- The `flowcontrol.apiserver.k8s.io/v1beta3` `FlowSchema` and `PriorityLevelConfiguration` APIs has been promoted to `flowcontrol.apiserver.k8s.io/v1`, with the following changes:

* Enable DefaultTolerationSeconds admission controller by default ([#41815](https://github.com/kubernetes/kubernetes/pull/41815), [@kevin-wangzefeng](https://github.com/kevin-wangzefeng))
* ResourceQuota ability to support default limited resources ([#36765](https://github.com/kubernetes/kubernetes/pull/36765), [@derekwaynecarr](https://github.com/derekwaynecarr))

- Implemented logging for failed transactions and the full table in `kube-proxy` with 
`nftables` when using log level 4 or higher. Logging is rate-limited to one entry every 24 hours 
to avoid performance issues. ([#128886](https://github.com/kubernetes/kubernetes/pull/128886), [@npinaeva](https://github.com/npinaeva))
- Implemented the `scheduler_cache_size` metric.

  - Please adapt your workloads to tolerate the same future taint preemptively.
    
- Kubeadm: improve the validation of serviceSubnet and podSubnet.
  ServiceSubnet has to be limited in size, due to implementation details, and the mask can not allocate more than 20 bits.

#### **Other**

*   Scheduler cache ignores updates to an assumed pod if updates are limited to pod annotations.  ([#54008](https://github.com/kubernetes/kubernetes/pull/54008),[ @yguo0905](https://github.com/yguo0905))

    limiter.acquire(Integer.MAX_VALUE / 4);
    limiter.acquire(Integer.MAX_VALUE / 2);
    limiter.acquire(Integer.MAX_VALUE);
    assertEvents("R0.00", "R0.00", "R0.00");

    limiter.setRate(1.0);
    limiter.acquire();
    limiter.acquire();
    limiter.acquire();
    assertEvents("R0.00", "R1.00", "R1.00");

    limiter.acquire(Integer.MAX_VALUE / 4);
    limiter.acquire(Integer.MAX_VALUE / 2);
    limiter.acquire(Integer.MAX_VALUE);
    assertEvents("R0.00", "R0.00", "R0.00");

    limiter.setRate(1.0);
    limiter.acquire();
    limiter.acquire();
    limiter.acquire();
    assertEvents("R0.00", "R1.00", "R1.00");

- Dynamic Resource Allocation (DRA): Added a feature so the number of ResourceClaim objects can be limited per namespace and by the number of devices requested through a specific class via the v1.ResourceQuota mechanism. ([#120611](https://github.com/kubernetes/kubernetes/pull/120611), [@pohly](https://github.com/pohly)) [SIG API Machinery, Apps, Auth,...

   *
   * Visited Nodes = {}
   * EndpointPair [N1, N2] - return
   * EndpointPair [N1, N3] - return
   * Visited Nodes = {N1}
   * EndpointPair [N2, N1] - skip
   * EndpointPair [N2, N3] - return
   * Visited Nodes = {N1, N2}
   * EndpointPair [N3, N1] - skip
   * EndpointPair [N3, N2] - skip
   * Visited Nodes = {N1, N2, N3}
   * EndpointPair [N4, N4] - return
   * Visited Nodes = {N1, N2, N3, N4}

  worked with the EndpointSliceController.) ([#91399](https://github.com/kubernetes/kubernetes/pull/91399), [@danwinship](https://github.com/danwinship)) [SIG Apps and Network]
- Fixes a bug evicting pods after a taint with a limited tolerationSeconds toleration is removed from a node ([#93722](https://github.com/kubernetes/kubernetes/pull/93722), [@liggitt](https://github.com/liggitt)) [SIG Apps and Node]