For me, reading about B+ trees and figuring out how many db-pages get touched in an operation was enough without having to write any implementations myself. For

For me, reading about B+ trees and figuring out how many db-pages get touched in an operation was enough without having to write any implementations myself. For others it could solidify understanding. If your db uses LSM Trees then characteristics are different again, so YMMV.