Question about "Using Atomic Transactions to Power an Idempotent API" about sorg HOT 1 CLOSED

Hey Bruno, sorry for the really late reply on this!

Good question, and thank you for reading!

I've been reading about isolation levels lately. But it's all theoretical - just showing DB features. Meaningful examples, how to apply those DB features to applications, are scarce.
If you have a resource (blog, book, something else) to recommend on this topic in the vein of your article linked above I'd be grateful!

I wish I had something better for you here, but I'd still reference the Postgres docs on isolation levels as the best canonical source of reading around.

There are a few other decent blog posts online, but like those docs they tend to focus more on the technical aspect of isolation levels than how to build applications with them (and you've probably read them). Two I'd call out from large-ish companies are:

• "Cross shard transactions at 10 million requests per second" https://dropbox.tech/infrastructure/cross-shard-transactions-at-10-million-requests-per-second
• "Why you should pick strong consistency, whenever possible" https://cloud.google.com/blog/products/gcp/why-you-should-pick-strong-consistency-whenever-possible

I've been looking for a book that covers this kind of thing in detail, but haven't found one to date. My hypothesis is that the intersection between know how on isolation levels and people that practically architecture apps is a far smaller specialization than you'd hope. There are a lot of technical database people who know about isolation levels and not much about building applications, and there are a lot of people who build applications and who don't know much about their underlying databases, but people (and by extension published material) with significant experience in both are relatively rare.

Anyway, going to close this out for now, but let me know if you find something!

from sorg.