prawn-stack VS pg_ivm

I wanted incremental refresh in Postgres as well and found that you can manage your own table to get something close.

Basically you create a regular table in place of a materialised one, only aggregate data newer than what's currently in the table then store the new aggregates in table. Repeat this an interval.

https://github.com/cadbox1/prawn-stack/blob/master/src/backe...

I use this to show page view data aggregated by hour without calculating it on each request using Lambda

https://prawn.cadell.dev/

I don't have the startup part yet but here's my one-person stack with Postgres, Node and React deployed on AWS with CDK using RDS, Lambda, S3 and Cloudfront. It's 100% in the free tier.

https://github.com/cadbox1/prawn-stack

FTA, because I don't like Jeopardy questions in headlines:

“Incremental View Maintenance (IVM) provides a method for keeping materialized views current by calculating and applying only the incremental changes, as opposed to the complete recomputation of contents performed by the REFRESH MATERIALIZED VIEW command.”

Article shows using the pg_ivm Postgres extension available here: https://github.com/sraoss/pg_ivm

There is an extension that provides some limited incremental MVIEW refresh: https://github.com/sraoss/pg_ivm

There’s awesome work being done on incremental view maintenance in postgres:

https://github.com/sraoss/pg_ivm

It makes sense to perform managed denormalization - use a materialized view or automatically refresh a table or foreign server (via FDW) using common triggers (like pg_ivm does). And it's fine to add a TTL to it and use as a read store... update on user login and make a partial index just for that. And that's how you could get CQRS...