assertThat(peer.takeFrame().type).isEqualTo(Http2.TYPE_RST_STREAM)
  }

  /**
   * Confirm that the client times out if the server stalls after 3 bytes. After the timeout the
   * connection is still considered healthy while we await the degraded pong. When that doesn't
   * arrive the connection goes unhealthy.
   */
  @Test fun readTimesOut() {
    // Write the mocking script.

* [scheduler] Fix issue new pod with affinity stuck at `creating` because node had been deleted but its pod still exists. ([#53647](https://github.com/kubernetes/kubernetes/pull/53647), [@wenlxie](https://github.com/wenlxie))

* The error handling behavior of `pkg/client/restclient.Result` has changed.  Calls to `Result.Raw()` will no longer parse the body, although they will still return errors that react to `pkg/api/errors.Is*()` as in previous releases.  Callers of `Get()` and `Into()` will continue to receive errors that are parsed from the body if the kind and apiVersion of the body match the `Status` object. ([#36001](...

   ([#86139](https://github.com/kubernetes/kubernetes/pull/86139), [@jasimmons](https://github.com/jasimmons))
- 'Graduated `JobTrackingWithFinalizers` to stable.
  Jobs created before the feature was enabled are still tracked without finalizers.
  Jobs tracked with finalizers have the annotation batch.kubernetes.io/job-tracking.
  If the annotation is present and the user attempts to remove it, the control plane adds it back.

    Node* node = graph->graph.FindNodeId(*pos);
    // FindNodeId() returns nullptr for nodes that have been deleted.
    // We aren't currently allowing nodes to be deleted, but it is safer
    // to still check.
    if (node != nullptr) return ToOperation(node);
    *pos += 1;
  }

  // No more nodes.
  return nullptr;
}

void TF_GraphToGraphDef(TF_Graph* graph, TF_Buffer* output_graph_def,

Traditionally, the Kubernetes scheduler was based on the assumption that additional persistent storage is available everywhere in the cluster and has infinite capacity. Topology constraints addressed the first point, but up to now pod scheduling was still done without considering that the remaining storage capacity may not be enough to start a new pod. [Storage capacity tracking](https://github.com/kubernetes/enhancements/tree/master/keps/sig-storage/1472-storage-capacity-tracking), a new alpha feature,...

      return removeFromChain(castForTesting(first), castForTesting(entry));
    }

    /**
     * Unsafely returns the value of the given entry if it's still live, or {@code null} otherwise.
     */
    @Nullable V getLiveValueForTesting(InternalEntry<K, V, ?> entry) {
      return getLiveValue(castForTesting(entry));
    }

- Omit nil or empty field when calculating container hash value to avoid hash changed. For a new field with a non-nil default value in the container spec, the hash would still get changed. ([#57741](https://github.com/kubernetes/kubernetes/pull/57741), [@dixudx](https://github.com/dixudx))

    * kubeadm upgrade can now rollback and restore etcd after an upgrade failure
* Update version of Istio addon from 0.5.1 to 0.6.0. ([#61911](https://github.com/kubernetes/kubernetes/pull/61911), [@ostromart](https://github.com/ostromart))
    * See https://istio.io/news/releases/0.x/announcing-0.6/ for full Isto release notes.

	}
	return pinfo.Index, nil
}

// getPoolIdxExistingNoLock returns the (first) found object pool index containing an object.
// If the object exists, but the latest version is a delete marker, the index with it is still returned.
// If the object does not exist ObjectNotFound error is returned.
// If any other error is found, it is returned.
// The check is skipped if there is only one pool, and 0, nil is always returned in that case.