> Why not store it on the EC2 instance directly? Since the data in S3 receives updates every few hours, querying it directly ensures queries are on up-to-date

> Why not store it on the EC2 instance directly? Since the data in S3 receives updates every few hours, querying it directly ensures queries are on up-to-date data, whereas creating a copy of the dataset on the EC2 would necessitate periodically checking for updates and moving the deltas across to the EC2 (not insurmountable; but complexity worth avoiding if possible).