Retrieve/Change Model Configuration with the REST API
I have jobs that will call a Model API. Ideally, I could scale up the number of instances hosting my model when my job runs and scale back down after it is done so that I could quickly finish my job but also have the model available to query at will. Right now, the model is configured with a lot of instances and I start and stop it using the REST API when I run my job.
Another issue I have noticed is that depending how busy our cluster is, I will not receive the configured number of instances when starting the Model. If I could retrieve the number of healthy nodes from the REST API, I could configure my workers calling the Model accordingly.
While all of this is doable with the GUI, being able to query this information programatically has clear advantages in the context of scheduled jobs.