ORACLE MAKES GLOBALLY DISTRIBUTED EXADATA DATABASE

Abstract

The service automatically distributes, stores, and synchronizes data across locations so applications can remain online even if an entire region experiences an outage, while also helping customers meet data residency requirements in regulated industries. The company highlights several categories that benefit directly: high-volume transaction processing, long-running analytics on streaming data, vector-enabled AI workloads that need proximity to business records, and global e-commerce systems that cannot tolerate downtime. Policy management for data residency, placement near end users, and cross-region fault tolerance aims to address a wide range of compliance and experience requirements under one managed control plane.

Oracle has announced the general availability of Oracle Globally Distributed Exadata Database on Exascale Infrastructure, a new managed service designed to run mission-critical databases that span multiple Oracle Cloud Infrastructure (OCI) regions without application rewrites. The service automatically distributes, stores, and synchronizes data across locations so applications can remain online even if an entire region experiences an outage, while also helping customers meet data residency requirements in regulated industries. The announcement on August 7, 2025, positions the offering squarely at teams that need consistent performance and availability characteristics for transactional, analytics, and emerging agent-based AI workloads that demand access to current operational data. A core design element of the service is a serverless control plane that elastically adjusts to workload change. Instead of provisioning fixed capacity tied to a single region, teams get a database that scales up or down as demand spikes or subsides—minimizing idle resources and reducing operational overhead. For organizations that experience unpredictable demand profiles (for example, retail during seasonal peaks or global consumer applications with diurnal load patterns), the combination of automatic elasticity and cross-region distribution lowers the need to over-provision for worst-case scenarios. Under the hood, Oracle employs Raft replication and an Active/Active/Active topology to coordinate writes and reads across data centers. This architecture is intended to provide fast, zero-data-loss failover characteristics while keeping read and write latencies low for users by placing data near them. The company highlights several categories that benefit directly: high-volume transaction processing, long-running analytics on streaming data, vector-enabled AI workloads that need proximity to business records, and global e-commerce systems that cannot tolerate downtime. The system's policy-driven placement allows administrators to encode residency and proximity rules so that sensitive tables remain confined to approved jurisdictions while less sensitive datasets can be positioned for performance. Compatibility with the full Oracle Database and SQL stack is a deliberate choice. By keeping the database semantics and SQL surface area consistent, Oracle reduces the need for major application rewrites or refactors when moving to a distributed footprint. Existing Oracle skills, tools, and application frameworks remain applicable, which should shorten adoption timelines for large estates already invested in Exadata-backed systems of record. For teams pursuing AI integration, Oracle points out that Exascale's elasticity and vector search support can be combined with data distribution policies to keep embeddings and proprietary customer data governed and performant in the same platform. Early customer commentary focuses on latency and availability. As one large online payments operator noted, the long-standing requirement has been "lightning-fast response times and mission-critical availability," and the new service is pitched as a path to even faster responses as global demand scales. For operations leaders, the selling proposition is measured not only in failover metrics, but also in simplifying multi-region database administration that previously demanded intricate orchestration across many clusters. Oracle's messaging frames the capability as taking distributed databases—historically complex and expensive—and making them accessible to a broader set of customers via a pay-for-what-you-use consumption model. While the distributed database landscape has multiple approaches—ranging from sharded SQL engines to globally consistent NewSQL stores—Oracle's entry centers on delivering these behaviors within its flagship database environment and Exascale infrastructure. For organizations already running Exadata, the path to a globally distributed posture becomes incremental rather than a ground-up platform change. Policy management for data residency, placement near end users, and cross-region fault tolerance aims to address a wide range of compliance and experience requirements under one managed control plane. About Oracle Oracle is a global provider of database software, cloud applications, and cloud infrastructure. Its database portfolio includes Exadata-based systems and managed services delivered on Oracle Cloud Infrastructure. For more information, visit www.oracle.com.