Exercise 1: Choosing a Journaling Mode for a Database Server — Possible Solution
====================================================================

Recommended mode: journal (full data and metadata journaling)

Explanation:

Per the chapter's own compare-table, journal mode is the ONLY one of
the three that protects both data and metadata -- ordered mode only
journals metadata (with data merely written before its own metadata
commits, not journaled itself), and writeback mode journals metadata
alone, with the chapter explicitly warning that "data itself could
still be corrupted after a crash even though the filesystem's own
structure stays consistent" under that mode.

For a database server specifically, the scenario explicitly states
that data integrity matters more than raw write speed -- exactly the
priority journal mode is built for, at the direct cost the chapter
names ("the slowest"). A database's own actual row/table data being
corrupted after a crash -- which ordered or writeback modes both
genuinely permit, since neither journals the DATA itself, only
metadata -- would be a far more serious problem for a database than
for, say, a scratch/cache filesystem where losing some in-flight data
after a crash is tolerable. Since the scenario has already decided
that integrity outweighs speed, journal mode is the correct choice
specifically because it's the only mode providing a real guarantee
against post-crash data corruption, not just metadata/structural
consistency.

WHY THIS WORKS AS AN ANSWER
------------------------------
This selects the correct mode based on the exercise's own stated
priority (integrity over speed) and justifies it by explaining
specifically what ordered/writeback modes leave unprotected (the
data itself, not just the filesystem structure) that journal mode
alone actually covers.