openGauss
activeby openGauss Community (Huawei-led) · since 2020 · also known as open Gauss
Huawei-originated open-source RDBMS forked from PostgreSQL, with in-memory and column engines.
Overview
openGauss is an enterprise relational database that Huawei open-sourced in 2020 from the kernel of its GaussDB, itself forked from PostgreSQL 9.2. It has diverged substantially: a thread-pool architecture instead of process-per-connection, the MOT in-memory engine, row (Astore/Ustore) and column storage, AI-assisted tuning and indexing, and in 7.0 a DataVec vector engine. It is widely deployed in China across banking, telecom and government as part of domestic-database programs, with many vendors shipping commercial distributions on top of it. Outside China, adoption, documentation quality in English and ecosystem support are limited.
Strengths & trade-offs
Strengths
- +Thread-pool model handles many connections without an external pooler
- +MOT in-memory engine for latency-sensitive OLTP tables
- +Row and column storage in one engine for mixed workloads
- +Strong security features: TDE, full-encrypted queries, row-level security
- +Large Chinese ecosystem of commercial distributions and ARM (Kunpeng) optimization
Trade-offs
- −Forked from PostgreSQL 9.2, so modern Postgres extensions rarely work unmodified
- −English documentation and community support are limited
- −Linux-only, best tested on openEuler and Kunpeng hardware
- −Few managed offerings outside Huawei Cloud
Great for
- +Chinese banking and telecom core systems
- +Domestic-database replacements for Oracle in China
- +HTAP workloads mixing row and column tables
- +On-prem deployments on ARM servers
Think twice for
- −Teams relying on the PostgreSQL extension ecosystem
- −Organizations outside China needing vendor support
- −Cloud-native deployments on Western hyperscalers
Capabilities
Native or official first-party support · 16 of 25 tracked
Scale & consistency
- Comfortable data size
- Large · ≤ 1 PB
- Horizontal scaling
- None
- Multi-region
- Read replicas
- Transactions
- Full ACID
- Consistency
- Strong
- Schema
- Strict
- Architecture
- Single node, Primary replica, Client server, In memory
- Query languages
- SQL, PL/pgSQL
Primary-standby with quorum replication; distributed scale-out is offered through ShardingSphere-based or commercial GaussDB editions.
Deployment
available on · Huawei Cloud
Ecosystem
Alternatives to openGauss
The extensible, standards-first open-source relational database.
opengauss vs postgresql →Huawei's distributed enterprise RDBMS built on openGauss, aimed at replacing Oracle in Chinese finance.
opengauss vs huawei-gaussdb →Chinese PostgreSQL-derived RDBMS with Oracle, MySQL and SQL Server compatibility modes.
opengauss vs kingbase →China's leading home-grown RDBMS; DM9 (2026) adds distributed, HTAP and AI features.
opengauss vs dameng →Alipay-proven distributed SQL with MySQL (and Oracle, in Enterprise) compatibility, HTAP and vector search.
opengauss vs oceanbase →Russian commercial PostgreSQL distribution with multimaster, 64-bit XIDs and compressed storage.
opengauss vs postgres-pro →openGauss FAQ
What is openGauss?
openGauss is a relational database by openGauss Community (Huawei-led), first released in 2020. openGauss is an enterprise relational database that Huawei open-sourced in 2020 from the kernel of its GaussDB, itself forked from PostgreSQL 9.2. It has diverged substantially: a thread-pool architecture instead of process-per-connection, the MOT in-memory engine, row (Astore/Ustore) and column storage, AI-assisted tuning and indexing, and in 7.0 a DataVec vector engine. It is widely deployed in China across banking, telecom and government as part of domestic-database programs, with many vendors shipping commercial distributions on top of it. Outside China, adoption, documentation quality in English and ecosystem support are limited.
Is openGauss open source?
Yes. openGauss is open source. License: Mulan Permissive Software License v2.
Does openGauss support ACID transactions?
Yes. openGauss supports full ACID transactions, with strong consistency.
Does openGauss scale horizontally?
openGauss does not scale horizontally; it runs on a single node, so you scale it up rather than out. It is comfortable at large scale (≤ 1 PB). It supports cross-region read replicas.
How can I run openGauss?
You can self-host it. Managed offerings include GaussDB. It is available on Huawei Cloud.
What is openGauss best for?
openGauss is a strong fit for chinese banking and telecom core systems; domestic-database replacements for Oracle in China; and HTAP workloads mixing row and column tables. Think twice for teams relying on the PostgreSQL extension ecosystem.
What are the best alternatives to openGauss?
Common alternatives to openGauss include PostgreSQL, Huawei GaussDB, KingbaseES, Dameng (DM) and OceanBase. The right choice depends on your workload, scale and hosting; WhichDB's smart match ranks them against your requirements.
How popular is openGauss?
openGauss ranks #357 of 844 databases on the WhichDB Score (6.7/100), and #114 among relational databases. The score combines public adoption signals such as package downloads, GitHub activity, job posts and developer discussion.