Ok, I'm running into another issue. I have a Kubernetes worker that starts jobs with a template that references a custom built docker image. When the job pod runs, it attempts to load the flow code from storage at

/opt/prefect/flows

and the image I have created has the flow .py files there. But it always gives this error: "Flow could not be retrieved from deployment."

how do you have your storage set up?

I don't have any storage, the files just exist in the docker image.

gotcha. haven't done this type of execution setup since I clone my code from gitlab, but I'll take a look at the docs

I think you can specify storage as none then provide your own docker image - as long as the path is set correctly it should pick it up

@Kevin Grismore I haven't, but I noticed it was looking there for the files so I just put them there when building the custom container.

in the deployment build from flow

Thanks @redsquare and @Kevin Grismore for jumping in. My specific issue was I wasn't using the right image. Then I ran into an issue where I needed to redefine the service account with additional permissions.

prefect kubernetes manifest agent > prefectAgent.yaml

is your friend for that.