Exercise 3: Why ws1 and a Backup Server Reach Opposite Conclusions on Question 2 — Possible Solution
====================================================================

Explanation:

Framework question 2 asks "Does the data need protection against
silent corruption over its lifetime?" -- the answer to this question
depends entirely on the DATA's own access pattern and role, not on the
disk hardware underneath it, which is why identical hardware can
produce opposite conclusions for two different workloads.

Per the chapter, ws1's web server involves "modest, predictable
storage needs" -- web content and logs. This kind of data is, by its
own nature, frequently accessed and frequently regenerated: logs
rotate and get replaced regularly, web content gets deployed/updated
through normal application workflows, and much of it is either
disposable or already backed up/version-controlled elsewhere (e.g. in
git, per the site's own deployment practices). If a bit of corruption
did occur, it would likely be noticed relatively quickly through
normal use, and much of the data has a natural refresh cycle that
limits how long any single corrupted block could persist unnoticed.
The chapter concludes ext4 is the right, "appropriately boring" choice
specifically because this exposure window is short and the
consequences of a rare, undetected bit-flip are low.

The backup/archive server is the structural opposite: per the chapter,
"backup data can sit untouched for months, and undetected bit rot
could go unnoticed until a restore is actually attempted and fails."
This data is written once and then deliberately left alone for long
stretches, specifically BECAUSE its whole purpose is to exist as a
static, unchanging safety net -- there's no natural refresh cycle
correcting for accumulated corruption the way ws1's logs and
redeployed web content have. The exposure window here is exactly as
long as the data sits unused, which for backup data can be very long,
and the consequence of undetected corruption is severe and delayed --
discovered only at the worst possible moment, during an actual
restore attempt.

Since both servers could run on physically identical disks, the
opposite conclusion comes entirely from how the DATA itself behaves
over time -- frequently refreshed and low-consequence for ws1, versus
long-dormant and high-consequence-if-corrupted for the backup server --
not from any difference in the underlying hardware's own likelihood of
bit rot, which is the same either way.

WHY THIS WORKS AS AN ANSWER
------------------------------
This identifies that question 2's answer depends on data access
pattern and consequence-of-corruption, not hardware, and explains the
specific difference in each scenario's own data lifecycle that
produces the opposite conclusion despite identical underlying disks.