I’d go with D since MCW auto-scales to handle concurrency with scheduled tasks, so you don’t need to worry about sizing for overlapping runs. That matches what’s in the official guide and most practice tests. Not 100% if they try to trick us with single-instance scenarios, but D seems safest overall. Agree?

D . MCW lets Snowflake auto-scale the warehouse to handle peak concurrency for scheduled tasks, so you don't need to size it manually for overlapping runs. The other answers focus too much on single runs or stream size, which doesn't cover all cases. Pretty sure D is what they're after but happy to hear other takes if I missed a use case.

Option D looks right since MCW in Snowflake automatically manages concurrency for tasks, so you don’t have to guess the warehouse size every time tasks might overlap. B feels less complete honestly, because stream content size doesn’t cover parallel task execution. I think D fits the typical scenario, but not 100% if there’s some edge case I’m missing. Agree?