Enhancing Disaster Recovery with Multi-Region Data Replication in Public Sector Databases

Abstract

As public sector organizations continue to migrate critical databases to cloud infrastructure, multi-region data replication for resilient disaster recovery (DR) is becoming increasingly important. This study provides a zero-based evaluation of multi-region replication for government systems in the context of RTO, RPO, MAPD, replication lag, and transaction commit assurance during failure simulations. A case study based on a custom benchmark framework implemented across three regions and two clouds showed active-active multi-region replication achieved 42% lower RTO and 60% better RPO compliance than previous DR standards. Replicated lag during failover also improved from an average of 4.2 seconds to 1.6 seconds with a 23% increase in stressed commit success rates. This research also highlights cross-region latency deviations as well as quantifying SLA breaches relating to various outage types. Centros emphasizes the advantages of active-active replication for enriched fault tolerance, reduced data loss, and enhanced service continuity for public sector data systems.