* Fix stackdriver metrics for node memory using wrong metric type ([#63535](https://github.co/kubernetes/kubernetes/pull/63535), [@serathius](https://github.com/serathius))
* fix mount unmount failure for a Windows pod ([#63272](https://github.com/kubernetes/kubernetes/pull/63272), [@andyzhangx](https://github.com/andyzhangx))

#### General Fixes and Reliability

* Because some plugins mount volume on sub-directory of volume path, we need to distinguish between volume path and mount path to fix issue in reconstruction. ([#74653](https://github.com/kubernetes/kubernetes/pull/74653), [@cofyc](https://github.com/cofyc))

pkg syscall (darwin-arm64), const SYS_UNDELETE ideal-int
pkg syscall (darwin-arm64), const SYS_UNLINK = 10
pkg syscall (darwin-arm64), const SYS_UNLINK ideal-int
pkg syscall (darwin-arm64), const SYS_UNMOUNT = 159
pkg syscall (darwin-arm64), const SYS_UNMOUNT ideal-int
pkg syscall (darwin-arm64), const SYS_UTIMES = 138
pkg syscall (darwin-arm64), const SYS_UTIMES ideal-int
pkg syscall (darwin-arm64), const SYS_VFORK = 66

- On compatible systems, a mounter's Unmount implementation is changed to not return an error when the specified target can be detected as not a mount point. On Linux, the behavior of detecting a mount point depends on `umount` command is validated when the mounter is created. Additionally, mount point checks will be skipped in CleanupMountPoint/CleanupMountWithForce if the mounter's Unmount having the changed behavior of not returning error when target is not a...

### Non-recursive Volume Ownership (FSGroup) graduates to Beta

By default, the `fsgroup` setting, if specified, recursively updates permissions for every file in a volume on every mount. This can make mount, and pod startup, very slow if the volume has many files. 

- Do not unmount and mount subpath bind mounts during container creation unless bind mount changes ([#105512](https://github.com/kubernetes/kubernetes/pull/105512), [@gnufied](https://github.com/gnufied)) [SIG Storage]

- On compatible systems, a mounter's Unmount implementation is changed to not return an error when the specified target can be detected as not a mount point. On Linux, the behavior of detecting a mount point depends on `umount` command is validated when the mounter is created. Additionally, mount point checks will be skipped in CleanupMountPoint/CleanupMountWithForce if the mounter's Unmount having the changed behavior of not returning error when target is not a...

- Fix: azure file mount timeout issue ([#88610](https://github.com/kubernetes/kubernetes/pull/88610), [@andyzhangx](https://github.com/andyzhangx)) [SIG Cloud Provider and Storage]
- Fix: corrupted mount point in csi driver ([#88569](https://github.com/kubernetes/kubernetes/pull/88569), [@andyzhangx](https://github.com/andyzhangx)) [SIG Storage]

pkg syscall (netbsd-arm64-cgo), const SYS_UNLINKAT ideal-int
pkg syscall (netbsd-arm64-cgo), const SYS_UNLINK ideal-int
pkg syscall (netbsd-arm64-cgo), const SYS_UNMOUNT = 22
pkg syscall (netbsd-arm64-cgo), const SYS_UNMOUNT ideal-int
pkg syscall (netbsd-arm64-cgo), const SYS_UTIMENSAT = 467
pkg syscall (netbsd-arm64-cgo), const SYS_UTIMENSAT ideal-int
pkg syscall (netbsd-arm64-cgo), const SYS_UTIMES = 420

- Fix: set "host is down" as corrupted mount. When SMB server is down, there is no way to terminate pod which is using SMB mount, would get an error. ([#101398](https://github.com/kubernetes/kubernetes/pull/101398), [@andyzhangx](https://github.com/andyzhangx))