I was leaning toward D, partial mesh, since it also has some redundancy and seems more scalable than a full mesh. But maybe I’m missing something about the "highest" part. Wouldn’t full mesh be overkill except in tiny setups?
Wouldn't tree topology (B) actually create several single points of failure at the higher nodes? I get why someone might pick D for scalability, but the question's all about availability, not cost. Full mesh (C) has no single points, so that's why it's usually picked for these types of questions.
C for sure. Full mesh means every device is connected to every other, so if something fails there are still tons of paths available. Partial mesh adds some redundancy but not to the same level. Pretty confident on this unless I'm misreading what they want by "highest." Agree?