then kube-proxy will send the traffic to the Terminating pod rather than dropping it. This
  patches up a race condition between when a pod is killed and when the external load
  balancer notices that it has been killed. ([#97238](https://github.com/kubernetes/kubernetes/pull/97238), [@andrewsykim](https://github.com/andrewsykim))

- Implement `post-filter` extension point for scheduling framework ([#78097](https://github.com/kubernetes/kubernetes/pull/78097), [@draveness](https://github.com/draveness))
- Add Bind extension point of the scheduling framework ([#78513](https://github.com/kubernetes/kubernetes/pull/78513), [@chenchun](https://github.com/chenchun))

### Feature

- A new metric `apiserver_request_filter_duration_seconds` has been introduced that 
  measures request filter latency in seconds. ([#95207](https://github.com/kubernetes/kubernetes/pull/95207), [@tkashem](https://github.com/tkashem)) [SIG API Machinery and Instrumentation]

Plugins that are enabled via `multiPoint` will automatically be registered for each individual extension point that they implement.
For example, a plugin that implements Score and Filter extensions can be simultaneously enabled for both.
This means entire plugins can be enabled and disabled without having to manually edit individual extension point settings.

  - `read_vs_write_request_count_watermarks` is removed. ([#110104](https://github.com/kubernetes/kubernetes/pull/110104), [@MikeSpreitzer](https://github.com/MikeSpreitzer)) [SIG API Machinery, Instrumentation...

- Fix a regression that the scheduler always goes through all Filter plugins. ([#114526](https://github.com/kubernetes/kubernetes/pull/114526), [@Huang-Wei](https://github.com/Huang-Wei)) [SIG Scheduling]

* fix: azure disk could not mounted on Standard_DC4s/DC2s instances ([#86612](https://github.com/kubernetes/kubernetes/pull/86612), [@andyzhangx](https://github.com/andyzhangx))
* Filter published OpenAPI schema by making nullable, required fields non-required in order to avoid kubectl to wrongly reject null values. ([#85722](https://github.com/kubernetes/kubernetes/pull/85722), [@sttts](https://github.com/sttts))

* Enable basic auth username rotation for GCI ([#44590](https://github.com/kubernetes/kubernetes/pull/44590), [@ihmccreery](https://github.com/ihmccreery))
* vSphere cloud provider: Filter out IPV6 node addresses. ([#45181](https://github.com/kubernetes/kubernetes/pull/45181), [@BaluDontu](https://github.com/BaluDontu))

behavior of some watch calls to the server when filtering on fields was incorrect. If watching objects with a filter, when an update was made that no longer matched the filter a DELETE event was correctly sent. However, the object that was returned by that delete was not the (correct) version before the update, but instead, the newer version. That meant the new object was not matched by the filter. This was a regression from behavior between cached watches on the server side and uncached watches, and...

- Added `apiserver_envelope_encryption_dek_cache_filled` to measure number of records in data encryption key (DEK) cache. ([#119878](https://github.com/kubernetes/kubernetes/pull/119878), [@ritazh](https://github.com/ritazh))