Exercise 1: One IP, Three Regions, Automatic Failover — Possible Solution
====================================================================

Yes, GCP's global load balancer can do this natively. A single anycast
IP address genuinely serves as the frontend for backend instances
across multiple regions at once, and real cross-region failover is
already built into that same global architecture - no additional
service needs to be layered on top to get multi-region routing or
failover.

An AWS Application Load Balancer alone cannot do this - an ALB is
inherently a regional construct, bound to one Region's own VPC. To get
one address routing across three regions with failover on AWS, the
team would need to add a separate service on top of three regional
ALBs, typically Route 53 (DNS-based routing/failover) or CloudFront,
rather than getting multi-region behavior from the load balancer
itself the way GCP provides it out of the box.

ANSWER: Yes, GCP's global load balancer handles this natively via a
single anycast IP with built-in cross-region failover. AWS's ALB alone
cannot - it is a regional service, so the team would need to add a
separate layer (such as Route 53) on top of three regional ALBs to
achieve the same real multi-region behavior.

WHY THIS WORKS AS AN ANSWER
------------------------------
This correctly identifies GCP's real global-by-default load balancing
architecture and contrasts it with AWS's genuinely regional ALB,
matching the chapter's own central structural finding.