Notebook jobs are running periodically, but very briefly to execute a small task. For the sake of cost savings I would like to make sure that they are only consuming resources during the execution.
It's not clear to me from the documentation whether a scheduled notebook job will spin up it's own environment during execution or require that there is an existing Sagemaker instance/app/kernel available on which it can execute.
When you schedule notebook jobs, your Jupyter notebooks run on SageMaker training instances. You are charged for the compute type you choose, based on the combined duration of use for all notebook jobs that run from the job definition. If you don’t specify a compute type, SageMaker assigns you a default Amazon EC2 instance type of ml.m5.large.
https://docs.aws.amazon.com/sagemaker/latest/dg/notebook-auto-run-pricing.html
