The 8 Functions That Break Your Postgres Indexes

There's a class of bug I see constantly in code review and almost never on a status page. The query is correct. The index exists. The plan is a sequential scan. -- index on users(email) exists SELECT