- Fixed an EndpointSlice Controller hashing bug that could cause EndpointSlices to incorrectly handle Pods with duplicate IP addresses. For example this could happen when a new Pod reused an IP that was also assigned to a Pod in a completed state. ([#115907](https://github.com/kubernetes/kubernetes/pull/115907), [@qinqon](https://github.com/qinqon)) [SIG Apps and Network]

	MRFWorkerMinLimit = 2

	// MRFWorkerAutoDefault is default number of mrf workers for "auto" mode
	MRFWorkerAutoDefault = 4

	// LargeWorkerCount is default number of workers assigned to large uploads ( >= 128MiB)
	LargeWorkerCount = 10
)

// NewReplicationPool creates a pool of replication workers of specified size

    * 2. IPv4: api-server will assign an IP from a `service-cluster-ip-range` that is `ipv4` (either the primary or the secondary, according to how they were configured).
    * 2. IPv6: api-server will assign an IP from a `service-cluster-ip-range` that is `ipv6` (either the primary or the secondary, according to how they were configured).

* Add `metrics-port` to kube-proxy cmd flags. ([#72682](https://github.com/kubernetes/kubernetes/pull/72682), [@whypro](https://github.com/whypro))
* Adds deleting pods created by DaemonSet assigned to not existing nodes. ([#73401](https://github.com/kubernetes/kubernetes/pull/73401), [@krzysztof-jastrzebski](https://github.com/krzysztof-jastrzebski))

- Kube-proxy, will restart in case it detects that the Node assigned pod.Spec.PodCIDRs have changed ([#113247](https://github.com/kubernetes/kubernetes/pull/113247), [@code-elinka](https://github.com/code-elinka)) [SIG Cloud Provider, Network and Storage]

### Apps

- Added deleting pods created by `DaemonSet` assigned to not existing nodes. ([#73401](https://github.com/kubernetes/kubernetes/pull/73401), [@krzysztof-jastrzebski](https://github.com/krzysztof-jastrzebski))

  It changes the dual-stack API wrt Service from a single ipFamily field to 3
  fields: ipFamilyPolicy (SingleStack, PreferDualStack, RequireDualStack),
  ipFamilies (a list of families assigned), and clusterIPs (inclusive of
  clusterIP).  Most users do not need to set anything at all, defaulting will
  handle it for them.  Services are single-stack unless the user asks for

* Admit critical pods in the kubelet ([#38836](https://github.com/kubernetes/kubernetes/pull/38836), [@bprashanth](https://github.com/bprashanth))
* assign -998 as the oom_score_adj for critical pods (e.g. kube-proxy) ([#39114](https://github.com/kubernetes/kubernetes/pull/39114), [@dchen1107](https://github.com/dchen1107))

* Admit critical pods in the kubelet ([#38836](https://github.com/kubernetes/kubernetes/pull/38836), [@bprashanth](https://github.com/bprashanth))
* assign -998 as the oom_score_adj for critical pods (e.g. kube-proxy) ([#39114](https://github.com/kubernetes/kubernetes/pull/39114), [@dchen1107](https://github.com/dchen1107))

- HostNetwork pods no longer depend on the PodIPs to be assigned to configure the defined hostAliases on the Pod ([#126460](https://github.com/kubernetes/kubernetes/pull/126460), [@aojea](https://github.com/aojea)) [SIG Network, Node and Testing]