* Add metrics exporter to the fluentd-gcp deployment ([#45096](https://github.com/kubernetes/kubernetes/pull/45096), [@crassirostris](https://github.com/crassirostris))
* Fix the bug where StartedAt time is not reported for exited containers. ([#45977](https://github.com/kubernetes/kubernetes/pull/45977), [@yujuhong](https://github.com/yujuhong))

- Fix kubectl printer to correctly handle timestamps of events emitted using events.k8s.io API ([#94226](https://github.com/kubernetes/kubernetes/pull/94226), [@ingvagabund](https://github.com/ingvagabund)) [SIG CLI]
  ```sh
  $ kubectl get event

		if rec := recover(); rec != nil {
			gridLogIf(ctx, fmt.Errorf("handleMessages: panic recovered: %v", rec))
			debug.PrintStack()
		}
		if debugPrint {
			fmt.Println("handleMessages: write goroutine exited")
		}
		cancel(ErrDisconnected)
		c.connChange.L.Lock()
		if atomic.CompareAndSwapUint32((*uint32)(&c.state), StateConnected, StateConnectionError) {
			c.connChange.Broadcast()
		}
		c.disconnected()

DaemonSet-related operations.
     * Behavior change:
        * DaemonSet pods will be created on nodes with .spec.unschedulable=true and will
not be evicted from nodes whose Ready condition is false.
        * Updates to the pod template are now permitted. To perform a rolling update of a
DaemonSet, update the pod template and then delete its pods one by one; they

# are present on the local host machine but not on the remote execution machine,
# leading to build failures. To resolve the issue, the following line is added
# to make sure all Genrule targets are excuted locally.
build:rbe_cross_compile_base_xla --config=rbe_cross_compile_base
build:rbe_cross_compile_base_xla --strategy=Genrule=standalone

# Due to the above strategy, all Genrule commands are executed locally, but the

				VersionID:  objInfo.VersionID,
			},
		}

	}
	names = append(names, ObjectToDelete{
		ObjectV: ObjectV{
			ObjectName: "dir/obj1",
			VersionID:  mustGetUUID(), // add a non-existent UUID.
		},
	})

	_, delErrs := obj.DeleteObjects(ctx, bucketName, names, ObjectOptions{
		Versioned: true,
	})
	for i := range delErrs {
		if delErrs[i] != nil {

  // pod is available (Ready for at least minReadySeconds) the old DaemonSet pod
  // on that node is marked deleted. If the old pod becomes unavailable for any
  // reason (Ready transitions to false, is evicted, or is drained) an updated
  // pod is immediatedly created on that node without considering surge limits.
  // Allowing surge implies the possibility that the resources consumed by the

* `kubectl rollout undo` now returns errors when attempting to rollback a deployment to a non-existent revision ([#70039](https://github.com/kubernetes/kubernetes/pull/70039), [@liggitt](https://github.com/liggitt))
    * `kubectl rollout undo` no longer uses the deprecated extensions/v1beta1 rollback API, which means that Events are no longer emitted when rolling back a deployment

- Kubeadm: when adding new control plane nodes with "kubeamd join", ensure that the etcd member addition is performed only if a given member URL does not already exist in the list of members. Similarly, on "kubeadm reset" only remove an etcd member if its ID exists. ([#127619](https://github.com/kubernetes/kubernetes/pull/127619), [@SataQiu](https://github.com/SataQiu)) [SIG Cluster Lifecycle]

### Other (Cleanup or Flake)

	}

	// Attempting to set a non-existent policy should fail.
	userDN := "uid=dillon,ou=people,ou=swengg,dc=min,dc=io"
	_, err = s.adm.AttachPolicyLDAP(ctx, madmin.PolicyAssociationReq{
		Policies: []string{policy + "x"},
		User:     userDN,
	})
	if err == nil {
		c.Fatalf("should not be able to attach non-existent policy")
	}

	userReq := madmin.PolicyAssociationReq{