refinery VS metrics

Great write up! At PeerDB, we’ve been using refinery https://github.com/rust-db/refinery to handle database migrations of our catalog Postgres database from Rust. It is easy, typesafe and gets the job done. Thought this would be useful for users building apps with Rust!

I don't think migrations need to be tied to ORM. We have refinery that allows you to write migrations in rust. I, personally, didn't like and prefer writing them in SQL - why learn how to migrations in X when you already know the SQL dialect that you're using.

The best solution I've ever seen is this Rust library https://github.com/cornucopia-rs/cornucopia

You write plain SQL for you schema (just a schema.sql is enough) and plain SQL functions for your queries. Then it generates Rust types and Rust functions from from that. If you don't use Rust, maybe there's a library like that for your favorite language.

Optionally, pair it with https://github.com/bikeshedder/tusker or https://github.com/blainehansen/postgres_migrator (both are based off https://github.com/djrobstep/migra) to generate migrations by diffing your schema.sql files, and https://github.com/rust-db/refinery to perform those migrations.

Now, if you have simple crud needs, you should probably use https://postgrest.org/en/stable/ and not an ORM. There are packages like https://www.npmjs.com/package/@supabase/postgrest-js (for JS / typescript) and probably for other languages too.

If you insist on an ORM, the best of the bunch is prisma https://www.prisma.io/ - outside of the typescript/javascript ecosystem it has ports for some other languages (with varying degrees of completion), the one I know about is the Rust one https://prisma.brendonovich.dev/introduction

What I personally use here is Refinery, but it's within the Rust ecosystem. It lets you write your migrations in pure SQL - which is what I prefer - but still requires about 5 lines of boilerplate Rust code.

Example refinery migrations

It's worth keeping an eye on Diesel's metrics suite (https://github.com/diesel-rs/metrics) as well; I found and fixed some suboptimal buffering that was affecting performance.

See here for some numbers. The relevant code lives inside the diesel github repository. Please also keep in mind that these are just numbers and you should run those these on your own and also run tests with your actual work load.

Performance is worse than in comparable frameworks

For the database part you might want to checkout a crate that's not based on sqlx as sqlx is known for providing non-optimal performance for the sqlite backend. rusqlite or diesel perform much better for this use case.

No it will likely not be less performant. See these numbers for some benchmark results for numbers. (As always with benchmarks: Please don't trust my numbers. To be sure you need to do your own benchmarks with your own use-case)

See here for some benchmark results. The benchmarks itself are in the diesel repository. Otherwise I believe there are numbers in the techempower benchmarks as well, although that includes other factors .

I'm developing a Rust web server backend in Axum that uses Postgres and performance will be pretty important since I plan to run it on one server for as long as possible. It seems like the postgres crate is about 2x faster than sqlx, and the postgres repository seems pretty active still.

Congratulations to the release. I know all of this is hard work. I would like to invite you to submit a ormlite implementation to the diesel benchmark collection. As soon as that's merged you will get regular reports here. The relevant code is here in the diesel repository.

In that case you may be interested in the metrics for different database libraries. diesel is doing rather well at the moment. sqlx is in the middle of a large rewrite that should improve performance, so we'll see how it compares after that

See here for some benchmark results for the diesel repository. Please keep in mind that as always with benchmarks, these numbers are not necessarily true for your usecase. Be sure to checkout at least the benchmark code and draw your own conclusions from there.