Bring your towel and join one of the fastest growing data communities. Welcome to our second-generation open source orchestration platform, a completely rethought approach to dataflow automation.

Prefect Community

<@ULVA73B9P> can you explain `PREFECT__STORAGE_BASE_PATH`  to me?

The `PREFECT__STORAGE_BASE_PATH` environment variable is used in Prefect to specify a default local path for file storage. This path acts as a base directory where Prefect will store or access files unless another path is specified. It ensures that file operations are restricted to a specific directory, providing a level of security and organization.

For more detailed information, you might want to check out the Prefect documentation on file storage <https://docs.prefect.io/latest/develop/results#specifying-a-default-filesystem|here>.

hmm <@U02KG43T675> this should be an implementation detail, so Im concerned there's some bug if you're asking about it

do you mind clarifying why you're asking?

yeah it's related to my previous questions (<https://prefect-community.slack.com/archives/CL09KU1K7/p1738004778709099>) about the `.deploy()` where it seems like the  `entrypoint_type` isn't being used/detected, hence the suggestion to use a `prefect.yaml` where you can specify the directory for the flow, ex:
```  entrypoint: flows/export/set_export_freq.py:set_export_freq
  path: /opt/prefect/arte_tasks/
  pull:
    - prefect.deployments.steps.set_working_directory:
        directory: /opt/prefect/arte_tasks```

I was hoping to skip using a `prefect.yaml` and simply have all the config for the Docker image build in the `.deploy()` since our Docker build is kind of complex needing to get env variables in order build correctly:
```    A_TOKEN = os.getenv("A_TOKEN")

    set_export_freq.deploy(
        name="set_export_freq",
        work_pool_name="arte-k8s",
        image=DockerImage(
                name="docker-arte-release.dr-uw2.adobeitc.com/arte-flows-dflake",
                tag="0.4",
                dockerfile="containers/arte-flows/Dockerfile",
                buildargs={"username": "arte", "pypitoken": A_TOKEN},
        ),
        entrypoint_type="arte_tasks/flows/export/set_export_freq.py:set_export_freq",
        parameters={"started_by": "dflake", "dry_run": True, "dpc_type": "prod", "verbose": True},
        #cron="0 9 * * *",
        description="Find rsids with > 1M daily hits. Set the export interval to 30 min if interval > 30 min.",
        tags=["automation", "export"]
    )```

aha

so `entrypoint_type` should probably be private, and you shouldnt have to set it. if you go to definition on that you'll see it defaults to an enum member which is literally "file_path"

<https://github.com/prefecthq/prefect/blob/main/src/prefect/flows.py#L1095>

whereas `entrypoint` is the kwarg you'd set in `from_source` if you're _not_ baking your code into your image

so i would totally remove `entrypoint_type` from your `.deploy()` to start, and then maybe see if there's some pathing issue in the dockerfile

we are baking our code into the image - so we can't use `from_source` right?  I'm not sure how the flow will be able to find the entrypoint then

I guess I was just overthinking it - thanks Nate!

yep so with `.deploy` if you're baking code into your image then you should only need to set your `image` and then you can handle pathing inside the `Dockerfile` itself