Add documentation for LogarithmicScaleDown featuregate

content/en/docs/concepts/workloads/controllers/replicaset.md

@@ -317,7 +317,8 @@ prioritize scaling down pods based on the following general algorithm:
     the pod with the lower value will come first.
  3. Pods on nodes with more replicas come before pods on nodes with fewer replicas.
  4. If the pods' creation times differ, the pod that was created more recently
-    comes before the older pod (the creation times are bucketed on an integer log scale)
+    comes before the older pod (the creation times are bucketed on an integer log scale
+    when the `LogarithmicScaleDown` [feature gate](/docs/reference/command-line-tools-reference/feature-gates/) is enabled)
     
 If all of the above match, then selection is random.
 

content/en/docs/reference/command-line-tools-reference/feature-gates.md

@@ -135,6 +135,7 @@ different Kubernetes components.
 | `LocalStorageCapacityIsolation` | `false` | Alpha | 1.7 | 1.9 |
 | `LocalStorageCapacityIsolation` | `true` | Beta | 1.10 | |
 | `LocalStorageCapacityIsolationFSQuotaMonitoring` | `false` | Alpha | 1.15 | |
+| `LogarithmicScaleDown` | `false` | Alpha | 1.21 | |
 | `MixedProtocolLBService` | `false` | Alpha | 1.20 | |
 | `NamespaceDefaultLabelName` | `true` | Beta | 1.21 | |
 | `NetworkPolicyEndPort` | `false` | Alpha | 1.21 |  |
@@ -670,6 +671,8 @@ Each feature gate is designed for enabling/disabling a specific feature:
   supports project quotas and they are enabled, use project quotas to monitor
   [emptyDir volume](/docs/concepts/storage/volumes/#emptydir) storage consumption rather than
   filesystem walk for better performance and accuracy.
+- `LogarithmicScaleDown`: Enable semi-random selection of pods to evict on controller scaledown
+  based on logarithmic bucketing of pod timestamps.
 - `MixedProtocolLBService`: Enable using different protocols in the same `LoadBalancer` type
   Service instance.
 - `MountContainers` (*deprecated*): Enable using utility containers on host as