ingress routes HTTP traffic to your services. You see Services and Ingress live in the cluster.
- Scaling groups: Node groups with auto-scale logic that you tune through multiple parameters. Your cluster grows with the load.
- Compute choice: x86/AMD64, ARM, GPU, and bare metal are available. You pick the right combination by provider and region.
- One-click upgrades: You roll out new Kubernetes versions with one click, in the maintenance window that fits your operations.
Private networking
Loopback provides your cluster's private network, the same with every provider. Your nodes talk to each other directly over it.
Cilium networking
Cilium is the CNI of the cluster. Loopback operates the managed cluster's private networking with it.
NGINX Ingress
NGINX Ingress routes HTTP traffic to your services. You see Services and Ingress live in the cluster.
Scaling groups
Node groups with auto-scale logic that you tune through multiple parameters. Your cluster grows with the load.
Compute choice
x86/AMD64, ARM, GPU, and bare metal are available. You pick the right combination by provider and region.
One-click upgrades
You roll out new Kubernetes versions with one click, in the maintenance window that fits your operations.
A consistent Kubernetes lifecycle.
From the control plane at the location you choose to the next upgrade. One operating model for the entire lifecycle.
01
Control plane
You create the control plane at the location you want.
02
Nodes
Then you create the nodes, also provider-agnostic, with private networking. The control plane and private network come from Loopback.
03
Scale
Auto-scale logic in scaling groups, adjustable individually through multiple parameters.
04
Update
One-click Kubernetes upgrades and maintenance windows.
Access
Kubernetes stays Kubernetes.
You copy or download the Admin Kubeconfig and OIDC Kubeconfig right in Loopback. The cluster stays standard Kubernetes, and you work with the tools you already know. Upgrades run through the platform, inside maintenance windows.
Kubeconfig
Download the Admin Kubeconfig and OIDC Kubeconfig in Loopback or copy them directly.
Standard access
Standard Kubernetes API. Manifests, Helm charts, and your usual tools work unchanged.
Maintenance windows
One-click upgrades including maintenance windows.
REST API
You can also drive the platform through the versioned REST API, for example from your pipeline.
What lives in the workspace
The workspace is the managed cluster in the project. Nodes, scaling groups, workloads, volumes, and endpoints stay in one context.
Compute / nodes
Create and manage hosts, with power actions, a stored SSH key, cordon/drain, and a web terminal. Node state stays visible in the workspace.
Scaling groups
Node groups with auto-scale logic that you tune through multiple parameters. Your cluster grows with the load.
Applications
Catalog deployments on the cluster and live workloads through the platform Kubernetes proxy. Applications live in the same workspace as nodes and endpoints.
Volumes
Persistent volumes in the workspace, including a live view. They stay in the cluster context alongside nodes and endpoints.
Endpoints
Kubernetes Services and Ingress as live networking in the cluster, with Cilium as the CNI and NGINX Ingress as an available ingress option. The load balancer sits in front of the project or directly on the workspace.
Upgrades
Cluster version, control-plane and host upgrades, and patch updates run through the platform. One-click upgrades including maintenance windows.
Kubeconfig
Admin Kubeconfig and OIDC Kubeconfig to copy or download. The cluster remains standard Kubernetes.
The Loopback UI for the workspace.
You see node state, CPU, memory, network, scaling groups, and versions at a glance. From the same view you scale, update, and manage your cluster.
A cluster in the project. Not an island.
Organization, project, and workspace build on each other: teams and access at the top, the managed cluster at the bottom. All in one account.
Organization
Account, teams, roles, and SSO. The frame for every resource.
Project
Shared context for clusters, storage, registry, and billing.
Workspace
The managed Kubernetes cluster. A project can hold more than one workspace.
Location stays a choice.
You choose provider and location per cluster. Interface, permissions, and billing stay the same everywhere.
- FSN1, NBG1, FRA1, FRA2Germany
- HEL1Finland
The facts.
| Control plane | HA, operated by Loopback |
|---|---|
| Private network | Operated by Loopback |
| CNI | Cilium |
| Ingress | NGINX as an option |
| Storage | Persistent volumes through the provider's CSI driver |
| Upgrades | One-click, including maintenance windows |
Keep working in the same project.
Object Storage, Registry, and Bundles share the account and project with the cluster.
A fixed price for the control plane.
The HA control plane is 50€ per month, billed hourly. Workers add on dynamically.
Kubernetes HA control plane
50€
per month, billed hourly Based on 720 hours
Plus dynamic costs for workers
- HA control plane
- Secure Network by FABRIC
- Scaling groups
- One-click upgrades
- Admin Kubeconfig and OIDC Kubeconfig
FAQ
Why not go straight to a cloud provider?
The selected providers supply the infrastructure. Loopback turns it into a platform: managed Kubernetes lifecycle, organizations, RBAC, SSO, REST API, and one bill across all providers. Your operations work the same way everywhere, whichever provider runs your cluster.
Do we need Kubernetes?
Kubernetes fits teams that need containers, portability, scaling, and standard orchestration. You can use Object Storage and Registry without a cluster too, each as a product of its own.
How much control do we keep?
You choose provider, region, and resources yourself. You keep standard Kubernetes access and S3-compatible interfaces, and you can take your data with you at any time through exactly those standard tools.
How does migration work?
Migrations run through the familiar Kubernetes, Helm, and registry tools. For assessment and guided migration, you can book our professional services.
Where is our data stored?
Loopback’s platform and core platform data are operated in Germany. The Germany location applies to the platform, not to your workloads. Where workloads run depends on the provider and location you choose.
What support is included?
Send standard questions by email to support@loopback.cloud. We respond Monday to Friday, 09:00 to 17:00 German time. Architecture workshops, migration assessment, and training are separate paid services.
How is pricing calculated?
You pay a Platform-Fee for the organization, as Start, Grow, or Scale. Dynamic usage costs add on: workers, Object Storage, Registry, load balancer, traffic, and the Kubernetes HA control plane. Monthly is the same rate times 720 hours, which is 30 days.
Is the control plane highly available?
Yes. Loopback operates a managed HA control plane.
Which providers and regions are supported?
The platform is provider-agnostic. You choose provider and location per cluster and run every cluster through the same interface, the same API, and the same bill.
Which CNI is used?
Cilium.
How do upgrades work?
Kubernetes upgrades run as one-click upgrades through the platform, including maintenance windows.
Can I use ARM, GPU, or bare metal?
x86/AMD64, ARM, GPU, and bare metal are available. You pick the right combination by provider and region.
How does persistent storage work?
Persistent volumes come from the CSI driver of the selected provider. You request them as usual with a PersistentVolumeClaim.
How do I automate Loopback?
Through the versioned REST API. You reach the platform functions in the interface and just as well through the API, with your account's API token and its permissions.