Timeout Controller

[wangeguo]

By design, Playbook has a lifecycle, from creation, initialization, compilation build, deployment, running to destruction.

Currently, we need to implement a new controller to perform periodic resource recovery (destruction) of Playbook. The following is a technical description:

1. The administrator can set a resource recovery duration, such as 2 weeks (default);
2. this duration is annotated on each Playbook resource definition;
3. Timeout Controller periodically scans for Playbooks that are about to expire and performs a series of actions, such as sending reminder emails or notifications (e.g. via Slack) 3 days or 2 days before the expiration date, and performs Playbook deletion on the same day;
4. Additional consideration is given to support Webhook notification of events in the future;

Reference:

1. Rust Kubernetes client and controller runtime https://github.com/kube-rs/kube
2. Kube controller-rs example: https://github.com/kube-rs/controller-rs
3. Amphitheatre Controllers: https://github.com/amphitheatre-app/amphitheatre/tree/master/controllers
4. k8s-ttl-controller https://github.com/fpetkovski/k8s-ttl-controller

Note: Contributors are requested to provide technical design solutions and implementation plans in the comments, as we need to communicate more fully.

[wangeguo]

Additional information: The Kubernetes platform itself has a ttl-controller that is only used for Job resource recovery, the code should be in the K8s repository and can also be viewed as a reference

[Guocork]

#235 This issue has been done.

[wangeguo]

#235 This issue has been done.

Cool，thanks very much！