nvflare.private.fed.client.client_executor module

class ClientExecutor(uid, startup)

Bases: object

To init the ClientExecutor.

Parameters

• uid – client name

• startup – startup folder

abort_task(client, job_id)

To abort the client executing task.

Parameters

• client – the FL client object

• job_id – the job_id

abort_train(client, job_id)

To abort the client training.

Parameters

• client – the FL client object

• job_id – the job_id

check_status(client, job_id) → str

To check the status of the running client.

Parameters

• client – the FL client object

• job_id – the job_id

Returns

A client status message

cleanup()

Cleanup.

get_errors(job_id)

Get the error information.

Returns

A dict of error information.

get_run_info(job_id) → dict

Get the run information.

Parameters

job_id – the job_id

Returns

A dict of run information.

reset_errors(job_id)

Reset the error information.

send_aux_command(shareable: Shareable, job_id)

To send the aux command to child process.

Parameters

• shareable – aux message Shareable

• job_id – the job_id

start_train(client, job_id, args, app_root, app_custom_folder, listen_port, allocated_resource, token, resource_consumer, resource_manager, target: str)

start_train method to start the FL client training.

Parameters

• client – the FL client object

• job_id – the job_id

• args – admin command arguments for starting the FL client training

• app_root – the root folder of the running APP

• app_custom_folder – FL application custom folder

• listen_port – port to listen the command.

• allocated_resource – allocated resources

• token – token from resource manager

• resource_consumer – resource consumer

• resource_manager – resource manager

• target – SP target location

class ProcessExecutor(uid, startup)

Bases: ClientExecutor

Run the Client executor in a child process.

To init the ProcessExecutor.

Parameters

• uid – client name

• startup – startup folder

abort_task(client, job_id)

To abort the client executing task.

Parameters

• client – the FL client object

• job_id – the job_id

abort_train(client, job_id)

To abort the client training.

Parameters

• client – the FL client object

• job_id – the job_id

check_status(client, job_id)

To check the status of the running client.

Parameters

• client – the FL client object

• job_id – the job_id

Returns

A client status message

close()

get_conn_client(job_id)

get_errors(job_id)

Get the error information.

Returns

A dict of error information.

get_run_info(job_id)

Get the run information.

Parameters

job_id – the job_id

Returns

A dict of run information.

get_run_processes_keys()

get_status(job_id)

reset_errors(job_id)

Reset the error information.

send_aux_command(shareable: Shareable, job_id)

To send the aux command to child process.

Parameters

• shareable – aux message Shareable

• job_id – the job_id

start_train(client, job_id, args, app_root, app_custom_folder, listen_port, allocated_resource, token, resource_consumer: ResourceConsumerSpec, resource_manager: ResourceManagerSpec, target: str)

start_train method to start the FL client training.

Parameters

• client – the FL client object

• job_id – the job_id

• args – admin command arguments for starting the FL client training

• app_root – the root folder of the running APP

• app_custom_folder – FL application custom folder

• listen_port – port to listen the command.

• allocated_resource – allocated resources

• token – token from resource manager

• resource_consumer – resource consumer

• resource_manager – resource manager

• target – SP target location