Open Source DocumentDB 1.0 Reaches General Availability

The Linux Foundation and contributing partners have reached general availability for DocumentDB 1.0, an open source, MIT-licensed document database built directly on PostgreSQL with MongoDB wire protocol compatibility. Microsoft originally announced the initiative in early 2025 to give developers full architectural visibility into a PostgreSQL-backed NoSQL datastore, subsequently placing it under the Linux Foundation alongside backing database organizations such as AWS, Cockroach Labs, Google, Snowflake, and Yugabyte. According to InfoQ's report, the release allows existing MongoDB drivers, query patterns, and client tools to operate without modifications while shifting underlying persistence to standard PostgreSQL engines.

This release analysis reviews the architectural design, licensing model, deployment packaging, early performance metrics, and known operational tradeoffs in version 1.0. Engineering leads evaluating alternative document backends, database administrators managing PostgreSQL infrastructure, and software developers seeking permissive alternatives to MongoDB's Server Side Public License can examine how the PostgreSQL extension and protocol gateway integrate into production clusters.

01PostgreSQL extension architecture and protocol gateway

DocumentDB splits its runtime into two core functional components: a PostgreSQL extension that provides MongoDB API-compatible data processing, and a translation gateway that converts incoming MongoDB protocol wire requests into native PostgreSQL calls. The version 1.0 general availability milestone prioritizes operational hardening across the query stack, incorporating specific fixes for aggregation pipelines, compound indexing structures, and geospatial lookups.

The runtime supports PostgreSQL versions 15 through 18 via prebuilt container images, while native Linux software packages are distributed specifically for PostgreSQL 17 and 18 installations. Production orchestration is assisted by an official Kubernetes operator designed to manage automated cluster lifecycle events. The permissive MIT license permits commercial use, private modification, and third-party managed hosting without enforcing source code disclosures required under MongoDB's SSPL.

Patty Chow, product manager at Microsoft, pointed out that alternative implementations give developers an opportunity to evaluate different approaches to storage and deployment while preserving familiar drivers and query patterns. FerretDB also participates in this ecosystem as a separate project that utilizes DocumentDB extensions as a backend storage engine while maintaining its own MongoDB-compatible interface.

02Benchmark performance throughput and query latency

Early benchmark testing conducted by Layerbase using standard YCSB mixed workloads revealed distinct behavioral characteristics across both PostgreSQL-oriented implementations. DocumentDB achieved 20% to 33% higher read throughput in these initial tests, whereas FerretDB maintained lower tail latency for write and update operations during mixed workloads, though analysts noted these figures remain inconclusive.

Independent validation of performance figures remains ongoing within open community tracking channels. An open GitHub issue logged by community testers reported higher overall query latency across several application workloads in DocumentDB 1.0 compared directly against MongoDB 9.0.4.

03Container images and Kubernetes deployment packages

Engineers deploying DocumentDB 1.0 can pull official container images that bundle support across PostgreSQL 15, 16, 17, and 18 directly into orchestration pipelines. Systems running bare-metal Linux hosts or virtual instances on PostgreSQL 17 or 18 can deploy dedicated distribution packages alongside the standalone gateway daemon.

docker pull ghcr.io/microsoft/documentdb:1.0.0
docker run -d --name documentdb -p 27017:27017 -p 5432:5432 ghcr.io/microsoft/documentdb:1.0.0

To configure application services, developers direct standard MongoDB connection strings toward the DocumentDB translation gateway port rather than a proprietary MongoDB daemon. Standard administrative tools, MongoDB shell clients, and ecosystem drivers link directly into the exposed endpoint without bespoke connection adapters.

mongosh "mongodb://localhost:27017/appdb"

04Community naming debates and ecosystem overlap

The choice of the DocumentDB moniker generated significant discussion across community forums due to established commercial naming overlap. Commenters on Hacker News highlighted frequent confusion with Amazon DocumentDB, while others on Reddit noted ambiguity surrounding historical product names like Azure DocumentDB and Azure Cosmos DB despite an official naming disambiguation guide published by the maintainers.

Architectural clarity also requires distinguishing DocumentDB from FerretDB, which implements its own MongoDB-compatible proxy layer over DocumentDB PostgreSQL extensions. User discussions on Reddit expressed varied experiences regarding driver parity across external compatibility layers, noting historical gaps when translating complex NoSQL operations onto relational backends.

The team has documented upgrade paths alongside known driver issues within the version 1.0 release notes. Engineering teams migrating existing codebases must account for edge-case query parsing behaviors and ongoing performance verifications between MongoDB wire protocol revisions.

05Release roadmap and long-term support branches

Project maintainers plan to partition upcoming releases into distinct long-term support branches alongside active development tracks. The formal versioning cadence and long-term support maintenance policy remain under active design discussion among the Linux Foundation member organizations.

Governance support across AWS, Cockroach Labs, Google, Snowflake, and Yugabyte signals broad industry participation in refining the storage extension layer. Further updates will address runtime tuning, transaction isolation behavior, and tighter integration across downstream projects utilizing the PostgreSQL extension stack.