* Time for attach/detach retry operations has been decreased from 10-12s to 2-6s ([#56974](https://github.com/kubernetes/kubernetes/pull/56974), [@gnufied](https://github.com/gnufied))

### Auth

* Contexts must be named in kubeconfigs. ([#56769](https://github.com/kubernetes/kubernetes/pull/56769), [@dixudx](https://github.com/dixudx))

- DRA API: Fixed the `tolerations` field in exact and sub requests to drop properly when the `DRADeviceTaints` API is disabled. ([#132927](https://github.com/kubernetes/kubernetes/pull/132927), [@pohly](https://github.com/pohly))

- Fixed a bug in the `dra_operations_duration_seconds` metric where the `is_error` label was recording inverted values. Error operations now correctly report `is_error="true"`, and successful operations report `is_error="false"`. ([#135227](https://github.com/kubernetes/kubernetes/pull/135227), [@hime](https://github.com/hime)) [SIG Node]

     ([#73977](https://github.com/kubernetes/kubernetes/pull/73977), [@khenidak](https://github.com/khenidak))
- Add scheduling support for RuntimeClasses. RuntimeClasses can now specify nodeSelector constraints & tolerations, which are merged into the PodSpec for pods using that RuntimeClass. ([#80825](https://github.com/kubernetes/kubernetes/pull/80825), [@tallclair](https://github.com/tallclair))

### Scheduling

- Tolerations with the same key and effect will be merged into one that has the value of the latest toleration for best effort pods. ([#75985](https://github.com/kubernetes/kubernetes/pull/75985), [@ravisantoshgudimetla](https://github.com/ravisantoshgudimetla))

   ([#122558](https://github.com/kubernetes/kubernetes/pull/122558), [@linxiulei](https://github.com/linxiulei))
- Fixed CEL estimated cost for expressions that perform operations on the result of `map()` operations (e.g., `.map(...).exists(...)` ) to have the correct estimated cost instead of an unbounded cost.
   ([#123562](https://github.com/kubernetes/kubernetes/pull/123562), [@jpbetz](https://github.com/jpbetz))

- Revised the logic for `DaemonSet` rolling update to exclude nodes if scheduling constraints are not met. This eliminates the problem of rolling updates to a `DaemonSet` getting stuck around tolerations. ([#119317](https://github.com/kubernetes/kubernetes/pull/119317), [@mochizuki875](https://github.com/mochizuki875))

  This allows the scheduler to handle cluster events faster with less memory.

  Specific node events include updates to taints, tolerations or allocatable.

- Moving `WindowsHostProcessContainers` feature to beta ([#106058](https://github.com/kubernetes/kubernetes/pull/106058), [@marosset](https://github.com/marosset))
- Node affinity, Node selectors, and tolerations are now mutable for Jobs that are suspended and have never been started ([#105479](https://github.com/kubernetes/kubernetes/pull/105479), [@ahg-g](https://github.com/ahg-g))

* fixes a memory leak in the kube-controller-manager observed when large numbers of pods with tolerations are created/deleted ([#65339](https://github.com/kubernetes/kubernetes/pull/65339), [@liggitt](https://github.com/liggitt))