A production MySQL database must survive an AZ failure with automatic failover and no manual intervention. Offloading read queries is a secondary nice-to-have. What is the PRIMARY mechanism?
A.Multi-AZ deployment — synchronous standby in another AZ
Synchronous standby + auto-failover = HA. Correct.
B.Read replicas in another AZ
Replicas are async — designed for read scaling, not HA failover.
C.Automated snapshots every hour
Backups restore after failure — they don't prevent downtime.
D.Multi-region active-active
Multi-region is DR, not AZ-failure HA.
Why this answer
Multi-AZ = synchronous replication to a standby in another AZ with automatic failover on infrastructure failure — the HA feature. Read replicas are async and serve reads (scaling), not automatic standby. Snapshots are backups (RPO gaps), not HA.
Correct answer: A · Last reviewed 2026-11-02
Official AWS references
Practice this interactively
Answer without seeing the solution first in the High Availability practice session.
Practice High Availability