nvflare.app_common.executors.client_api.external_process_backend module
External-process backend for ClientAPIExecutor.
The backend owns the launched trainer process/process group and its authenticated session with the
Client Job (CJ) Cell. Task and result Shareables use Cell/F3 directly, including lazy payload transfer. See
docs/design/client_api_execution_modes.md and the trainer counterpart in
nvflare/client/cell/api.py.
- class ExternalProcessBackend[source]
Bases:
CellBackendBaseLaunches and owns the external trainer process/group, bridged over the CJ cell.
- abort(fl_ctx: FLContext) None[source]
Notify the trainer of task cancellation and latch abortive run teardown.
An accepted lazy result can outlive
execute()while another client downloads it. Explicit run abort therefore carries a context marker so it remains distinguishable from normal END_RUN and CCWF task cancellation. Callers predating the marker retain the original run-abort behavior.
- execute(task_name: str, shareable: Shareable, fl_ctx: FLContext, abort_signal: Signal) Shareable[source]
Executes one task on the trainer and returns its result.
The backend delivers the task to the trainer (TASK_READY over Cell, or DataBus for in_process), waits for the result within the executor-configured task/result bounds, and returns the result Shareable.
Contract: this method must always return a Shareable and must not hang past abort: when abort_signal is triggered, the backend notifies/stops the trainer per its mode’s lifecycle ownership and returns
make_reply(ReturnCode.TASK_ABORTED). On failure it should return an error reply (e.g. ReturnCode.EXECUTION_EXCEPTION) rather than raise; exceptions that do escape are converted to EXECUTION_EXCEPTION replies by the executor, except UnsafeJobError which the executor lets propagate so ClientRunner can apply its dedicated UNSAFE_JOB handling.- Parameters:
task_name – name of the task.
shareable – the task data.
fl_ctx – the FLContext of the task.
abort_signal – checked during execution; triggered means the task is aborted.
- Returns:
The result Shareable (an error reply on failure/abort - never None).
- finalize(fl_ctx: FLContext) None[source]
Releases backend resources. Called when the executor handles END_RUN.
The backend tears down per its mode’s lifecycle ownership: stop the in-process trainer thread; send SHUTDOWN and stop the owned process tree (honoring the executor’s shutdown_timeout and stop_grace_period, and pending payload terminal state) for external_process.
Contract: must be idempotent and must not raise. Not called if
initialize()raised (see the cleanup-on-failure contract oninitialize()).- Parameters:
fl_ctx – the FLContext of the END_RUN event.
- initialize(context: ClientAPIBackendContext, fl_ctx: FLContext) None[source]
Prepares the backend for the run. Called once when the executor handles START_RUN.
contextis the frozen snapshot of the executor’s configuration plus a back-reference to the executor (forfire_log_analyticsand logging). A backend should read all of its config fromcontextrather than from executor private attributes, and should retain what it needs forexecute/finalize.The backend sets up its control plane here (DataBus wiring for in_process; Cell session machinery, bootstrap config, and - per launch_once policy - trainer launch for external_process).
Contract: raise an exception on any setup failure. The executor converts the exception into system_panic so the job fails cleanly instead of hanging while tasks wait on a backend that never became ready.
Cleanup-on-failure contract:
initialize()must be exception-safe and self-unwinding - if it raises, it must first release any partial setup it already made (threads started, processes launched, listeners/tokens registered, files written). The executor does NOT callfinalize()on a backend whoseinitialize()raised, becausefinalize()cannot assume a consistently half-initialized backend. Own your own rollback.- Parameters:
context – the frozen backend configuration and executor back-reference.
fl_ctx – the FLContext of the START_RUN event.