op.getLoc(), batch_variance.getType(), batch_variance,
          factor_const_op, /*broadcast_dimensions=*/DenseI64ArrayAttr());

      // Convert back to input type to stay aligned with expected output type
      // for TF op.
      y_out = rewriter.create<mhlo::ConvertOp>(op.getLoc(), y_out,
                                               input_element_type);

  // +optional
  optional string value = 2;
}

// PodExecOptions is the query options to a Pod's remote exec call.
// ---
// TODO: This is largely identical to PodAttachOptions above, make sure they stay in sync and see about merging
// and also when we cut V2, we should export a "StreamOptions" or somesuch that contains Stdin, Stdout, Stder and TTY
message PodExecOptions {

#### Continuous Improvement
- Project sustainability is not just about features. Many SIGs have been working on  improving test coverage, ensuring the basics stay reliable, and stability of the core feature set and working on maturing existing features and cleaning up the backlog.

#### Extensibility

	* Start recording cloud provider metrics for AWS ([#43477](https://github.com/kubernetes/kubernetes/pull/43477), [@gnufied](https://github.com/gnufied))

	* AWS: Batch DescribeInstance calls with nodeNames to 150 limit, to stay within AWS filter limits. ([#47516](https://github.com/kubernetes/kubernetes/pull/47516), [@gnufied](https://github.com/gnufied))

  * Advanced scheduling can be performed with [taints and tolerations](https://kubernetes.io/docs/concepts/configuration/assign-pod-node/#taints-and-tolerations-beta-feature)
* You can now specify (per pod) how long a pod should stay bound to a node, when there is a node problem.
* Various storage features have graduated to GA:
  * StorageClass pre-installed and set as default on Azure, AWS, GCE, OpenStack, and vSphere

  // +optional
  optional string value = 2;
}

// PodExecOptions is the query options to a Pod's remote exec call.
// ---
// TODO: This is largely identical to PodAttachOptions above, make sure they stay in sync and see about merging
// and also when we cut V2, we should export a "StreamOptions" or somesuch that contains Stdin, Stdout, Stder and TTY
message PodExecOptions {

	// SchedulingGates is an opaque list of values that if specified will block scheduling the pod.
	// If schedulingGates is not empty, the pod will stay in the SchedulingGated state and the
	// scheduler will not attempt to schedule the pod.
	//
	// SchedulingGates can only be set at pod creation time, and be removed only afterwards.
	//
	// +patchMergeKey=name

	// +optional
	OS *PodOS

	// SchedulingGates is an opaque list of values that if specified will block scheduling the pod.
	// If schedulingGates is not empty, the pod will stay in the SchedulingGated state and the
	// scheduler will not attempt to schedule the pod.
	//
	// SchedulingGates can only be set at pod creation time, and be removed only afterwards.
	//
	// +optional

	"schedulingGates":               "SchedulingGates is an opaque list of values that if specified will block scheduling the pod. If schedulingGates is not empty, the pod will stay in the SchedulingGated state and the scheduler will not attempt to schedule the pod.\n\nSchedulingGates can only be set at pod creation time, and be removed only afterwards.",

* AWS: Batch DescribeInstance calls with nodeNames to 150 limit, to stay within AWS filter limits. ([#47516](https://github.com/kubernetes/kubernetes/pull/47516), [@gnufied](https://github.com/gnufied))