Spawning Child Jobs#

Orchestration Python can spawn sub-jobs of another workflow inside of a workflow. This is commonly useful if your main workflow needs to initiate another workflow or set of workflows.

Create Child Job#

The workflow can create a sub job in the request queue of a lab using artificial.workflows.util.create_child_job()

Spawn Child Job#

The workflow can a sub job and schedule it to run ASAP using artificial.workflows.util.spawn_child_job()

Spawn Children Example Code#

parent_job_wf.py

from artificial.workflows.decorators import workflow
from artificial.workflows.util import create_child_job, spawn_child_job
# Import the child workflow
from child_job_wf import child_job

@workflow('Job Spawning', 'wf_parent', 'lab_c9eae09e-ceed-46d1-8390-87ff7dd8586d')
async def parent() -> None:

    num_samples = 1
    # This job will spawn and go into the request queue
    # The function name and params match the child workflows function definition in the child workflow
    await create_child_job(
        child=child_job(plate_barcode='Test_1', num_samples=num_samples),
        job_name='Job 1'
    )

    num_samples += 1
    # This job will spawn and be immediately scheduled
    # The function name and params match the child workflows function definition in the child workflow
    await spawn_child_job(
        child=child_job(plate_barcode='Test_2', num_samples=num_samples),
        job_name='Running Job 1'
    )

child_job_wf.py

from artificial.workflows.decorators import workflow
from artificial.workflows.runtime import show_info

@workflow('Child Job', 'wf_child', 'lab_c9eae09e-ceed-46d1-8390-87ff7dd8586d')
# This is the function name and definition being passed into create/spawn child
async def child_job(plate_barcode: str, num_samples: int) -> None:
    await show_info(f'Barcode: {plate_barcode}, num_samples: {num_samples}')