I want to use the k8sRunLauncher but my kubernetes cluster requires resource limits. I read in docs.dagster.io/deployment/guides/kubernetes/customizing-your-deployment that I can push that requirement into the tag of the job. But this means I have to do that for every job. Is there a way to change the default of

@job

? So I don't have to modify every job?

cc @chris for tips on applying tags across jobs

we don't have a great way of doing this to definition level objects unfortunately. Right now, the only way is to actually alter each definition object. You can use a factory pattern to make this easier in the future perhaps? But it's on our radar to improve the story around definition-level tags

@johann Is this now made easier with 0.14.10? I see a includeConfigInLaunchedRuns key in the newest docs here https://docs.dagster.io/deployment/guides/kubernetes/deploying-with-helm#configure-your-user-deployment

Yep! That gives you one place to specify k8s config for your gRPC server and every run that launches for it

Very cool, that is exactly what I want, but is there maybe an example of a filled out deployment? Because I can't use helm in my environment and so I have to manually fill in a lot of stuff. It seems we copy the job configuration underneath the DAGSTER_CLI_API_GRPC_CONTAINER_CONTEXT env var, but I'm not entirely sure how to nest the values there. (list $deployment | fromYaml | toJson| quote)

Hmm yeah we don’t support the non helm use case very well here… if you just locally run

helm template

, you can see what the value of the env will be

I see this is implemented in 0.14.13! Thanks @johann!