* Fix issues related to the eviction manager.

* Fix inconsistent Prometheus cAdvisor metrics.

* Fix issues related to the local storage allocatable feature.

### Auth

  * Add flag -t as shorthand for --tty
* Resources
  * Improved node stability by *optionally* evicting pods upon memory pressure - [Design Doc](https://github.com/kubernetes/kubernetes/blob/release-1.3/docs/proposals/kubelet-eviction.md)
  * **alpha**: NVIDIA GPU support ([#24836](https://github.com/kubernetes/kubernetes/pull/24836), [@therc](https://github.com/therc))
  * Adding loadBalancer services and nodeports services to quota system

- Fixed two regressions introduced by the PodDisruptionConditions feature (on by default in 1.26):
  - pod eviction API calls returned spurious precondition errors and required a second evict API call to succeed
  - dry-run eviction API calls persisted a DisruptionTarget condition into the pod being evicted ([#116750](https://github.com/kubernetes/kubernetes/pull/116750), [@atiratree](https://github.com/atiratree)) [SIG API Machinery and Testing]

- Fix the bug where spec.terminationGracePeriodSeconds of the pod will always be overwritten by the MaxPodGracePeriodSeconds of the soft eviction, you can enable the `AllowOverwriteTerminationGracePeriodSeconds` feature gate, which will restore the previous behavior.  If you do need to set this, please file an issue with the Kubernetes project to help contributors understand why you need it....

  - PreemptionByKubeScheduler (Pod preempted by kube-scheduler)
  - DeletionByTaintManager (Pod deleted by taint manager due to NoExecute taint)
  - EvictionByEvictionAPI (Pod evicted by Eviction API)
  - DeletionByPodGC (an orphaned Pod deleted by PodGC) ([#110959](https://github.com/kubernetes/kubernetes/pull/110959), [@mimowo](https://github.com/mimowo))

- The `pod/eviction` subresource now accepts `policy/v1` eviction requests in addition to `policy/v1beta1` eviction requests ([#100724](https://github.com/kubernetes/kubernetes/pull/100724), [@liggitt](https://github.com/liggitt))

    filesystem.setFaultyDelete(cacheDir / "a.0", true)
    taskFaker.runNextTask()

    // Confirm we prevent edits after a trim failure.
    assertThat(cache.edit("c")).isNull()

    // A successful eviction should allow new writes.
    filesystem.setFaultyDelete(cacheDir / "a.0", false)
    cache.evictAll()
    set("c", "cc", "cc")
    assertValue("c", "cc", "cc")
  }

  @ParameterizedTest

- Pod eviction now honors graceful deletion by default if no delete options are provided in the eviction request. ([#72730](https://github.com/kubernetes/kubernetes/pull/72730), [@liggitt](https://github.com/liggitt))

### Auth

- Fixed a missing behavior where Windows nodes did not implement memory-pressure eviction. ([#122922](https://github.com/kubernetes/kubernetes/pull/122922), [@marosset](https://github.com/marosset)) [SIG Node, Testing and Windows]

  E0303 17:25:03.436781 63839 summary_sys_containers.go:47] Failed to get system container stats for "/docker": failed to get cgroup stats for "/docker": failed to get container info for "/docker": unknown container "/docker"