* `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

    fun retryOnConnectionFailure(retryOnConnectionFailure: Boolean) =
      apply {
        this.retryOnConnectionFailure = retryOnConnectionFailure
      }

    /**
     * Configure this client to perform fast fallbacks by attempting multiple connections
     * concurrently, returning once any connection connects successfully.
     *
     * This implements Happy Eyeballs ([RFC 6555][rfc_6555]), balancing connect latency vs.

* Add automatic etcd 3.2->3.1 and 3.1->3.0 minor version rollback support to gcr.io/google_container/etcd images. For HA clusters, all members must be stopped before performing a rollback. ([#59298](https://github.com/kubernetes/kubernetes/pull/59298), [@jpbetz](https://github.com/jpbetz))

		case isSysErrInvalidArg(err):
			st, _ := Lstat(filePath)
			if st != nil && st.IsDir() {
				// Linux returns InvalidArg for directory O_DIRECT
				// we need to keep this fallback code to return correct
				// errors upwards.
				return nil, dmTime, errFileNotFound
			}
			return nil, dmTime, errUnsupportedDisk
		}
		return nil, dmTime, err
	}
	defer f.Close()

  - Better detection of installed CRI.
  - Better output for image pull errors.
- Improved air-gapped and offline support
  - kubeadm now handles air-gapped environments by using the local client version as a fallback.
  - Some kubeadm commands are now allowed to work in a completely offline mode.
- Certificate handling improvements:
  - Renew certs as part of upgrade.

* Restored the ability of kubectl running inside a pod to consume resource files specifying a different namespace than the one the pod is...

  - `kube-apiserver` now only aggregates openapi schemas from `/openapi/v2` endpoints of aggregated API servers. The fallback to aggregate from `/swagger.json` has been removed. Ensure aggregated API servers provide schema information via `/openapi/v2` (available since v1.10). ([#73441](https://github.com/kubernetes/kubernetes/pull/73441), [@roycaihw](https://github.com/roycaihw))...

- Kubeadm: support automatic retry after failing to pull image ([#86899](https://github.com/kubernetes/kubernetes/pull/86899), [@SataQiu](https://github.com/SataQiu)) [SIG Cluster Lifecycle]
- Kubeadm: upgrade supports fallback to the nearest known etcd version if an unknown k8s version is passed ([#88373](https://github.com/kubernetes/kubernetes/pull/88373), [@SataQiu](https://github.com/SataQiu)) [SIG Cluster Lifecycle]

* Remove the restart-kube-proxy and restart-apiserver functions ([#23180](https://github.com/kubernetes/kubernetes/pull/23180), [@roberthbailey](https://github.com/roberthbailey))
* Copy annotations back from RS to Deployment on rollback ([#23160](https://github.com/kubernetes/kubernetes/pull/23160), [@janetkuo](https://github.com/janetkuo))

- The fallback behavior of the Downward API's `resourceFieldRef` field was updated to account for pod-level resources: if container-level limits were not set, pod-level limits were now used before falling back to node allocatable resources. ([#132605](https://...