Pricing the Second Region Before You Attach It: Replicated Write I/Os Ignore I/O-Optimized

Decision drivers

Replicated write I/O rate ($0.20/M in the us-east-1 example) multiplied per secondary Region.
Replicated write I/Os billed under both Aurora Standard and I/O-Optimized.
Replication data transfer (~2.75 KB per write I/O at $0.02/GB in the same example).
Headless secondary clusters as storage-only DR (must add an instance before switchover/failover).

Production premise

What you're trading: DR posture for a bill that scales as N × full copy. Every secondary Region pays full retail — its own instances, its own storage (the same GB as the primary), its own read I/Os — plus replicated write I/Os billed per secondary at the same count as the primary's in-region write I/Os ($0.20 per million in AWS's us-east-1 worked example), plus inter-region transfer for the replication stream itself (~2.75 KB per write I/O at $0.02/GB in the same example). Aurora I/O-Optimized eliminates in-region I/O charges but not replicated write I/Os: AWS's own I/O-Optimized example still bills the secondary $9/month for 45 million replicated writes.

Failure mode & inflection point

Write-amplified workloads (chatty commits, small transactions) discover that going I/O-Optimized didn't fix global costs: 45M writes/month is $9 per secondary — trivial; 4.5B writes/month (a sustained ~100 write I/Os/sec) is $900 per secondary per month before a single instance hour, multiplied by every secondary Region. Storage-heavy databases re-pay full GB-month rates in every Region, under whatever configuration each Region runs.

Production guardrail

Default to one warm secondary in the Region pair closest to your primary (lowest transfer cost and lag), added only when a named requirement — RPO in seconds, a latency SLO, or data residency — exists. Configure secondaries as Aurora Standard: I/O-Optimized's higher storage rate buys nothing for a Region that mostly receives writes. If the secondary is for DR only and "add an instance first" fits your RTO, a headless secondary cluster (storage only, no compute) is the cost floor.