Description
Over the years many users have been frustrated that it is hard to programmatically get an ML job config from one cluster and then put it into a different cluster because the get jobs API returns augmented config documents containing more fields than the config required by the put job API. The put job API will reject configs that contain unknown fields. This creates a burden on the client implementing the transfer process to know which fields need to be kept, and this is especially difficult because the list varies from release to release.
We can address this by adding an option to the get jobs endpoint to return configs containing all the fields required by put job, but no more.
The same option and filtering functionality should be added to every other get endpoint we have that returns more than the corresponding put endpoint requires.