- The `WaitFor` and `WaitForWithContext` functions in the wait package have now been marked private. Callers should use the equivalent `Poll*` method with a zero duration interval. ([#115116](https://github.com/kubernetes/kubernetes/pull/115116), [@smarterclayton](https://github.com/smarterclayton))

- Kube-proxy handles Terminating EndpointSlices conditions and enables zero downtime deployments for Services with ExternalTrafficPolicy=Local author: @andrewsykim ([#117718](https://github.com/kubernetes/kubernetes/pull/117718), [@aojea](https://github.com/aojea)) [SIG Network, Testing and Windows]

* Change CollisionCount from int64 to int32 across controllers ([#50575](https://github.com/kubernetes/kubernetes/pull/50575), [@dixudx](https://github.com/dixudx))
* fix GPU resource validation that incorrectly allows zero limits ([#50218](https://github.com/kubernetes/kubernetes/pull/50218), [@dixudx](https://github.com/dixudx))

- In the PriorityLevelConfiguration object, the `nominalConcurrencyShares` field now accepts a zero value in both the `flowcontrol.apiserver.k8s.io/v1` and `flowcontrol.apiserver.k8s.io/v1beta3` APIs.
   ([#123001](https://github.com/kubernetes/kubernetes/pull/123001), [@tkashem](https://github.com/tkashem))

- Fixed CEL estimated cost of `replace()` to handle a zero length replacement string correctly.
  Previously this would cause the estimated cost to be higher than it should be. ([#120097](https://github.com/kubernetes/kubernetes/pull/120097), [@jpbetz](https://github.com/jpbetz)) [SIG API Machinery]

- Fixed relative CPU priority for pods where containers explicitly request zero cpu by giving the lowest priority instead of falling back to the cpu limit to avoid possible cpu starvation of other pods. ([#108832](https://github.com/kubernetes/kubernetes/pull/108832), [@waynepeking348](https://github.com/waynepeking348))