Hi all! I am running into an issue with my user code deployment. For some context, I am running Dagster 0.13.9 deploying from Helm on GKE. My user code is structured as follows (no workspace.yaml file).

/
+-- dagster_workspace/
    +-- data_platform_dagster/
        +-- repository.py

I am testing the

dagster api grpc

cli command in my user code container in interactive mode. When I run

dagster api grpc -h 0.0.0.0 -p 3030 --python-file repository.py --working-directory /dagster_workspace/data_platform_dagster

or

dagster api grpc -h 0.0.0.0 -p 3030 --python-file /dagster_workspace/data_platform_dagster/repository.py

I get the following error:

dagster.core.errors.DagsterImportError: Encountered ImportError: `cannot import name 'k8s_job_executor' from 'dagster' (/usr/local/lib/python3.8/site-packages/dagster/__init__.py)` while importing module repository from file /dagster_workspace/data_platform_dagster/repository.py. Local modules were resolved using the working directory `/dagster_workspace/data_platform_dagster`. If another working directory should be used, please explicitly specify the appropriate path using the `-d` or `--working-directory` for CLI based targets or the `working_directory` configuration option for `python_file`-based workspace.yaml targets.

I appreciate any direction. Thanks!

Hi Arslan - I think you want

from dagster_k8s import k8s_job_executor

(not

from dagster import k8s_job_executor

)

I knew it was something simple 🙃. Thanks for your help!