kube-proxy is responsible for, rather than only counting the number of rules
  that it re-synced on the last sync. The new `sync_proxy_rules_iptables_last`
  metric now gives the latter number. ([#119140](https://github.com/kubernetes/kubernetes/pull/119140), [@danwinship](https://github.com/danwinship)) [SIG Network]

- Scheduler alpha metrics binding_duration_seconds and scheduling_algorithm_preemption_evaluation_seconds are deprecated, Both of those metrics are now covered as part of framework_extension_point_duration_seconds, the former as a PostFilter the latter and a Bind plugin. The plan is to remove both in 1.21 ([#95001](https://github.com/kubernetes/kubernetes/pull/95001), [@arghya88](https://github.com/arghya88)) [SIG Instrumentation and Scheduling]

- Kubeadm: Fix a bug where kubeadm cannot parse kubelet's version if the latter dumps logs on the standard error. ([#85351](https://github.com/kubernetes/kubernetes/pull/85351), [@rosti](https://github.com/rosti))

type parameters.
<code>D0[int, string]</code> and <code>struct{ x int; y string }</code> are
different because the former is an <a href="#Instantiations">instantiated</a>
defined type while the latter is a type literal
(but they are still <a href="#Assignability">assignable</a>).
</p>

<h3 id="Assignability">Assignability</h3>

<p>

* kubead now creates kube-proxy with a toleration to run on all nodes, no matter the taint. ([#62390](https://github.com/kubernetes/kubernetes/pull/62390), [@discordianfish](https://github.com/discordianfish))

- Graduating `pending_pods`, `preemption_attempts_total`, `preemption_victims` and `schedule_attempts_total` metrics to stable. Also `e2e_scheduling_duration_seconds` is renamed to `scheduling_attempt_duration_seconds` and the latter is graduated to stable. ([#105941](https://github.com/kubernetes/kubernetes/pull/105941), [@rezakrimi](https://github.com/rezakrimi)) [SIG Instrumentation, Scheduling and Testing]

  // Kubelet retries. Later errors (e.g. loading or validating a checkpointed config) will result in
  // a rollback to LastKnownGood. In the latter case, it is usually possible to resolve the error
  // by fixing the config assigned in Spec.ConfigSource.
  // You can find additional information for debugging by searching the error message in the Kubelet log.

    * Action required:
    * Anyone who generated server cert for admission webhooks need to regenerate the cert. Previously, when generating server cert for the admission webhook, the CN value doesn't matter. Now you must set it to the DNS name of the webhook service, i.e., `<service.Name>.<service.Namespace>.svc`.

    * Action required:
    * Anyone who generated server cert for admission webhooks need to regenerate the cert. Previously, when generating server cert for the admission webhook, the CN value doesn't matter. Now you must set it to the DNS name of the webhook service, i.e., `<service.Name>.<service.Namespace>.svc`.

* Add `--enable-admission-plugins` `--disable-admission-plugins` flags and deprecate `--admission-control`. When using the separate flag, the order in which they're specified doesn't matter.  ([#58123](https://github.com/kubernetes/kubernetes/pull/58123), [@hzxuzhonghu](https://github.com/hzxuzhonghu))