Exercise 3: Why Data Egress Needs a Destination Breakdown — Possible Solution
====================================================================

Per the chapter, data transfer to different destinations -- cross-
region, cross-AZ, or out to the internet -- is priced VERY differently
from one another. This means total data transfer volume ALONE, without
knowing where that data actually went, tells you almost nothing about
where the actual COST is coming from, or what the appropriate fix
would even be.

Concretely, why total volume alone is insufficient:

Two scenarios could produce the IDENTICAL total data transfer volume
figure while having completely different cost implications and
completely different root causes:

SCENARIO A: A large volume of data moving CROSS-AZ within the same
region (e.g. between an application server and a database in a
different availability zone) -- typically priced much lower than
cross-region or internet-bound transfer, and often simply reflects
normal, expected application architecture (Chapter 1's own AZ
redundancy patterns).

SCENARIO B: The SAME total volume, but moving OUT TO THE INTERNET
(egress) -- typically the most expensive category by a wide margin,
and per this chapter's own examples, often indicates a genuine problem
worth investigating, like a misconfigured backup job sending far more
data cross-region than intended, or a cache misconfiguration causing
repeated, unnecessary re-fetching of the same data from origin.

If only the TOTAL volume were examined, both scenarios would look
identical on paper, even though Scenario A might be completely normal
and expected, while Scenario B represents a genuine, fixable cost
anomaly. Breaking the cost down BY DESTINATION is what actually
reveals which category the traffic falls into -- allowing an
investigator to correctly identify whether there's even a problem to
fix at all, and if so, WHERE (a backup job's target region, a caching
layer's configuration, etc.) rather than just knowing "a lot of data
moved somewhere."

WHY THIS WORKS AS AN ANSWER
------------------------------
This uses a concrete side-by-side comparison to show that identical
TOTAL volume numbers can represent completely different cost profiles
and completely different (or nonexistent) problems depending on
destination -- directly justifying the chapter's own instruction to
"break down data transfer costs specifically by destination... since
these are priced very differently," rather than treating that
instruction as an arbitrary preference.