To explain and exemplify my approach, I’ve created a Docker Container Block with the

alpine

image and I’ve edited the

command

to

ls

(This will make the agent run

ls

instead of running the flow). To build and apply the deploy, I created an empty dummy flow (just because it is required by the command), called

flow_1

, and ran:

prefect deployment build flow_1.py:flow_1 -n deploy -q test -ib docker-container/not-python -a

flow_1

:

@flow
def flow_1():
    pass

After running the deploy, the agent pulls the

alpine

image, runs the containers, runs the command

ls

and kills the container, as expected. The only problem is that the agent does not send updates regarding the flow run status to the server, since the flow was not run. So the status of the flow run remains

Pending

forever. Is there a way to update the status in this circumstances? Maybe I could use an image which contains python and edit the

command

of the Docker Container Block to something like

ls && python flow_1.py

(just an example, probably the python command required will be more complex than that) ? This way both

ls

and the flow will run, and the server will be updated with the

Complete

status of the flow run.

Docker Container Block: