Challange 1 of Chapter 3 Networking Domain –
I agree that C, F, H answer are possible, but i would like your opinion regarding D as i think it could be possible too.. probably 4th in the ladder.. just for the sake of understanding all possibilities
you commented about answer D that :
"Although VPCs A and D are peered, the routing table in VPC A does not have an entry for VPC D’s CIDR block. Well this is one of those all or nothing items. And if we didn’t have a route in the routing table in A, then we wouldn’t be able to reach D. But here it seems like we are able to reach it, it’s just a little slower than they want."
isn’t also possible/true that if you miss step of adding the entry, your traffic instead of going via the peering might go via the normal INTERNET as it might get out using the 0.0.0.0/0 route? so that would explain the SLOWNESS of the traffic.. client is expecting direct traffic VPC A to VPC D without going out.. instead it is going VPC A -> Internet -> VPC D.
in short.. if i forget one step of the peering? do i black hole my traffic or most likely it will use the previous way (if there was one obviously) thank you!
I partially agree. Although I see your point, and make sense (internal vs. external traffic latency), we would have to assume that the resources inside the VPCs (EC2 instances, for example) are public, i.e., Internet Facing… Which is not mentioned on the question. That why (I guess) this option was discarded.