* Update the GCE windows node image to include hot fixes since July. ([#81233](https://github.com/kubernetes/kubernetes/pull/81233), [@YangLu1031](https://github.com/YangLu1031))
* switch to VM Update call in attach/detach disk operation, original CreateOrUpdate call may lead to orphaned VMs or blocked resources ([#81208](https://github.com/kubernetes/kubernetes/pull/81208), [@andyzhangx](https://github.com/andyzhangx))

- 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

    * inside of which a write lock was intended to prevent races against
    * VM creation, but the usage of the sync.RWMutex type is incorrect,
    * plus is placed in a defer call so the locking is effectively
    * nonexistent.  This patch adds an anonymous function call inside
    * the loop to correctly handle acquiring the write lock with deferred
    * unlock to run at that function's completion, following the normal

LIST calls into multiple smaller chunks. A client can specify a limit to the number of results to return. If more results exist, a token is returned that allows the client to continue the previous list call repeatedly until all results are retrieved.  The resulting list is identical to a list call that does not perform chunking, thanks to capabilities provided by etcd3.  This allows the server to use less memory and CPU when very large lists are returned. This feature is gated as APIListChunking...

calls into multiple smaller chunks.  A client can specify a limit to the number of results to return, and if more results exist a token will be returned that allows the client to continue the previous list call repeatedly until all results are retrieved.  The resulting list is identical to a list call that does not perform chunking thanks to capabilities provided by etcd3.  This allows the server to use less memory and CPU responding with very large lists.  This feature is gated as APIListChunking and...

   * Returns true when all work currently in progress by the watchdog have completed. This method
   * creates more work for the watchdog and waits for that work to be executed. When it is, we know
   * work that preceded this call is complete.
   */
  private fun awaitWatchdogIdle() {
    val latch = CountDownLatch(1)
    val watchdogJob: AsyncTimeout =
      object : AsyncTimeout() {
        override fun timedOut() {

- Fixed panic when accessing CustomResources of a CRD with `x-kubernetes-int-or-string`. ([#83787](https://github.com/kubernetes/kubernetes/pull/83787), [@sttts](https://github.com/sttts))

* Remove broken getvolumename and pass PV or volume name to attach call ([#46249](https://github.com/kubernetes/kubernetes/pull/46249), [@chakri-nelluri](https://github.com/chakri-nelluri))

* Make watch cache treat resourceVersion consistent with uncached watch ([#24008](https://github.com/kubernetes/kubernetes/pull/24008), [@liggitt](https://github.com/liggitt))

### Other notable changes

* Trusty: Add retry in curl commands ([#24749](https://github.com/kubernetes/kubernetes/pull/24749), [@andyzheng0831](https://github.com/andyzheng0831))

  Noteable changes:
  1. Populate new `RuntimeHandler` field in CRI's `ImageSpec` struct during image pulls from container runtimes.
  2. Pass `runtimeHandler` field in `RemoveImage()` call to container runtime in `kubelet`'s image garbage collection. ([#121456](https://github.com/kubernetes/kubernetes/pull/121456), [@kiashok](https://github.com/kiashok))