Exercise 3: Explaining an Archive-Tier Retrieval Surprise — Possible Solution
====================================================================

Why this happened:

The old backup had almost certainly been moved, automatically, by a
LIFECYCLE POLICY into an archive/cold storage tier at some point after
it aged past whatever threshold that policy defined -- a completely
routine, intended behavior, not an error. Per the chapter, archive
tiers trade a much lower ongoing storage cost for slower, and often
separately charged, retrieval -- "retrieving data from an archive/cold
storage tier can take anywhere from minutes to several hours, and
often carries a real retrieval cost on top of the storage savings."
The customer's expectation of instant, free access -- reasonable if
they were thinking of the object as sitting in standard/frequent-access
storage -- simply doesn't match how the archive tier it had actually
been moved into behaves.

What a support engineer should proactively communicate:

1. WHICH DATA HAS A LIFECYCLE POLICY APPLIED, and roughly what
   threshold triggers a tier transition (e.g. "backups older than 90
   days move to archive storage"), so customers understand in advance
   that older data won't behave the same way as recently-created data.

2. THAT RETRIEVAL FROM ARCHIVE TIERS IS NEITHER INSTANT NOR FREE --
   setting the expectation explicitly, before a customer discovers it
   unexpectedly during an actual restore, ideally as part of onboarding
   or whenever a lifecycle policy is first configured.

3. FOR TIME-SENSITIVE RESTORES, that retrieval delay should be
   factored into any recovery planning -- if a customer might need
   fast access to old backups during a real incident, archive tier
   retrieval times could meaningfully affect how quickly they can
   actually recover, which is worth knowing well before an emergency,
   not during one.

WHY THIS WORKS AS AN ANSWER
------------------------------
This directly explains the surprise using the chapter's own
lifecycle-policy and archive-tier material, then turns the chapter's
own warn-box ("a real support conversation distinction... worth having
proactively, before a lifecycle policy quietly moves data somewhere
slower to retrieve") into concrete, actionable communication points a
support engineer could actually use -- rather than only explaining the
mechanism after the fact, once the surprise has already happened.