Black Friday sale: an extra 40% off every course until Monday, November 30.New Year sale: an extra 30% off every course until Friday, January 1.Memorial Day sale: an extra 25% off every course until Monday, May 31.SysAdmin Day sale: an extra 25% off every course until Sunday, August 1.Back to School sale: an extra 25% off every course until Sunday, August 22.Labor Day sale: an extra 25% off every course until Monday, September 6.Black Friday sale: an extra 40% off every course until Monday, November 29.New Year sale: an extra 30% off every course until Saturday, January 1.Choose your certification →

Copyright (c) 2026 MindMesh Academy. All rights reserved. This content is proprietary and may not be reproduced or distributed without permission.

3.2.1.7. Cost-Aware Multi-Region Data Replication (S3 Replication, DynamoDB Global Tables)

3.2.1.7. Cost-Aware Multi-Region Data Replication (S3 Replication, DynamoDB Global Tables)

💡 First Principle: Every replica is a second copy you store, a second write you pay for, and inter-Region transfer for each byte. Replicate only the data a recovery or latency requirement names, into the cheapest form that still meets its restore objective.

Scenario: A company replicates every object in a 2 PB bucket to a second Region in the "S3 Standard" class with Replication Time Control enabled, and runs a "DynamoDB" global table in four Regions, although only one prefix is needed for disaster recovery, an RPO of a day is acceptable, and only two Regions serve users.

  • "S3" replication: Each replicated GB pays inter-Region transfer, replication PUT requests, and full storage in the destination. Reduce all three by filtering the rule to the prefixes or tags that recovery needs, writing replicas straight into a cheaper storage class (for example "S3 Glacier Instant Retrieval" for rarely read DR copies that must still be readable in milliseconds, or "Glacier Deep Archive" when retrieval in hours is acceptable), and adding Replication Time Control (a 15-minute SLA at an extra per-GB fee) only when an RPO or compliance rule requires it. Existing objects need a one-time Batch Replication job; live replication covers only new writes.
  • "DynamoDB" global tables: Every replica Region pays for storage and for replicated writes (priced the same as single-Region writes since November 2024) plus cross-Region transfer. Keep replicas only in Regions that serve traffic or need a seconds-level RPO; for a DR Region where an RPO of hours is fine, a cross-Region backup copy is far cheaper than a permanently written replica.
  • Relational databases: "Aurora Global Database" keeps a running secondary cluster plus replicated write traffic for a seconds-level RPO; cross-Region automated backup or snapshot copies cost only storage and transfer for an RPO of hours. Match the mechanism to the stated RPO and RTO, not to the strongest option available.

⚠️ Exam Trap: Enabling S3 Replication Time Control, or keeping an always-written global table replica, "to be safe" when the stated RPO is hours. Both buy a seconds-to-minutes RPO the requirement never asked for; a filtered replication rule into a colder class, or a scheduled cross-Region backup copy, meets an RPO of hours for a fraction of the cost.

See how it connects
Alvin Varughese
Written byAlvin Varughese
Founder•26 professional certifications earned