- Fix a scheduler preemption issue where the victim pod was not deleted due to incorrect status patching. This issue occurred when the preemptor and victim pods had different QoS classes in their status, causing the preemption to fail entirely. ([#126691](https://github.com/kubernetes/kubernetes/pull/126691), [@Huang-Wei](https://github.com/Huang-Wei)) [SIG Scheduling]

- Fixed a scheduler preemption issue where the victim pod was not deleted due to incorrect status patching. This issue occurred when the preemptor and victim pods had different QoS classes in their status, causing the preemption to fail entirely. ([#126644](https://github.com/kubernetes/kubernetes/pull/126644), [@Huang-Wei](https://github.com/Huang-Wei))

- Fix a scheduler preemption issue where the victim pod was not deleted due to incorrect status patching. This issue occurred when the preemptor and victim pods had different QoS classes in their status, causing the preemption to fail entirely. ([#126695](https://github.com/kubernetes/kubernetes/pull/126695), [@Huang-Wei](https://github.com/Huang-Wei)) [SIG Scheduling]

- `GlusterFS` in-tree storage driver which was deprecated in kubernetes 1.25 release is now removed entirely in `v1.26`. Volumes must be migrated to an alternate storage solution before upgrading to `v1.26`. ([#112015](https://github.com/kubernetes/kubernetes/pull/112015), [@humblec](https://github.com/humblec))

## Changes by Kind

- The formerly publicly-available cAdvisor web UI that the kubelet started using `--cadvisor-port` has been entirely removed in 1.12. The recommended way to run cAdvisor if you still need it, is via a DaemonSet. ([#65707](https://github.com/kubernetes/kubernetes/pull/65707), [@dims](https://github.com/dims))

* kubectl delete does not use reapers for removing objects anymore, but relies on server-side GC entirely ([#63979](https://github.com/kubernetes/kubernetes/pull/63979), [@soltysh](https://github.com/soltysh))

  - The `metadata.selfLink` field is deprecated in individual and list objects. It will no longer be returned starting in v1.20, and the field will be removed entirely in v1.21. ([#80978](https://github.com/kubernetes/kubernetes/pull/80978), [@wojtek-t](https://github.com/wojtek-t))

- Skipped pod backoff entirely when the `PodMaxBackoffDuration` kube-scheduler option was set to zero and the `SchedulerPopFromBackoffQ` feature gate was enabled. ([#131965](https://github.com/kubernetes/kubernetes/pull/131965), [@macsko](https://github.com/macsko))

- Fix a scheduler preemption issue where the victim pod was not deleted due to incorrect status patching. This issue occurred when the preemptor and victim pods had different QoS classes in their status, causing the preemption to fail entirely. ([#126694](https://github.com/kubernetes/kubernetes/pull/126694), [@Huang-Wei](https://github.com/Huang-Wei)) [SIG Scheduling]

 * class to read a response that is larger than the entire memory allocated to the current process.
 * It can even stream a response larger than the total storage on the current device, which is a
 * common requirement for video streaming applications.
 *
 * Because this class does not buffer the full response in memory, the application may not
 * re-read the bytes of the response. Use this one shot to read the entire response into memory with