EDB Releases PGD 6.4 with Quorum Commit, Bringing True Distributed Consistency to Mission-Critical Postgres

Native connection pooling and large object support further simplify high-durability architectures and strengthen a sovereign foundation for banking, payments, and national infrastructure

WILMINGTON, Del., May 28, 2026 /PRNewswire/ -- EnterpriseDB (EDB), the leading sovereign AI and data company, today announced EDB Postgres Distributed (PGD) 6.4, the distributed transactional foundation of EDB Postgres^® AI. The release introduces three landmark capabilities: Quorum Commit for true cross-node distributed consistency, integrated connection pooling through the native Connection Manager, and full support for PostgreSQL large objects. Together, they deliver the durability guarantees and architectural simplicity that Tier 1 financial and infrastructure applications demand.

Demand for sovereign control over critical data is accelerating. Gartner forecasts that by 2030, more than 75% of European and Middle Eastern enterprises will geopatriate workloads — up from less than 5% in 2025. AI and data sovereignty are clear strategic imperatives.

But control over where data lives is only half the picture; organizations also need certainty about its state. PGD 6.4 meets that demand, bringing the strongest distributed consistency Postgres has ever offered.

Quorum Commit: The end of the distributed double-spend

The most consequential addition in PGD 6.4 is Quorum Commit; a pre-commit coordination mechanism that enforces a single, coherent global truth across every node in the cluster before any transaction is locally committed. Unlike traditional synchronous replication, which confirms only that a replica has received a record, Quorum Commit coordinates the content of concurrent writes across data centers in real time.

Consider a credit card shared across geographies. Under standard replication, two simultaneous charges can each pass a local balance check and commit independently, leaving the bank to resolve an overspend after the fact. With Quorum Commit, the cluster reaches consensus before either charge completes: The first transaction wins the quorum vote and is committed; the second sees the updated balance and is declined at the point of sale without any post-commit cleanup.

"Postgres has become the de facto standard for modern applications. Yet until now, organizations running high-value workloads in banking, payments, or telecom were forced to fall back on legacy enterprise RDBMS for their strongest consistency requirements. PGD 6.4 changes that, delivering the same distributed consistency those systems were built on, now fully Postgres native," said Jozef de Vries, SVP, Database Engineering at EDB.

For financial institutions migrating away from proprietary databases, Quorum Commit removes the last major technical barrier: a fully Postgres-native path to the same consistency guarantees offered by high-end commercial systems with none of the licensing overhead.

Connection Manager pooling: Enterprise scale, zero external dependencies

PGD 6.4 extends the built-in Connection Manager introduced in PGD 6.0 with native connection pooling, eliminating the need for external poolers such as pgBouncer in most production topologies. Because the Connection Manager is integrated directly with PGD's Raft consensus layer, it provides capabilities no external proxy can match: cluster-aware routing at both the cluster and region levels, automatic failover routing that responds to consensus changes in real time, and unified observability through PostgreSQL's own logging and monitoring views.

The architectural payoff is significant at scale. Organizations running hundreds of distinct clusters such as payment processors operating across thousands of CPU cores gain a simpler, more resilient stack with one fewer component to deploy, version, and monitor. Operational complexity is reduced without any sacrifice in throughput or connection headroom.

Large object support: Broadening the scope of distributed Postgres

PGD 6.4 adds full replication support for PostgreSQL large objects—binary data structures that were previously unavailable in distributed Postgres environments. This brings a class of workloads into the PGD fold that were previously excluded: applications managing scanned documents, image archives, binary payloads, or mixed transactional and unstructured data within a single database. Financial institutions, government agencies, and healthcare operators commonly maintain exactly these hybrid schemas alongside their core transactional tables.

PGD 6.4 is now available as a stand-alone distribution and will be available as part of EDB Postgres AI in June. For technical documentation, release notes, and upgrade guidance, visit enterprisedb.com/docs/pgd.

About EDB
EDB Postgres^® AI (EDB PG AI) is the first open, enterprise-grade sovereign data and AI platform—secure, compliant, and scalable, on-premises and across clouds. Built on Postgres, the world's leading database, EDB PG AI unifies transactional, analytical, and AI workloads, enabling organizations to operationalize their data and LLMs while maintaining control over sovereign environments. EDB PG AI is supported by a global partner network and delivers up to 99.999% availability as well as hybrid management and a built-in AI factory. As one of the most active contributors to the PostgreSQL project, EDB is deeply invested in the vitality of the global community. To learn more, visit www.enterprisedb.com.

EnterpriseDB and EDB are registered trademarks of EnterpriseDB Corporation. Postgres and PostgreSQL are registered trademarks of the PostgreSQL Community Association of Canada and used with their permission. All other trademarks are owned by their respective owners.

Media Contact:
Anwar Ahmad
[email protected]
(347) 206-5651

SOURCE EnterpriseDB