Class TpuClient (2.26.0)

Manages TPU nodes and other resources.

TPU API v1

Equality

Instances of this class created via copy-construction or copy-assignment always compare equal. Instances created with equal std::shared_ptr<*Connection> objects compare equal. Objects that compare equal share the same underlying resources.

Performance

Creating a new instance of this class is a relatively expensive operation, new objects establish new connections to the service. In contrast, copy-construction, move-construction, and the corresponding assignment operations are relatively efficient as the copies share all underlying resources.

Thread Safety

Concurrent access to different instances of this class, even if they compare equal, is guaranteed to work. Two or more threads operating on the same instance of this class is not guaranteed to work. Since copy-construction and move-construction is a relatively efficient operation, consider using such a copy when using this class from multiple threads.

Constructors

TpuClient(TpuClient const &)

Copy and move support

Parameter
Name Description
TpuClient const &

TpuClient(TpuClient &&)

Copy and move support

Parameter
Name Description
TpuClient &&

TpuClient(std::shared_ptr< TpuConnection >, Options)

Parameters
Name Description
connection std::shared_ptr< TpuConnection >
opts Options

Operators

operator=(TpuClient const &)

Copy and move support

Parameter
Name Description
TpuClient const &
Returns
Type Description
TpuClient &

operator=(TpuClient &&)

Copy and move support

Parameter
Name Description
TpuClient &&
Returns
Type Description
TpuClient &

Functions

ListNodes(std::string const &, Options)

Lists nodes.

Parameters
Name Description
parent std::string const &

Required. The parent resource name.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
StreamRange< google::cloud::tpu::v1::Node >

a StreamRange to iterate of the results. See the documentation of this type for details. In brief, this class has begin() and end() member functions returning a iterator class meeting the input iterator requirements. The value type for this iterator is a StatusOr as the iteration may fail even after some values are retrieved successfully, for example, if there is a network disconnect. An empty set of results does not indicate an error, it indicates that there are no resources meeting the request criteria. On a successful iteration the StatusOr<T> contains elements of type google.cloud.tpu.v1.Node, or rather, the C++ class generated by Protobuf from that type. Please consult the Protobuf documentation for details on the Protobuf mapping rules.

ListNodes(google::cloud::tpu::v1::ListNodesRequest, Options)

Lists nodes.

Parameters
Name Description
request google::cloud::tpu::v1::ListNodesRequest

Unary RPCs, such as the one wrapped by this function, receive a single request proto message which includes all the inputs for the RPC. In this case, the proto message is a google.cloud.tpu.v1.ListNodesRequest. Proto messages are converted to C++ classes by Protobuf, using the Protobuf mapping rules.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
StreamRange< google::cloud::tpu::v1::Node >

a StreamRange to iterate of the results. See the documentation of this type for details. In brief, this class has begin() and end() member functions returning a iterator class meeting the input iterator requirements. The value type for this iterator is a StatusOr as the iteration may fail even after some values are retrieved successfully, for example, if there is a network disconnect. An empty set of results does not indicate an error, it indicates that there are no resources meeting the request criteria. On a successful iteration the StatusOr<T> contains elements of type google.cloud.tpu.v1.Node, or rather, the C++ class generated by Protobuf from that type. Please consult the Protobuf documentation for details on the Protobuf mapping rules.

GetNode(std::string const &, Options)

Gets the details of a node.

Parameters
Name Description
name std::string const &

Required. The resource name.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
StatusOr< google::cloud::tpu::v1::Node >

the result of the RPC. The response message type (google.cloud.tpu.v1.Node) is mapped to a C++ class using the Protobuf mapping rules. If the request fails, the StatusOr contains the error details.

GetNode(google::cloud::tpu::v1::GetNodeRequest const &, Options)

Gets the details of a node.

Parameters
Name Description
request google::cloud::tpu::v1::GetNodeRequest const &

Unary RPCs, such as the one wrapped by this function, receive a single request proto message which includes all the inputs for the RPC. In this case, the proto message is a google.cloud.tpu.v1.GetNodeRequest. Proto messages are converted to C++ classes by Protobuf, using the Protobuf mapping rules.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
StatusOr< google::cloud::tpu::v1::Node >

the result of the RPC. The response message type (google.cloud.tpu.v1.Node) is mapped to a C++ class using the Protobuf mapping rules. If the request fails, the StatusOr contains the error details.

CreateNode(std::string const &, google::cloud::tpu::v1::Node const &, std::string const &, Options)

Creates a node.

Parameters
Name Description
parent std::string const &

Required. The parent resource name.

node google::cloud::tpu::v1::Node const &

Required. The node.

node_id std::string const &

The unqualified resource name.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
future< StatusOr< google::cloud::tpu::v1::Node > >

A future that becomes satisfied when the LRO (Long Running Operation) completes or the polling policy in effect for this call is exhausted. The future is satisfied with an error if the LRO completes with an error or the polling policy is exhausted. In this case the StatusOr returned by the future contains the error. If the LRO completes successfully the value of the future contains the LRO's result. For this RPC the result is a google.cloud.tpu.v1.Node proto message. The C++ class representing this message is created by Protobuf, using the Protobuf mapping rules.

CreateNode(ExperimentalTag, NoAwaitTag, std::string const &, google::cloud::tpu::v1::Node const &, std::string const &, Options)

Creates a node.

Specifying the NoAwaitTag immediately returns the [google::longrunning::Operation] that corresponds to the Long Running Operation that has been started. No polling for operation status occurs.

Parameters
Name Description
ExperimentalTag
NoAwaitTag
parent std::string const &
node google::cloud::tpu::v1::Node const &
node_id std::string const &
opts Options
Returns
Type Description
StatusOr< google::longrunning::Operation >

CreateNode(google::cloud::tpu::v1::CreateNodeRequest const &, Options)

Creates a node.

Parameters
Name Description
request google::cloud::tpu::v1::CreateNodeRequest const &

Unary RPCs, such as the one wrapped by this function, receive a single request proto message which includes all the inputs for the RPC. In this case, the proto message is a google.cloud.tpu.v1.CreateNodeRequest. Proto messages are converted to C++ classes by Protobuf, using the Protobuf mapping rules.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
future< StatusOr< google::cloud::tpu::v1::Node > >

A future that becomes satisfied when the LRO (Long Running Operation) completes or the polling policy in effect for this call is exhausted. The future is satisfied with an error if the LRO completes with an error or the polling policy is exhausted. In this case the StatusOr returned by the future contains the error. If the LRO completes successfully the value of the future contains the LRO's result. For this RPC the result is a google.cloud.tpu.v1.Node proto message. The C++ class representing this message is created by Protobuf, using the Protobuf mapping rules.

CreateNode(ExperimentalTag, NoAwaitTag, google::cloud::tpu::v1::CreateNodeRequest const &, Options)

Creates a node.

Specifying the NoAwaitTag immediately returns the [google::longrunning::Operation] that corresponds to the Long Running Operation that has been started. No polling for operation status occurs.

Parameters
Name Description
ExperimentalTag
NoAwaitTag
request google::cloud::tpu::v1::CreateNodeRequest const &
opts Options
Returns
Type Description
StatusOr< google::longrunning::Operation >

CreateNode(ExperimentalTag, google::longrunning::Operation const &, Options)

Creates a node.

This method accepts a google::longrunning::Operation that corresponds to a previously started Long Running Operation (LRO) and polls the status of the LRO in the background.

Parameters
Name Description
ExperimentalTag
operation google::longrunning::Operation const &
opts Options
Returns
Type Description
future< StatusOr< google::cloud::tpu::v1::Node > >

DeleteNode(std::string const &, Options)

Deletes a node.

Parameters
Name Description
name std::string const &

Required. The resource name.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
future< StatusOr< google::cloud::tpu::v1::Node > >

A future that becomes satisfied when the LRO (Long Running Operation) completes or the polling policy in effect for this call is exhausted. The future is satisfied with an error if the LRO completes with an error or the polling policy is exhausted. In this case the StatusOr returned by the future contains the error. If the LRO completes successfully the value of the future contains the LRO's result. For this RPC the result is a google.cloud.tpu.v1.Node proto message. The C++ class representing this message is created by Protobuf, using the Protobuf mapping rules.

DeleteNode(ExperimentalTag, NoAwaitTag, std::string const &, Options)

Deletes a node.

Specifying the NoAwaitTag immediately returns the [google::longrunning::Operation] that corresponds to the Long Running Operation that has been started. No polling for operation status occurs.

Parameters
Name Description
ExperimentalTag
NoAwaitTag
name std::string const &
opts Options
Returns
Type Description
StatusOr< google::longrunning::Operation >

DeleteNode(google::cloud::tpu::v1::DeleteNodeRequest const &, Options)

Deletes a node.

Parameters
Name Description
request google::cloud::tpu::v1::DeleteNodeRequest const &

Unary RPCs, such as the one wrapped by this function, receive a single request proto message which includes all the inputs for the RPC. In this case, the proto message is a google.cloud.tpu.v1.DeleteNodeRequest. Proto messages are converted to C++ classes by Protobuf, using the Protobuf mapping rules.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
future< StatusOr< google::cloud::tpu::v1::Node > >

A future that becomes satisfied when the LRO (Long Running Operation) completes or the polling policy in effect for this call is exhausted. The future is satisfied with an error if the LRO completes with an error or the polling policy is exhausted. In this case the StatusOr returned by the future contains the error. If the LRO completes successfully the value of the future contains the LRO's result. For this RPC the result is a google.cloud.tpu.v1.Node proto message. The C++ class representing this message is created by Protobuf, using the Protobuf mapping rules.

DeleteNode(ExperimentalTag, NoAwaitTag, google::cloud::tpu::v1::DeleteNodeRequest const &, Options)

Deletes a node.

Specifying the NoAwaitTag immediately returns the [google::longrunning::Operation] that corresponds to the Long Running Operation that has been started. No polling for operation status occurs.

Parameters
Name Description
ExperimentalTag
NoAwaitTag
request google::cloud::tpu::v1::DeleteNodeRequest const &
opts Options
Returns
Type Description
StatusOr< google::longrunning::Operation >

DeleteNode(ExperimentalTag, google::longrunning::Operation const &, Options)

Deletes a node.

This method accepts a google::longrunning::Operation that corresponds to a previously started Long Running Operation (LRO) and polls the status of the LRO in the background.

Parameters
Name Description
ExperimentalTag
operation google::longrunning::Operation const &
opts Options
Returns
Type Description
future< StatusOr< google::cloud::tpu::v1::Node > >

ReimageNode(google::cloud::tpu::v1::ReimageNodeRequest const &, Options)

Reimages a node's OS.

Parameters
Name Description
request google::cloud::tpu::v1::ReimageNodeRequest const &

Unary RPCs, such as the one wrapped by this function, receive a single request proto message which includes all the inputs for the RPC. In this case, the proto message is a google.cloud.tpu.v1.ReimageNodeRequest. Proto messages are converted to C++ classes by Protobuf, using the Protobuf mapping rules.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
future< StatusOr< google::cloud::tpu::v1::Node > >

A future that becomes satisfied when the LRO (Long Running Operation) completes or the polling policy in effect for this call is exhausted. The future is satisfied with an error if the LRO completes with an error or the polling policy is exhausted. In this case the StatusOr returned by the future contains the error. If the LRO completes successfully the value of the future contains the LRO's result. For this RPC the result is a google.cloud.tpu.v1.Node proto message. The C++ class representing this message is created by Protobuf, using the Protobuf mapping rules.

ReimageNode(ExperimentalTag, NoAwaitTag, google::cloud::tpu::v1::ReimageNodeRequest const &, Options)

Reimages a node's OS.

Specifying the NoAwaitTag immediately returns the [google::longrunning::Operation] that corresponds to the Long Running Operation that has been started. No polling for operation status occurs.

Parameters
Name Description
ExperimentalTag
NoAwaitTag
request google::cloud::tpu::v1::ReimageNodeRequest const &
opts Options
Returns
Type Description
StatusOr< google::longrunning::Operation >

ReimageNode(ExperimentalTag, google::longrunning::Operation const &, Options)

Reimages a node's OS.

This method accepts a google::longrunning::Operation that corresponds to a previously started Long Running Operation (LRO) and polls the status of the LRO in the background.

Parameters
Name Description
ExperimentalTag
operation google::longrunning::Operation const &
opts Options
Returns
Type Description
future< StatusOr< google::cloud::tpu::v1::Node > >

StopNode(google::cloud::tpu::v1::StopNodeRequest const &, Options)

Stops a node, this operation is only available with single TPU nodes.

Parameters
Name Description
request google::cloud::tpu::v1::StopNodeRequest const &

Unary RPCs, such as the one wrapped by this function, receive a single request proto message which includes all the inputs for the RPC. In this case, the proto message is a google.cloud.tpu.v1.StopNodeRequest. Proto messages are converted to C++ classes by Protobuf, using the Protobuf mapping rules.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
future< StatusOr< google::cloud::tpu::v1::Node > >

A future that becomes satisfied when the LRO (Long Running Operation) completes or the polling policy in effect for this call is exhausted. The future is satisfied with an error if the LRO completes with an error or the polling policy is exhausted. In this case the StatusOr returned by the future contains the error. If the LRO completes successfully the value of the future contains the LRO's result. For this RPC the result is a google.cloud.tpu.v1.Node proto message. The C++ class representing this message is created by Protobuf, using the Protobuf mapping rules.

StopNode(ExperimentalTag, NoAwaitTag, google::cloud::tpu::v1::StopNodeRequest const &, Options)

Stops a node, this operation is only available with single TPU nodes.

Specifying the NoAwaitTag immediately returns the [google::longrunning::Operation] that corresponds to the Long Running Operation that has been started. No polling for operation status occurs.

Parameters
Name Description
ExperimentalTag
NoAwaitTag
request google::cloud::tpu::v1::StopNodeRequest const &
opts Options
Returns
Type Description
StatusOr< google::longrunning::Operation >

StopNode(ExperimentalTag, google::longrunning::Operation const &, Options)

Stops a node, this operation is only available with single TPU nodes.

This method accepts a google::longrunning::Operation that corresponds to a previously started Long Running Operation (LRO) and polls the status of the LRO in the background.

Parameters
Name Description
ExperimentalTag
operation google::longrunning::Operation const &
opts Options
Returns
Type Description
future< StatusOr< google::cloud::tpu::v1::Node > >

StartNode(google::cloud::tpu::v1::StartNodeRequest const &, Options)

Starts a node.

Parameters
Name Description
request google::cloud::tpu::v1::StartNodeRequest const &

Unary RPCs, such as the one wrapped by this function, receive a single request proto message which includes all the inputs for the RPC. In this case, the proto message is a google.cloud.tpu.v1.StartNodeRequest. Proto messages are converted to C++ classes by Protobuf, using the Protobuf mapping rules.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
future< StatusOr< google::cloud::tpu::v1::Node > >

A future that becomes satisfied when the LRO (Long Running Operation) completes or the polling policy in effect for this call is exhausted. The future is satisfied with an error if the LRO completes with an error or the polling policy is exhausted. In this case the StatusOr returned by the future contains the error. If the LRO completes successfully the value of the future contains the LRO's result. For this RPC the result is a google.cloud.tpu.v1.Node proto message. The C++ class representing this message is created by Protobuf, using the Protobuf mapping rules.

StartNode(ExperimentalTag, NoAwaitTag, google::cloud::tpu::v1::StartNodeRequest const &, Options)

Starts a node.

Specifying the NoAwaitTag immediately returns the [google::longrunning::Operation] that corresponds to the Long Running Operation that has been started. No polling for operation status occurs.

Parameters
Name Description
ExperimentalTag
NoAwaitTag
request google::cloud::tpu::v1::StartNodeRequest const &
opts Options
Returns
Type Description
StatusOr< google::longrunning::Operation >

StartNode(ExperimentalTag, google::longrunning::Operation const &, Options)

Starts a node.

This method accepts a google::longrunning::Operation that corresponds to a previously started Long Running Operation (LRO) and polls the status of the LRO in the background.

Parameters
Name Description
ExperimentalTag
operation google::longrunning::Operation const &
opts Options
Returns
Type Description
future< StatusOr< google::cloud::tpu::v1::Node > >

ListTensorFlowVersions(std::string const &, Options)

List TensorFlow versions supported by this API.

Parameters
Name Description
parent std::string const &

Required. The parent resource name.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
StreamRange< google::cloud::tpu::v1::TensorFlowVersion >

a StreamRange to iterate of the results. See the documentation of this type for details. In brief, this class has begin() and end() member functions returning a iterator class meeting the input iterator requirements. The value type for this iterator is a StatusOr as the iteration may fail even after some values are retrieved successfully, for example, if there is a network disconnect. An empty set of results does not indicate an error, it indicates that there are no resources meeting the request criteria. On a successful iteration the StatusOr<T> contains elements of type google.cloud.tpu.v1.TensorFlowVersion, or rather, the C++ class generated by Protobuf from that type. Please consult the Protobuf documentation for details on the Protobuf mapping rules.

ListTensorFlowVersions(google::cloud::tpu::v1::ListTensorFlowVersionsRequest, Options)

List TensorFlow versions supported by this API.

Parameters
Name Description
request google::cloud::tpu::v1::ListTensorFlowVersionsRequest

Unary RPCs, such as the one wrapped by this function, receive a single request proto message which includes all the inputs for the RPC. In this case, the proto message is a google.cloud.tpu.v1.ListTensorFlowVersionsRequest. Proto messages are converted to C++ classes by Protobuf, using the Protobuf mapping rules.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
StreamRange< google::cloud::tpu::v1::TensorFlowVersion >

a StreamRange to iterate of the results. See the documentation of this type for details. In brief, this class has begin() and end() member functions returning a iterator class meeting the input iterator requirements. The value type for this iterator is a StatusOr as the iteration may fail even after some values are retrieved successfully, for example, if there is a network disconnect. An empty set of results does not indicate an error, it indicates that there are no resources meeting the request criteria. On a successful iteration the StatusOr<T> contains elements of type google.cloud.tpu.v1.TensorFlowVersion, or rather, the C++ class generated by Protobuf from that type. Please consult the Protobuf documentation for details on the Protobuf mapping rules.

GetTensorFlowVersion(std::string const &, Options)

Gets TensorFlow Version.

Parameters
Name Description
name std::string const &

Required. The resource name.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
StatusOr< google::cloud::tpu::v1::TensorFlowVersion >

the result of the RPC. The response message type (google.cloud.tpu.v1.TensorFlowVersion) is mapped to a C++ class using the Protobuf mapping rules. If the request fails, the StatusOr contains the error details.

GetTensorFlowVersion(google::cloud::tpu::v1::GetTensorFlowVersionRequest const &, Options)

Gets TensorFlow Version.

Parameters
Name Description
request google::cloud::tpu::v1::GetTensorFlowVersionRequest const &

Unary RPCs, such as the one wrapped by this function, receive a single request proto message which includes all the inputs for the RPC. In this case, the proto message is a google.cloud.tpu.v1.GetTensorFlowVersionRequest. Proto messages are converted to C++ classes by Protobuf, using the Protobuf mapping rules.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
StatusOr< google::cloud::tpu::v1::TensorFlowVersion >

the result of the RPC. The response message type (google.cloud.tpu.v1.TensorFlowVersion) is mapped to a C++ class using the Protobuf mapping rules. If the request fails, the StatusOr contains the error details.

ListAcceleratorTypes(std::string const &, Options)

Lists accelerator types supported by this API.

Parameters
Name Description
parent std::string const &

Required. The parent resource name.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
StreamRange< google::cloud::tpu::v1::AcceleratorType >

a StreamRange to iterate of the results. See the documentation of this type for details. In brief, this class has begin() and end() member functions returning a iterator class meeting the input iterator requirements. The value type for this iterator is a StatusOr as the iteration may fail even after some values are retrieved successfully, for example, if there is a network disconnect. An empty set of results does not indicate an error, it indicates that there are no resources meeting the request criteria. On a successful iteration the StatusOr<T> contains elements of type google.cloud.tpu.v1.AcceleratorType, or rather, the C++ class generated by Protobuf from that type. Please consult the Protobuf documentation for details on the Protobuf mapping rules.

ListAcceleratorTypes(google::cloud::tpu::v1::ListAcceleratorTypesRequest, Options)

Lists accelerator types supported by this API.

Parameters
Name Description
request google::cloud::tpu::v1::ListAcceleratorTypesRequest

Unary RPCs, such as the one wrapped by this function, receive a single request proto message which includes all the inputs for the RPC. In this case, the proto message is a google.cloud.tpu.v1.ListAcceleratorTypesRequest. Proto messages are converted to C++ classes by Protobuf, using the Protobuf mapping rules.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
StreamRange< google::cloud::tpu::v1::AcceleratorType >

a StreamRange to iterate of the results. See the documentation of this type for details. In brief, this class has begin() and end() member functions returning a iterator class meeting the input iterator requirements. The value type for this iterator is a StatusOr as the iteration may fail even after some values are retrieved successfully, for example, if there is a network disconnect. An empty set of results does not indicate an error, it indicates that there are no resources meeting the request criteria. On a successful iteration the StatusOr<T> contains elements of type google.cloud.tpu.v1.AcceleratorType, or rather, the C++ class generated by Protobuf from that type. Please consult the Protobuf documentation for details on the Protobuf mapping rules.

GetAcceleratorType(std::string const &, Options)

Gets AcceleratorType.

Parameters
Name Description
name std::string const &

Required. The resource name.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
StatusOr< google::cloud::tpu::v1::AcceleratorType >

the result of the RPC. The response message type (google.cloud.tpu.v1.AcceleratorType) is mapped to a C++ class using the Protobuf mapping rules. If the request fails, the StatusOr contains the error details.

GetAcceleratorType(google::cloud::tpu::v1::GetAcceleratorTypeRequest const &, Options)

Gets AcceleratorType.

Parameters
Name Description
request google::cloud::tpu::v1::GetAcceleratorTypeRequest const &

Unary RPCs, such as the one wrapped by this function, receive a single request proto message which includes all the inputs for the RPC. In this case, the proto message is a google.cloud.tpu.v1.GetAcceleratorTypeRequest. Proto messages are converted to C++ classes by Protobuf, using the Protobuf mapping rules.

opts Options

Optional. Override the class-level options, such as retry and backoff policies.

Returns
Type Description
StatusOr< google::cloud::tpu::v1::AcceleratorType >

the result of the RPC. The response message type (google.cloud.tpu.v1.AcceleratorType) is mapped to a C++ class using the Protobuf mapping rules. If the request fails, the StatusOr contains the error details.