*  Fix: Don't crash after a web socket fails its connection upgrade. We incorrectly released
    the web socket's connections back to the pool before their resources were cleaned up.

 *  Fix: Don't infinite loop when a received web socket message has self-terminating compressed
    data.

 *  Fix: Don't fail the call when the response code is ‘HTTP 102 Processing’ or ‘HTTP 103 Early
    Hints’.

- **Stateful Application Support**
  - Enhanced persistence capabilities (eg: `StorageClasses`, new volume plugins)
  - New resources and scheduler features (eg: `ScheduledJob` resource, pod/node affinity/anti-affinity)
- **Cluster Federation**
  - Global Multi-cluster HTTP(S) Ingress across GCE and GKE clusters.
  - Expanded support for federated hybrid-cloud resources including ReplicaSets, Secrets, Namespaces and Events.

- Changed the error message displayed when a pod is trying to attach a volume that does not match the label/selector from "x node(s) had volume node affinity conflict" to "x node(s) didn't match PersistentVolume's node affinity". ([#129887](https://github.com/kubernetes/kubernetes/pull/129887), [@rhrmo](https://github.com/rhrmo)) [SIG Scheduling and Storage]

    filesystem.deleteRecursively(cacheDir)
    assertThat(cache["a"]).isNull()
  }

  /**
   * We had a long-lived bug where [DiskLruCache.trimToSize] could infinite loop if entries
   * being edited required deletion for the operation to complete.
   */
  @Test
  fun trimToSizeWithActiveEdit() {
    val expectedByteCount = if (windows) 10L else 0L

    /** Group cache expiry time in seconds. */
    protected long groupCacheExpiry = 10 * 60L;

    /** Maximum depth for processing nested groups to prevent infinite loops. */
    protected int maxGroupDepth = 10;

    /** Use V2 endpoint. */
    protected boolean useV2Endpoint = true;

    /**
     * Initializes the Entra ID authenticator.

* Cloud Controller Manager now sets Node.Spec.ProviderID ([#50730](https://github.com/kubernetes/kubernetes/pull/50730), [@andrewsykim](https://github.com/andrewsykim))
* Parameterize session affinity timeout seconds in service API for Client IP based session affinity. ([#49850](https://github.com/kubernetes/kubernetes/pull/49850), [@m1093782566](https://github.com/m1093782566))

- Fixed 'kubeadm upgrade' infinite loop waiting for pod restart ([#69886](https://github.com/kubernetes/kubernetes/pull/69886), [@bart0sh](https://github.com/bart0sh))

{min-height:1.2em}.placeholder-glow .placeholder{animation:placeholder-glow 2s ease-in-out infinite}@keyframes placeholder-glow{50%{opacity:.2}}.placeholder-wave{-webkit-mask-image:linear-gradient(130deg,#000 55%,rgba(0,0,0,0.8) 75%,#000 95%);mask-image:linear-gradient(130deg,#000 55%,rgba(0,0,0,0.8) 75%,#000 95%);-webkit-mask-size:200% 100%;mask-size:200% 100%;animation:placeholder-wave 2s linear infinite}@keyframes placeholder-wave{100%{-webkit-mask-position:-200% 0%;mask-position:-200% 0%}}.c...

- The calculations for Pod topology spread skew now exclude nodes that
  don't match the node affinity/selector. This may lead to unschedulable pods if you previously had pods
  matching the spreading selector on those excluded nodes (not matching the node affinity/selector),
  especially when the `topologyKey` is not node-level. Revisit the node affinity and/or pod selector in the

    assertThat(response3.body.string()).isEqualTo("def")
    assertThat(server.takeRequest().exchangeIndex).isEqualTo(1) // New connection.
  }

  /**
   * We had a bug where we'd perform infinite retries of route that fail with connection shutdown
   * errors. The problem was that the logic that decided whether to reuse a route didn't track
   * certain HTTP/2 errors. https://github.com/square/okhttp/issues/5547
   */