Operations
Once your clusters and workloads are running, day-2 operations keep them healthy. Kupe Cloud handles infrastructure operations (node patching, control plane maintenance) — your team owns application-level operations and Kubernetes version upgrades.
Shared responsibility
Section titled “Shared responsibility”| Kupe manages | Your team manages |
|---|---|
| Host node OS patching and security updates | Application deployments and rollbacks |
| Control plane availability and backups | Namespace and RBAC design |
| Kupe platform component upgrades (Argo CD, Grafana, and other managed platform services) | Kubernetes minor version upgrades (console-guided) |
| Storage and network infrastructure | Workload health, scaling, and capacity planning |
Kubernetes upgrades
Section titled “Kubernetes upgrades”Kupe supports in-place Kubernetes minor version upgrades through the console. Upgrades follow a guided workflow:
- Check compatibility — review release notes and test in a non-production cluster.
- Initiate upgrade — start the upgrade from the console cluster settings.
- Monitor progress — the console shows upgrade status and any issues.
- Validate — confirm workloads are healthy, APIs are responding, and no deprecation warnings appear.
Common operational tasks
Section titled “Common operational tasks”- Investigate unhealthy pods — events, logs, and describe output in the console.
- Roll back a bad deployment — revert the Git commit and let Argo CD reconcile.
- Scale workloads — adjust replica counts or HPA targets in your manifests.
- Respond to alerts — follow the triage path from alert to dashboard to logs.
- Decommission a cluster — remove workloads, verify no external dependencies, then delete.
Guides
Section titled “Guides”- Upgrade Clusters — step-by-step Kubernetes version upgrade workflow.