The cloud-native orchestrator for the whole development lifecycle, with integrated lineage and observability.

dagster

Hey all, looking for clarity on the LaunchRunMutation using dagster's graphql API.
Where can I find the repositoryLocationName?
Should runConfigData include required resources, partitions_def? Is there an example I can use for this type?

Thanks in advance!

hi <@U05BPDR70BH>! for repositoryLocationName, you can find this if you go to the "Deployment" tab in the UI. If you're using the new `Definitions` API, then the things listed there will be the repository (code) location names, and the repository will be `__repository__` . If you're using `@repository`, then that tab will show you `{code_location_name}@{repository_name}` .

As for runConfigData, this just needs to be the actual run config you'd like to supply to the run, so essentially just a dictionary of config in the same shape that you'd pass into `@job(config={...})`. One example would be
```{
  "ops": {"my_op": {"config": {"foo": 1, "bar": "baz"}}}
}```

I also wrote up the bit about the repository args into a github discussion for better visibility (as it's pretty confusing at the moment): <https://github.com/dagster-io/dagster/discussions/14752>

<@U01J51Y6B9D> One of the things that I want to pass in config is for what partition to run the config for. For example: I want to run this job for partition 1, but using the launchrunmutation. How would I be able to achieve this?

ah I see -- partition keys can be set via the tags, see: <https://github.com/dagster-io/dagster/discussions/14442>

<@U01J51Y6B9D> Can this be done in a pure graphql query (not pythong client)? Thanks again!

yep! under the hood, the python client is just executing a graphql query. tags get passed via `{"executionMetadata": {"tags": {"dagster/partition": "..."}}}`

Gotcha. Thanks <@U01J51Y6B9D> , you rock!