Release artifacts for 1.18.7 and 1.17.10 are incomplete. __Do not use these releases.__

# Changelog since v1.18.6

## Changes by Kind

### Bug or Regression

- Do not add nodes labeled with kubernetes.azure.com/managed=false to backend pool of load balancer. ([#93034](https://github.com/kubernetes/kubernetes/pull/93034), [@matthias50](https://github.com/matthias50)) [SIG Cloud Provider]

Function1 implements scala.Serializable { public void ConfigurationParser$$anonfun$4(ConfigurationParser); } xsbt/boot/Labeled.class package xsbt.boot; public final synchronized class Labeled implements scala.collection.IterableLike { private final String label; private final scala.Option value; public final String label(); public final scala.Option value(); public void Labeled(String, scala.Option); } xsbt/boot/CheckProxy$.class package xsbt.boot; public final synchronized class CheckProxy$ { public...

- Graduated the `LegacyServiceAccountTokenTracking` feature gate to GA. The usage of auto-generated secret-based service account token now produces warnings, and relevant Secrets are labeled with a last-used timestamp (label key `kubernetes.io/legacy-token-last-used`). ([#117591](https://github.com/kubernetes/kubernetes/pull/117591), [@zshihang](https://github.com/zshihang)) [SIG API Machinery, Auth and Testing]

- Cordoned nodes are now deregistered from AWS target groups. ([#85920](https://github.com/kubernetes/kubernetes/pull/85920), [@hoelzro](https://github.com/hoelzro)) [SIG Cloud Provider]
- Do not add nodes labeled with kubernetes.azure.com/managed=false to backend pool of load balancer. ([#93034](https://github.com/kubernetes/kubernetes/pull/93034), [@matthias50](https://github.com/matthias50)) [SIG Cloud Provider]

        to a labeled dataset, in one function call. Note that it doesn't perform
        image data augmentation (which is meant to be done using preprocessing
        layers).
    *   **[`text_dataset_from_directory`](https://www.tensorflow.org/api_docs/python/tf/keras/preprocessing/text_dataset_from_directory)**
        takes you from a structured directory of text files to a labeled
        dataset, in one function call.

      </match>
    </magic>
  </mime-type>
  <mime-type type="application/vnd.wqd">
    <glob pattern="*.wqd"/>
  </mime-type>
  <mime-type type="application/vnd.wrq-hp3000-labelled"/>
  <mime-type type="application/vnd.wt.stf">
    <glob pattern="*.stf"/>
  </mime-type>
  <mime-type type="application/vnd.wv.csp+wbxml"/>
  <mime-type type="application/vnd.wv.csp+xml"/>

* Replace *_admission_latencies_milliseconds_summary and *_admission_latencies_milliseconds metrics due to reporting wrong unit (was labelled milliseconds, but reported seconds), and multiple naming guideline violations (units should be in base units and "duration" is the best practice labelling to measure the time a request takes). Please convert to use *_admission_duration_seconds and...

* Fix problems with scaling up the cluster when unschedulable pods have some persistent volume claims. ([#44860](https://github.com/kubernetes/kubernetes/pull/44860), [@mwielgus](https://github.com/mwielgus))
* Exclude nodes labeled as master from LoadBalancer / NodePort; restores documented behaviour. ([#44745](https://github.com/kubernetes/kubernetes/pull/44745), [@justinsb](https://github.com/justinsb))

## Changelog since v1.16.13

## Changes by Kind

### Bug or Regression

- Do not add nodes labeled with kubernetes.azure.com/managed=false to backend pool of load balancer. ([#93034](https://github.com/kubernetes/kubernetes/pull/93034), [@matthias50](https://github.com/matthias50)) [SIG Cloud Provider]

- Do not add nodes labeled with kubernetes.azure.com/managed=false to backend pool of load balancer. ([#93034](https://github.com/kubernetes/kubernetes/pull/93034), [@matthias50](https://github.com/matthias50)) [SIG Cloud Provider]