- Fix missing cadvisor machine metrics. ([#97006](https://github.com/kubernetes/kubernetes/pull/97006), [@lingsamuel](https://github.com/lingsamuel)) [SIG Node]
- Fix: azure file latency issue for metadata-heavy workloads ([#97082](https://github.com/kubernetes/kubernetes/pull/97082), [@andyzhangx](https://github.com/andyzhangx)) [SIG Cloud Provider and Storage]

*   [beta] Flag `--chunk-size={SIZE}` is added to `kubectl get` to customize the number of results returned in large lists of resources. This reduces the perceived latency of managing large clusters because the server returns the first set of results to the client much more quickly. Pass 0 to disable this feature.([#53768](https://github.com/kubernetes/kubernetes/pull/53768),[ @smarterclayton](https://github.com/smarterclayton))...

	want := opts.TraceTypes()
	if !want.Contains(trcInfo.TraceType) {
		return false
	}

	isHTTP := trcInfo.TraceType.Overlaps(madmin.TraceInternal|madmin.TraceS3) && trcInfo.HTTP != nil

	// Check latency...
	if opts.Threshold > 0 && trcInfo.Duration < opts.Threshold {
		return false
	}

	// Check internal path
	isInternal := isHTTP && HasPrefix(trcInfo.HTTP.ReqInfo.Path, minioReservedBucketPath+SlashSeparator)

### Bug or Regression

- An inefficient lock in EndpointSlice controller metrics cache has been reworked. Network programming latency may be significantly reduced in certain scenarios, especially in clusters with a large number of Services. ([#107169](https://github.com/kubernetes/kubernetes/pull/107169), [@robscott](https://github.com/robscott)) [SIG Apps and Network]

## External Dependencies

- Resolves a performance regression in default 1.31+ configurations, related to the ConsistentListFromCache feature, where rapid create / update API requests across different namespaces encounter increased latency. ([#130156](https://github.com/kubernetes/kubernetes/pull/130156), [@AwesomePatrol](https://github.com/AwesomePatrol)) [SIG API Machinery]

### Other (Cleanup or Flake)

			// use target endpoint metrics from node which has been up the longest
			if v2.LastOnline.After(v.LastOnline) || v2.LastOnline.IsZero() {
				v2.Endpoint = v.Endpoint
				v2.LastOnline = v.LastOnline
				v2.Latency = v.Latency
				v2.Online = v.Online
				v2.TotalDowntime = v.TotalDowntime
				v2.DeploymentID = v.DeploymentID
			}
			v2.ReplicatedCount += v.ReplicatedCount
			v2.ReplicatedSize += v.ReplicatedSize

- Fixed a bug where high latency `kube-apiserver` caused scheduling throughput degradation. ([#134154](https://github.com/kubernetes/kubernetes/pull/134154), [@macsko](https://github.com/macsko))

	smallFileThreshold = 128 * humanize.KiByte // Optimized for NVMe/SSDs

	// For hardrives it is possible to set this to a lower value to avoid any
	// spike in latency. But currently we are simply keeping it optimal for SSDs.

	// bigFileThreshold is the point where we add readahead to put operations.
	bigFileThreshold = 128 * humanize.MiByte

- Reverted regression that prevented `client-go` latency metrics to be reported with a template URL to avoid label cardinality. ([#111752](https://github.com/kubernetes/kubernetes/pull/111752), [@aanm](https://github.com/aanm))