Ask HN: What footguns does a SQL dev need to know when learning MongoDB/NoSQL?

I'm aware of some of the weird behavior and dangerous mistakes I can make with SQL.

- Missing indices on join tables.

- Postgres columns are case sensitve. MySQL columns are not.

- Missing limits on subqueries.

- Etc.

But now I find myself working on an API backed by a NoSQL db. What do SQL devs like me need to know when they wander into a NoSQL dev shop?

I'm curious about the weird footguns as well as the different data modeling approaches. For example, all I know are join tables, but now I'm learning that I can just put a list of objects inside a document.