Managed Services
Managed On-Premises Kubernetes
Operate self-managed Kubernetes control planes, nodes, networks, storage, workloads, security, upgrades, backup, and recovery.
Self-managed Kubernetes requires ownership from hardware through workloads. We manage the platform lifecycle while maintaining clear responsibility boundaries with application and data-centre teams.
When this service is needed
- On-premises clusters run production services.
- Control planes, etcd, certificates, and add-ons require maintenance.
- Upgrades and recovery are not tested regularly.
- Teams need support across hardware, OS, and Kubernetes.
Scope
- Hardware, OS, control-plane, etcd, node, network, and storage health.
- Certificates, keys, accounts, RBAC, policies, secrets, and registries.
- Workload status, capacity, scheduling, autoscaling, and disruption.
- Patches, versions, add-ons, CNI, CSI, ingress, and lifecycle compatibility.
- Metrics, logs, auditing, alerts, incidents, and problem management.
- etcd backup, resource backup, restore drills, and cluster recovery.
Operating model
- Onboard: establish inventory, baselines, access, dependencies, and ownership.
- Observe: monitor service, capacity, security, backup, and cost indicators.
- Operate: handle incidents, requests, changes, patches, and routine work.
- Improve: review recurring problems, risks, technical debt, and optimisation.
- Report: communicate activity, findings, trends, and next priorities.
Service controls
- Responsibility matrix and escalation paths.
- Change records, approval, and rollback for material changes.
- Runbooks and evidence for recurring work.
- Recurring review of capacity, lifecycle, risk, and cost.
Intended outcomes
The on-premises cluster gains end-to-end operations, a planned lifecycle, and tested, documented recovery capability.