re: downtime, since it’s a single node cluster, you’ll see a downtime of about 20 minutes for your dataset