Feature Requests

pgsearch ( https://github.com/paradedb/paradedb/tree/dev/pg_search ) is a Postgres extension that enables full text search over heap tables using the BM25 algorithm. The docs don't mention it at the time of this post: https://docs.render.com/postgresql-extensions As many companies #justusepostgres it would be very helpful to have something like pgsearch supported by Render.

Currently, you can only create read replicas in the same region. It would be nice to create read replicas in other regions. That way, I could have my primary in Oregon and my read replica in Frankfurt.

the feature would make db management easier, as it allows you to restore your database to a specific moment in time, which is particularly useful for recovering from mistakes like faulty migrations or data corruption

pganalyze has a lot of very useful query stats and other monitoring data that Render's PostgreSQL doesn't provide out of the box. Similar to the Datadog integration, we should have a way to easily integrate pganalyze with Render.

Allow for wal_level to be set to 'logical'. https://community.render.com/t/postgres-docker-disk-render-config/3010 https://github.com/supabase/realtime

As a user I can log into the shell for servers and would like the ability to do the same for databases.

This is an extension that implements stronger full-text-search support in Postgres - more docs here: https://blog.paradedb.com/pages/introducing_bm25 This would be great because it'd enable people to use Render Postgres for more without having to add ElasticSearch to their stack.

We occasionally want to allow access to the prod database via our personal IP addresses. However I only want that access to last an hour and then have it be automatically revoked.

Right now replicas have the same credentials as the primary database. This limits its usefulness as a way to provide certain people or applications with read-only access. In addition to that, it means the Replica database is incompatible with Ruby on Rails's native "multiple database" functionality as it requires having a distinct user for each database: https://guides.rubyonrails.org/active_record_multiple_databases.html#setting-up-your-application >> Second, the username for the writers and replicas should be different, and the replica user's database permissions should be set to only read and not write. I expect few customers needing to specify their own credentials. Anything that is different from the primary database should suffice both use cases.

Being able to see which SQL queries are negatively impacting app performance.