- Sometimes, the scheduler incorrectly placed a pod in the "unschedulable" queue instead of the "backoff" queue. This happened when some plugin previously declared the pod as "unschedulable" and then in a later attempt encounters some other error. Scheduling of that pod then got delayed by up to five minutes, after which periodic flushing moved the pod back into the "active" queue. ([#120334](https://github.com/kubernetes/kubernetes/pull/120334), [@pohly...

    - [Certificates API](#certificates-api)
    - [Cluster Autoscaler](#cluster-autoscaler)
    - [Deployment](#deployment)
    - [Federation](#federation)
    - [Internal Storage Layer](#internal-storage-layer)
    - [Node Components](#node-components)
    - [kubectl](#kubectl)
    - [RBAC](#rbac)
    - [Scheduling](#scheduling)
    - [Service](#service)
    - [StatefulSet](#statefulset)

- Fixed a race condition in the iptables mode of kube-proxy in `1.27` and later
  that could result in some updates getting lost (e.g., when a service gets a
  new endpoint, the rules for the new endpoint might not be added until
  much later). ([#122204](https://github.com/kubernetes/kubernetes/pull/122204), [@danwinship](https://github.com/danwinship))

* etcd3: dereference the UID pointer for a readable error message. ([#33349](https://github.com/kubernetes/kubernetes/pull/33349), [@madhusudancs](https://github.com/madhusudancs))
* etcd3: pass SelectionPredicate instead of Filter to storage layer ([#31190](https://github.com/kubernetes/kubernetes/pull/31190), [@hongchaodeng](https://github.com/hongchaodeng))

- Support for go modules in Kubernetes Core.
- Continued preparation for cloud provider extraction and code organization.  The cloud provider code has been moved to kubernetes/legacy-cloud-providers for easier removal later and external consumption.
- Kubectl [get and describe](https://github.com/kubernetes/enhancements/issues/515) now works with extensions

[beta] User-provided apiservers can be aggregated (served along with) the rest of the Kubernetes API. See [Extending the Kubernetes API with the aggregation layer](https://kubernetes.io/docs/concepts/api-extension/apiserver-aggregation/), [Configure the aggregation layer](https://kubernetes.io/docs/tasks/access-kubernetes-api/configure-aggregation-layer/), and [Setup an extension API server](https://kubernetes.io/docs/tasks/access-kubernetes-api/setup-extension-api-server/).

* [alpha] Adding...

- Fixes a race condition in the iptables mode of kube-proxy in 1.27 and later
  that could result in some updates getting lost (e.g., when a service gets a
  new endpoint, the rules for the new endpoint might not be added until
  much later). ([#122756](https://github.com/kubernetes/kubernetes/pull/122756), [@hakman](https://github.com/hakman)) [SIG Network]

* Provide flags to use etcd3 backed storage ([#24455](https://github.com/kubernetes/kubernetes/pull/24455), [@hongchaodeng](https://github.com/hongchaodeng))

### Other notable changes

* Fix hyperkube's layer caching, and remove --make-symlinks at build time ([#25693](https://github.com/kubernetes/kubernetes/pull/25693), [@luxas](https://github.com/luxas))

- Sometimes, the scheduler incorrectly placed a pod in the "unschedulable" queue instead of the "backoff" queue. This happened when some plugin previously declared the pod as "unschedulable" and then in a later attempt encounters some other error. Scheduling of that pod then got delayed by up to five minutes, after which periodic flushing moved the pod back into the "active" queue. ([#120334](https://github.com/kubernetes/kubernetes/pull/120334), [@pohly...