- DRA: when scheduling many pods very rapidly, sometimes the same device was allocated twice for different ResourceClaims due races between data processing in different goroutines. Depending on whether DRA drivers check for this during NodePrepareResources (they should, but maybe not all implement this...

  This eliminates the problem of rolling updates to a DaemonSet getting stuck around tolerations. ([#120792](https://github.com/kubernetes/kubernetes/pull/120792), [@mochizuki875](https://github.com/mochizuki875)) [SIG Apps and Testing]

  |   improving on that
  |
  | o use enums where appropriate (eg dependency scope)
  |
  | o a number of elements have a groupId/artifactId and sometimes version. It would be good to have them all extend one
  |   definition of these types
  |
-->
<model xmlns="http://codehaus-plexus.github.io/MODELLO/2.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"

   ([#122068](https://github.com/kubernetes/kubernetes/pull/122068), [@caohe](https://github.com/caohe))
- Fixed an issue where `AvailableBytes` sometimes did not report correctly on WindowsNodes when the `PodAndContainerStatsFromCRI` feature was enabled.
   ([#122846](https://github.com/kubernetes/kubernetes/pull/122846), [@marosset](https://github.com/marosset))

  which is not allowed. ([#121632](https://github.com/kubernetes/kubernetes/pull/121632), [@kerthcet](https://github.com/kerthcet))

*   `tf.function`/AutoGraph:
    *   `AutoGraph` now includes into TensorFlow loops any variables that are
        closed over by local functions. Previously, such variables were
        sometimes incorrectly ignored.
    *   functions returned by the `get_concrete_function` method of
        `tf.function` objects can now be called with arguments consistent with

- Fixed a regression that sometimes prevented `kubectl port-forward` to work when TCP and UDP services were configured on the same port ([#94728](https://github.com/kubernetes/kubernetes/pull/94728), [@amorenoz](https://github.com/amorenoz)) [SIG CLI]

### Bug or Regression

* Fix a bug in GCE multizonal clusters where PersistentVolumes were sometimes created in zones without nodes. ([#52322](https://github.com/kubernetes/kubernetes/pull/52322), [@davidz627](https://github.com/davidz627))

* Fix a bug in GCE multizonal clusters where PersistentVolumes were sometimes created in zones without nodes. ([#52322](https://github.com/kubernetes/kubernetes/pull/52322), [@davidz627](https://github.com/davidz627))