👋 Hello, team! We are a small start up - trying to use Prefect cloud. We have set up ECS push work pool with AWS. We have one image with all the pipelines. So I am trying to create image outside flow.deploy(). And I am trying to configure entry point type pointing the module in path: `flows/my_flow.py`:

@flow(name='Test Flow', task_runner=ConcurrentTaskRunner(), log_prints=True)
def my_flow():

my_flow.deploy(  # -- Production Schedule
        name="my_flow_deployment",
        work_pool_name="work_poll_ecs_fargate",
        image=DeploymentImage(
           name="prefect-ecs-prod-flows:latest",
           platform="linux/amd64",
        ),
        build=False,
            entrypoint_type="flows.my_flow:my_flow"
        # schedule=IntervalSchedule(
        #     interval=timedelta(days=1),
        #     anchor_date=datetime(year=2024, month=3, day=1, hour=8)
        # )
    )

This is how I created the deployment:

python -m <http://flows.my|flows.my>_flow

. Deployment was created but when I try running the deployment it crashed with no logs. Please anyone let me understand whether using on repo for all image is a good practice and if using same repos for all flow how to specify dentrypoint correctly

hi @krishna!

entrypoint_type="flows.my_flow:my_flow"

^ this is not the intended use of

entrypoint_type

- this wants an `Enum` that you shouldnt have to worry about I would remove

entrypoint_type

here and instead add a

from_source

like this > whether using on repo for all image is a good practice whether its good practice depends on your use case and organizational practices! I will say it is a common pattern