- The AWS credential provider can now obtain ECR credentials even without the AWS cloud provider or being on an EC2 instance. Additionally, AWS credential provider caching has been improved to honor the ECR credential timeout. ([#75587](https://github.com/kubernetes/kubernetes/pull/75587), [@tiffanyfay](https://github.com/tiffanyfay))


### Azure

      ...

                          documentation for details. \n <gateway:experimental:description>
                          \n When the BackendRef points to a Kubernetes Service, implementations
                          SHOULD honor the appProtocol field if it is set for the
                          target Service Port. \n Implementations supporting appProtocol
                          SHOULD recognize the Kubernetes Standard Application Protocols

the issued certificate to determine the actual duration.\n\nThe v1.22+ in-tree implementations of the well-known Kubernetes signers will honor this field as long as the requested duration is not greater than the maximum duration they will honor per the --cluster-signing-duration CLI flag to the Kubernetes controller manager.\n\nCertificate signers may not honor this field for various reasons:\n\n  1. Old signer that is unaware of the field (such as the in-tree\n     implementations prior to v1.22)\n...

- Kubectl: fix the bug that kubectl autoscale does not honor '--name' flag ([#91855](https://github.com/kubernetes/kubernetes/pull/91855), [@SataQiu](https://github.com/SataQiu)) [SIG CLI]
- Kubectl: fix the bug that kubectl scale does not honor '--timeout' flag ([#91858](https://github.com/kubernetes/kubernetes/pull/91858), [@SataQiu](https://github.com/SataQiu)) [SIG CLI]

- In a scheduler with `Permit` plugins, when a Pod is rejected during `WaitOnPermit`, the scheduler records the plugin.
  The scheduler will use the record to honor cluster events and queueing `hints registered` for the plugin, to inform whether to retry the pod. ([#119785](https://github.com/kubernetes/kubernetes/pull/119785), [@sanposhiho](https://github.com/sanposhiho))

- `kubectl exec` and `kubectl attach` now honor the `--quiet` flag which suppresses output from the local binary that could be confused by a script with the remote command output (all non-failure output is hidden). In addition, print inline with exec and attach the list of alternate containers when...

	nodeTaintsField := subFldPath0.Child("nodeTaintsPolicy")
	labelSelectorField := subFldPath0.Child("labelSelector")
	unknown := core.NodeInclusionPolicy("Unknown")
	ignore := core.NodeInclusionPolicyIgnore
	honor := core.NodeInclusionPolicyHonor

	testCases := []struct {
		name            string
		constraints     []core.TopologySpreadConstraint
		wantFieldErrors field.ErrorList

	}
	p[0] = b
	return 1, nil
}

// Verifies that the Transport doesn't reuse a connection in the case
// where the server replies before the request has been fully
// written. We still honor that reply (see TestIssue3595), but don't
// send future requests on the connection because it's then in a
// questionable state.
// golang.org/issue/7569
func TestTransportNoReuseAfterEarlyResponse(t *testing.T) {

- Fixes a regression in the pod binding subresource to honor the `metadata.uid` precondition.