- Open a CLI:
kubectl debug -n kube-system -it --image alpine node/k8s-control-1 - Open Dashboard:
talosctl dashboard - Create an etcd backup:
talosctl etcd snapshot db.snapshot
- Reconcile:
flux reconcile source git cluster -n flux-system - Show kustomizations:
flux get kustomizations --watch - Show everything:
flux get all -A
- Show all resources:
kubectl get all -A - Show events for an app:
kubectl events -n [name] --watch - Redirect port:
kubectl port-forward -n [namespace] service/[service_name] [local_port]:[pod_port] - Get a secret:
kubectl get secret -n namespace secret-name -o jsonpath="{.data.token}" | base64 -d
- Report engine-image migration status:
scripts/longhorn-engine-migrate.sh --mode report - Migrate one eligible detached V1 volume and wait for convergence:
scripts/longhorn-engine-migrate.sh --mode detached --limit 1 - Migrate one eligible attached V1 volume during a low-traffic window:
scripts/longhorn-engine-migrate.sh --mode attached --limit 1 - Before manual migration, confirm automatic engine upgrades are disabled:
kubectl -n longhorn-system get settings.longhorn.io concurrent-automatic-engine-upgrade-per-node-limit -o jsonpath='{.value}{"\n"}'(must be0) - Repeat each command until the report has no V1 volumes on the old image. Do not delete the old EngineImage manually: Longhorn removes a non-default, unreferenced EngineImage after about 60 minutes.