A bill shows surprising data-transfer charges. Review reveals a chatty microservice pair deployed across different AZs in the same VPC. What's the cost-effective fix while keeping AZ redundancy?
A.Same-AZ placement only — accept reduced resilience
Single-AZ sacrifices the resilience requirement.
B.Co-locate chatty services per-AZ with a redundant copy in each AZ (affinity-aware placement)
Per-AZ full-stack placement keeps HA without cross-AZ calls. Correct.
C.Move one service to a different region
Cross-region is worse — higher transfer costs and worse latency.
D.Increase NAT gateway capacity
NAT capacity is unrelated to inter-AZ pricing.
Why this answer
Inter-AZ traffic within a VPC is billed ~$0.01/GB each way — chatty cross-AZ calls are a classic hidden cost. The pattern: deploy independent service pairs per AZ and prefer same-AZ routing, keeping redundancy (each AZ has a full stack) while eliminating cross-AZ chatter.
Correct answer: B · Last reviewed 2026-11-02
Official AWS references
Practice this interactively
Answer without seeing the solution first in the Cost-Optimized Databases & Networking practice session.
Practice Cost-Optimized Databases & Networking