Auto Scaling

Now with the @function(...) and @signal(...) decorators, you have the option to customize how your code will run, for starters a good auto scaling strategy can be beneficial to make sure your code gets the proper scaling

1. Create a QueueAutoScaler instance by importing the QueueAutoScaler class

   from snowcell.workload import QueueAutoScaler 
   
   asg = QueueAutoScaler(max_containers=3, tasks_per_container=2)

2. Add the auto scaling instance to your decorator: This will now run your block of code with the asg policy you defined and attached

   @client.function(name="my task", cpu=1.5, memory=2, autoscaler=asg)
   def predict(**inputs):
       return inputs['x'] + inputs['y']
   
   predict(x=3, y=2)