i 'm deploying dagster with CeleryK8sLauncher, rep...
# ask-community
p
i 'm deploying dagster with CeleryK8sLauncher, repository is loaded from user-deployment grpc. Then in launchpad UI, i can not see the below
execution
config. If i config the workspace to load repository directly from package location, the config will display correctly.
Copy code
execution:
  celery-k8s:
    config:
      broker:
      env: DAGSTER_CELERY_BROKER_URL
      backend:
        env: DAGSTER_CELERY_BACKEND_URL
      job_namespace: <abc>
y
cc @daniel is it an oversight in launchpad or an intentional behavior that needs better error handling?
d
I haven't seen this problem reported before. Is there a way for us to reproduce it ourselves?
p
my setup is on kubernetes with an specific namespace (not default namspace) 1. an user deployment 2. CeleryK8sLauncher 3. i expect to see the following config appear in launch pad,
runs
table:
Copy code
execution:
  celery-k8s:
    config:
      broker:
      env: DAGSTER_CELERY_BROKER_URL
      backend:
        env: DAGSTER_CELERY_BACKEND_URL
      job_namespace: <abc>
d
I can't think of a reason why changing the run launcher would change what's shown in the Launchpad tab. Are you sure that's the only thing that is changing when what you are seeing the Launchpad changes?
p
when i change from user-deployment to package, the
execution
config show correctly.
maybe there is an issue from grpc api.
d
What precisely do you mean when you say change from user-deployment to package? Can you give an example, is there example code that we could use to try to reproduce the problem ourselves locally?
p
i mean changing the workspace.yaml from:
Copy code
load_from:
  - grpc_server:
      host: abcdf
      port: 3030
      location_name: abcdf
to
Copy code
load_from:
  - python_package: 
      package_name: abcdf
d
This is when running dagit locally?
p
sorry, i did deploy on real environment, not locally.
d
Ok - what you're describing is not expected, but I think we are going to need to have a way to reproduce the problem ourselves in order to be more helpful in figuring out what's going on
👍 1
p
i 'll let you know if i can reproduce it locally.