What it does
Kubernetes Kapsule is Scaleway's managed Kubernetes service for containerized applications. A cluster wizard configures its control plane, private network, and node pools; Kubernetes workloads are then managed with standard tools such as kubectl. Scaleway operates the control plane and core components and offers autoscaling and node autohealing. This reduces cluster-maintenance work, but teams still define their applications, node capacity, networking, and persistent storage.
Cluster creation is infrastructure design
The documented console flow begins under Containers > Kubernetes. Choose a region, create a cluster, select Kapsule rather than Kosmos, and set its Kubernetes version and shared or dedicated control-plane offer. The guide then has you attach a Private Network and configure one or more pools: availability zone, node type, system volume, node count, and optional autoscaling or autoheal. Review the configuration and submit it. Scaleway supplies a kubeconfig for local kubectl access; the console, CLI, API, and Terraform are also documented management paths. These are vendor-published steps, not a cluster created for this listing.
For an illustrative web service, place stateless replicas in a node pool and use standard Kubernetes manifests and an ingress controller. Decide storage separately: Kapsule can replace or restart nodes, so local node files are not durable. The product FAQ points stateful workloads to Kubernetes persistent volumes, with Scaleway block storage available through its storage class.
What the provider manages, and what it bills
Scaleway says it manages the Kubernetes control plane and core components, including the cloud controller manager and autoscaler. Autoscaling and autohealing can reduce routine cluster work, but workload manifests, scheduling constraints, and resource sizing remain the team's responsibility. The FAQ says there is no additional control-plane charge; nodes use the same pricing as underlying Compute Instances, while load balancers and persistent volumes are billed separately. The pricing index opened for this research listed Kapsule among container products but did not expose its node rates, so a region- and pool-specific cost still needs to be checked before deployment.
Pay attention to network and placement choices before submitting: the guide says a cluster cannot be moved to another Organization or Project after creation, and its Private Network cannot be detached later. Custom cloud-init can also override node defaults; Scaleway warns that it may break nodes and is not supported for troubleshooting.
Strengths
- Scaleway manages the Kubernetes control plane and core components rather than requiring the team to operate them itself.
- Clusters support autoscaling, node autohealing, private-network integration, and standard Kubernetes tooling.
- The product can be managed through a developer API, CLI, or Terraform as well as the Scaleway console.
Watch out for
- Nodes may be restarted or replaced, so stateful workloads need persistent volumes instead of relying on local node files.
- The cluster's Organization, Project, and Private Network choices are difficult to revise after creation; node and add-on costs are separate from the control plane.
Best for
- Teams that need a managed, standard Kubernetes cluster on Scaleway Instances
- Container workloads that benefit from autoscaling, node recovery, and private-network links to other Scaleway services
Sources
- Kubernetes Kapsule | Orchestrate your containers in the cloud | Scaleway (Oct 25, 2026)
- How to create a Kubernetes Kapsule cluster | Scaleway Documentation (Oct 25, 2026)
- Pricing | Scaleway (Oct 25, 2026)