In my case I would have ~10 million rows with a

vendor_id

+

vendor_name

or 10 million rows with a

vendor_id

and then a

vendors

collection with the

name

There's definitely going to be additional overhead with join. However, there are some use cases that're simply not possible without a join. If you can express your queries in a flat schema, you should consider that first.

@Kishore Nallan yes there are definetely use cases where this is going to be a game changer for us as well where we will happily take the performance degredation. I will continue flattening when it's possible. Thank you for swift and precise response 🙏