- `response.patch` and `response.patchType` are not permitted from validating admission webhooks
  - `apiVersion: "admission.k8s.io/v1"` is required
  - `kind: "AdmissionReview"` is required
  - `response.uid: "<value of request.uid>"` is required

  attempted first, and if it fails with a corrupt resource error then it triggers unsafe delete.
  In addition, when this feature is enabled, the 'details' field of 'Status' from the LIST response
  includes information that identifies the corrupt object(s).
  NOTE: unsafe deletion ignores finalizer constraints, and skips precondition checks.

  <mime-type type="application/mbms-msk+xml"/>
  <mime-type type="application/mbms-msk-response+xml"/>
  <mime-type type="application/mbms-protection-description+xml"/>
  <mime-type type="application/mbms-reception-report+xml"/>
  <mime-type type="application/mbms-register+xml"/>
  <mime-type type="application/mbms-register-response+xml"/>
  <mime-type type="application/mbms-user-service-description+xml"/>

- Improved how the API server responds to **list** requests where the response format negotiates to Protobuf. List responses in Protobuf are marshalled one element at the time, drastically reducing memory needed to serve large collections. Streaming list responses can be disabled via the `StreamingCollectionEncodingToProtobuf`...

- Resolves a spurious "Unknown discovery response content-type" error in client-go discovery requests by tolerating extra content-type parameters in API responses ([#117637](https://github.com/kubernetes/kubernetes/pull/117637), [@seans3](https://github.com/seans3)) [SIG API Machinery]

- A kubelet parameter and config option has been added to change the CFS quota period from the default 100ms to some other value between 1µs and 1s. This was done to improve response latencies for workloads running in clusters with guaranteed and burstable QoS classes. ([#63437](https://github.com/kubernetes/kubernetes/pull/63437), [@szuecs](https://github.com/szuecs)) Courtesy of SIG API Machinery, SIG Apps, SIG Architecture,...

## Before Upgrading

* When Response is a `metav1.Status`, it is no longer copied into the audit.Event status. Only the "status", "reason" and "code" fields are set.  For example, when we run `kubectl get pods abc`, the API Server returns a status object:

  The API Server will no longer proxy non-101 responses for upgrade requests. This could break proxied backends (such as an extension API server) that respond to upgrade requests with a non-101 response code. ([#92941](https://github.com/kubernetes/kubernetes/pull/92941), [@tallclair](https://github.com/tallclair)) [SIG API Machinery]
 
## Changes by Kind

### Bug or Regression

### API Change

- A new field `omitManagedFields` has been added to both `audit.Policy` and `audit.PolicyRule` 
  so cluster operators can opt in to omit managed fields of the request and response bodies from 
  being written to the API audit log. ([#94986](https://github.com/kubernetes/kubernetes/pull/94986), [@tkashem](https://github.com/tkashem)) [SIG API Machinery, Auth, Cloud Provider and Testing]

- Added `apiserver_watch_list_duration_seconds` metrics which will measure response latency distribution in seconds for watchlist requests broken by group, version, resource and scope. ([#120490](https://github.com/kubernetes/kubernetes/pull/120490), [@p0lyn0mial](https://github.com/p0lyn0mial))