1.22 addressed a long-standing issue in the Kubelet where terminating pods were [vulnerable to race conditions](https://github.com/kubernetes/kubernetes/pull/102344) leading to early shutdown, resource leaks, or long delays in actually completing pod shutdown. As a consequence of this change the Kubelet now correctly takes into account the resources of running and terminating pods when deciding to accept new pods, since terminating pods are still holding on to those resources. This stricter handling...

- A terminating pod on a node that is not caused by preemption no longer prevents `kube-scheduler` from preempting pods on that node

- Fixed the preStop hook will block the pod termination grace period ([#115835](https://github.com/kubernetes/kubernetes/pull/115835), [@HirazawaUi](https://github.com/HirazawaUi)) [SIG Node and Testing]

- Add a `serving` and `terminating` condition to the EndpointSlice API.
  `serving` tracks the readiness of endpoints regardless of their terminating state. This is distinct from `ready` since `ready` is only true when pods are not terminating. 

- Kubelet: Fix the volume manager didn't check the device mount state in the actual state of the world before marking the volume as detached. It may cause a pod to be stuck in the Terminating state due to the above issue when it was deleted. ([#129064](https://github.com/kubernetes/kubernetes/pull/129064), [@carlory](https://github.com/carlory)) [SIG Node]

## Dependencies

### Added
_Nothing has changed._

- EndpointSlice should not contain endpoints for terminating pods ([#89056](https://github.com/kubernetes/kubernetes/pull/89056), [@andrewsykim](https://github.com/andrewsykim)) [SIG Apps and Network]

- The `ProxyTerminatingEndpoints` feature is now Beta and enabled by default. When enabled, kube-proxy will attempt to route traffic to terminating pods when the traffic policy is `Local` and there are only terminating pods remaining on a node. ([#113363](https://github.com/kubernetes/kubernetes/pull/113363), [@andrewsykim](https://github.com/andrewsykim))
- The `goroutines` metric is newly added in the scheduler. 

    - Traffic to a service IP now starts getting rejected (as opposed to
      merely dropped) as soon as there are no longer any *usable*
      endpoints, rather than waiting until all of the terminating
      endpoints have terminated even when those terminating endpoints
      were not being used.
  
    - Chains for endpoints that won't be used are no longer output to

- Fixed a bug where hard evictions due to resource pressure allowed pods
  to use the full termination grace period instead of shutting down instantly. 
  This bug also affected force deleted pods. Both cases now receive a termination grace period of 1 second. ([#124063](https://github.com/kubernetes/kubernetes/pull/124063), [@olyazavr](https://github.com/olyazavr))

* Fix race condition introduced by graceful termination which can lead to a deadlock in kube-proxy ([#72361](https://github.com/kubernetes/kubernetes/pull/72361), [@lbernail](https://github.com/lbernail))
* Support graceful termination with IPVS when deleting a service ([#71895](https://github.com/kubernetes/kubernetes/pull/71895), [@lbernail](https://github.com/lbernail))