How do you configure policies to archive data when the dataset gets too large, like how you can with Glacier policies in AWS S3?
The comparable capability in Foundry is Retention Policies, which let you delete data according to various criteria (see docs for specifics).
You can also write a pipeline job that truncates a dataset and ships certain slices of that dataset to another storage system for long-term storage (e.g. AWS Glacier). This is useful if you want to preserve the data, but not in S3 or whichever backing system your Foundry datasets rely on.