M201 How Data is Stored on Disk

I am using Atlas for my Database and collections. On watching this video I would like to know how you can set up my Atlas database to take advantage of Paralelisation of I/O? It would seem that while this class is good, it appears to be geared to a manual install. Is there any reference or way to set this up on my atlas cluster?

Also… If I use this format, would I use one cluster for the database and another for the index? or are the disks contained on the same cluster? To quote the lecture.

/data/db/

hello/
collections/
index/

How is this structure in Atlas? And how can you verify it in Atlas?