Kubernetes Operator for Automated Policy-Based Checkpointing

The Checkpoint-Restore Operator manages checkpoint retention well but has no way to create checkpoints automatically. This project adds a new CheckpointSchedule CRD and controller that triggers checkpoint creation through four mechanisms: cron-based time schedules, CPU/memory resource thresholds, Kubernetes cluster events (node drain, pod eviction, preemption), and pod annotations for application-level signaling. All triggers call the existing Kubelet Checkpoint API via the Kubernetes API server proxy. Deliverables include the CheckpointSchedule CRD, a CheckpointCreator that calls the Kubelet API, four independent trigger subsystems, a TriggerManager to coordinate them, a full controller-runtime reconciler, RBAC updates, unit and BATS integration tests, and sample configuration. Together with the existing retention policies, this turns the operator into a complete checkpoint lifecycle manager — admins declare both when checkpoints are created and how long they are kept, with no custom scripting required.