> Are the readable replicas supposed to be long-lived Right now, yes, they should be relatively long-lived (e.g. hours). Each node keeps a full copy of the dat

> Are the readable replicas supposed to be long-lived Right now, yes, they should be relatively long-lived (e.g. hours). Each node keeps a full copy of the database so it's recommended to use persistent volumes with it so it doesn't need to re-snapshot on boot. We do have plans to make this work with very short-lived requests for platforms like Lambda or Vercel. That'll use transactionally-aware paging on-demand. You could even run it in a browser, although I'm not sure why you would want to. :)