Exercise 2: Why a Zonal MIG Alone Isn't Protected Against a Zone Failure — Possible Solution
====================================================================

Autoscaling and autohealing both operate entirely within whichever
zone the MIG is configured to use - a zonal Managed Instance Group
places every one of its instances in that single real zone. If that
zone itself goes down, every instance in the group goes down with it
at the same time, and there is nothing in that zone left to autoscale
or autoheal into - both features protect against individual instance
failures, not against the entire zone disappearing.

Autoscaling and autohealing being active does not change this - they
are real, useful features, but they operate on the assumption that
the zone itself is healthy. The real fix is to use a regional MIG
instead, which automatically spreads instances across multiple zones
in the region - if one zone fails, the instances in the other zones
keep the service running, giving genuine zone-failure protection that
a zonal MIG cannot provide on its own.

ANSWER: The assumption is wrong because autoscaling and autohealing
both operate within the group's single zone - if that zone fails,
every instance fails with it, since there is no other zone for the
group to fall back on. The fix is to use a regional MIG instead, which
spreads instances across multiple zones and so survives a single
zone's own failure.

WHY THIS WORKS AS AN ANSWER
------------------------------
This correctly distinguishes instance-level protection (autoscaling/
autohealing) from zone-level protection (regional vs. zonal MIG),
directly reusing the chapter's own explicit warning that mirrors
Azure's own scale-set caveat.