* Don't recreate lb cloud resources on kcm restart ([#29082](https://github.com/kubernetes/kubernetes/pull/29082), [@bprashanth](https://github.com/bprashanth))
* List all nodes and occupy cidr map before starting allocations ([#29062](https://github.com/kubernetes/kubernetes/pull/29062), [@bprashanth](https://github.com/bprashanth))

- Reduced memory usage/allocations during wait for volume attachment. ([#126575](https://github.com/kubernetes/kubernetes/pull/126575), [@Lucaber](https://github.com/Lucaber)) [SIG Node and Storage]

  - Added scheduler-owned field in `ResourceClaim.Status` to specify how much device capacity is reserved for a specific request.
  - Added an additional identifier to `ResourceClaim.Status` for the device supports multiple allocations.

- Introduced a breaking change to the `resource.k8s.io` API in its `AllocationResult` struct. This change allows a kubelet plugin for the `DynamicResourceAllocation` feature to service allocations from multiple resource driver controllers. ([#116332](https://github.com/kubernetes/kubernetes/pull/116332), [@klueska](https://github.com/klueska))

 * Improvements and fixes in Keras loss masking:

### Serverside Apply - Beta 2

- Fixed bug in `TopologyManager` for ensuring aligned allocations on machines with more than 2 NUMA nodes ([#108052](https://github.com/kubernetes/kubernetes/pull/108052), [@klueska](https://github.com/klueska))

- DRA: when the scheduler had to deallocate a claim after a node became unsuitable for a pod, it might have needed more attempts than really necessary. This was fixed by first disabling allocations. ([#120428](https://github.com/kubernetes/kubernetes/pull/120428), [@pohly](https://github.com/pohly))

    * and cloud_cidr_allocator. (ALPHA FEATURE)
    * Note: for this change, the other allocators still exist and are the default.
    * It supports two modes:
        * CIDR range allocations done within the cluster that are then propagated out to the cloud provider.
        * Cloud provider managed IPAM that is then reflected into the cluster.

    * and cloud_cidr_allocator. (ALPHA FEATURE)
    * Note: for this change, the other allocators still exist and are the default.
    * It supports two modes:
        * CIDR range allocations done within the cluster that are then propagated out to the cloud provider.
        * Cloud provider managed IPAM that is then reflected into the cluster.