- During device plugin allocation, resources requested by the pod can only be allocated if the device plugin has registered itself to kubelet AND healthy devices are present on the node to be allocated. If these conditions are not sattsfied, the pod would fail with `UnexpectedAdmissionError` error. ([#116376](https://github.com/kuber...

                            int num_underlying_devices,
                            TFE_CustomDevice* device, void** device_info) {
  device->copy_tensor_to_device = &CopyToParallelDevice;
  device->copy_tensor_from_device = &CopyTensorFromParallelDevice;
  device->delete_device = &DeleteParallelDevice;
  device->execute = &ParallelDeviceExecute;
  std::vector<std::string> underlying_devices_vector;

  virtual const string& Name() const = 0;

  // Returns the operation's device name.
  //
  // The value returned may be different from the one set by SetDeviceName, but
  // it will be compatible with it: the name will be updated by device placement
  // logic to refer to the specific device chosen.
  //
  // Example: If one calls `op->SetDeviceName("/device:GPU")`, the value

- [IPv4/IPv6 dual-stack](https://kubernetes.io/docs/concepts/services-networking/dual-stack) enables the allocation of both IPv4 and IPv6 addresses to Pods and Services.
- [API Server Network Proxy](https://github.com/kubernetes/enhancements/blob/master/keps/sig-api-machinery/1281-network-proxy/README.md) going alpha in 1.16.

==============================================================================*/

#ifndef TENSORFLOW_C_EAGER_CUSTOM_DEVICE_TESTUTIL_H_
#define TENSORFLOW_C_EAGER_CUSTOM_DEVICE_TESTUTIL_H_

// A simple logging device to test custom device registration.
#include <memory>

#include "tensorflow/c/c_api.h"
#include "tensorflow/c/eager/c_api.h"
#include "tensorflow/c/eager/c_api_experimental.h"
#include "tensorflow/c/tf_status.h"

  ASSERT_EQ(TF_OK, TF_GetCode(status)) << TF_Message(status);

  // The device of var_handle0 is local device which is the same before and
  // after cluster update. Remove resource with valid device should succeed.
  TFE_Op* op = TFE_NewOp(ctx, "DestroyResourceOp", status);
  ASSERT_EQ(TF_OK, TF_GetCode(status)) << TF_Message(status);
  TFE_OpAddInput(op, var_handle0, status);

  ASSERT_EQ(TF_OK, TF_GetCode(status)) << TF_Message(status);
  const DLTensor* dltensor_out = &dlm_out->dl_tensor;
  EXPECT_EQ(dltensor_out->device.device_type, dltensor_in->device.device_type);
  EXPECT_EQ(dltensor_out->device.device_id, dltensor_in->device.device_id);
  EXPECT_EQ(dltensor_out->ndim, dltensor_in->ndim);
  EXPECT_EQ(dltensor_out->dtype.code, dltensor_in->dtype.code);

  repeated k8s.io.api.admissionregistration.v1.RuleWithOperations rules = 3;

  // FailurePolicy defines how unrecognized errors from the admission endpoint are handled -
  // allowed values are Ignore or Fail. Defaults to Ignore.
  // +optional
  optional string failurePolicy = 4;

  // matchPolicy defines how the "rules" list is used to match incoming requests.
  // Allowed values are "Exact" or "Equivalent".
  //

- Until this release, the iscsi plugin was waiting 10 seconds for a path to appear in the device list. However this timeout is not enough, or is less than the default device discovery timeout in most systems, which prevents certain devices from being discovered. This timeout has been raised to 30 seconds, which should help to avoid mount issues due to device discovery. ([#78475](https://github.com/kubernetes/kubernetes/pull/78475), [@humblec](https://github.com/humblec))...

-   type: input id: OS attributes: label: OS Platform and Distribution
    description: placeholder: e.g., Linux Ubuntu 16.04 validations: required:
    false

-   type: input id: Mobile attributes: label: Mobile device description:
    placeholder: e.g., Linux Ubuntu 16.04 validations: required: false

-   type: input id: Python attributes: label: Python version description:
    placeholder: e.g., 3.9 validations: required: false