我的征尘是星辰大海。。。

The dirt and dust from my pilgrimage forms oceans of stars...

-------当记忆的篇章变得零碎，当追忆的图片变得模糊，我们只能求助于数字存储的永恒的回忆

作者:黄教授

手机视频列表

国产数据库行业现状扫描

视频

音频

原始脚本

国产数据库行业现状白描、格局、乱象与底层逻辑解析。
国产数据库行业自信创政策全面推进以来，迎来爆发式增长。
市场参与者短时间内激增至167家，形成头部厂商与中小厂商混战，自研产品与开源改造产品并存的复杂格局。
本文以中立第三方视角。
对行业现状形成根源、不同场景选型规则及未来趋势进行客观纪实与深度解析，还原产业真实面貌。
一。
早期政务与国企数据库采购需求与选型错配的历史现状。
在信创国产化启动前，国内党政机关、事业单位及多数国企的核心数据管理系统。
普遍采用 Oracle  IBM  DB 二等国外商用数据库。
这一选型现状与实际业务需求形成明显错配，是行业长期存在的既定事实。
从业务场景来看，基层党政机关、普通事业单位的核心数据需求集中于人事档案、工资报表、政务台账、公文管理等基础场景。
无高并发访问、海量交易处理、跨地域容灾等严苛要求，业务复杂度极低，轻量级本地数据库即可满足运行需求。
而大型央企、传统国企的核心业务多围绕 ERP 系统、生产物料管理、内部办公系统展开，交易频次与并发量极远低于金融、互联网行业。
对数据库的高性能分布式能力要求有限。
在此背景下，大规模采用国外高端商用数据库，核心源于早期技术选型的惯性思维。
行业采购惯例，以及国外厂商长期的市场布局。
此类采购产生的高额授权费、年度运维服务费，持续形成财政资金外流。
且核心技术与源码控制权掌握在外企手中，存在数据安全与供应链潜在风险。
这也是后续信创国产化政策推进的核心背景。
二、当前行业格局。
167家厂商分层，技术实力两极分化。
截至2026年，国内注册运营的国产数据库厂商已达167家。
看似竞争激烈，实则技术路径与实力呈现清晰分层，行业集中度极低，中小厂商占比远超头部企业。
第一梯队全资源核心厂商。
该梯队厂商数量极少，具备完全自主知识产权的数据库内核，无开源数据库底层代码依赖，研发周期长，技术投入大。
产品经过大规模高并发高可用场景验证。
代表厂商为蚂蚁集团旗下 OceanBase 华为 GaussDB 产品主要覆盖金融核心交易。
央企核心系统、高端政务等对稳定性、安全性、分布式能力要求极高的场景，是行业内技术壁垒最高、市场认可度最强的阵营。
第二梯队开源深度改造合规厂商。
此类厂商基于 PostgreSQL PG 开源数据库进行深度二次开发。
PG 采用 BSD 类宽松开源协议。
允许商用闭源代码修改与品牌重塑，无开源传染限制，属于合规合法的技术路径。
厂商会重写存储引擎。
事务管理、国产软硬件适配模块，产品满足信创测评要求，具备稳定的运维与迭代能力。
主要服务于地方政务、中型国企。
传统行业数字化场景是当前信创市场的主力参与者之一。
第三梯队，浅层改造与套壳厂商，该梯队占行业总量的百分之九十左右。
是167家厂商的主体构成。
多数厂商基于 MySQL 开源数据库进行浅层修改，无视 MySQL GPL v 二强开源协议要求。
仅更换界面、修改参数、重新包装品牌，谎称自主研发，规避合规审核。
部分厂商基于 PG 做表层优化。
无核心研发与长期迭代能力，多为中小型初创企业。
核心目标为承接地方信创项目与政策补贴，无持续技术投入规划。
三、行业乱象根源，政策体制与利益交织的客观成因。
167家厂商混战的行业乱象，并非技术层面的供需失衡。
而是政策落地、地方考核、招标机制多重因素共同作用的结果，属于产业发展初期的阶段性特征。
其一，信创政策与地方政绩考核导向。
信创国产化率是各级政务部门的硬性考核指标。
各地未完成指标，扶持本土数字经济企业，允许本土中小数据库厂商参与项目招标，形成一地一厂的分散格局。
行业准入门槛阶段性降低，大量中小厂商得以入场。
其二，项目拆分式招标机制，政务与国企数字化项目。
普遍将 ERP 系统、数据库、中间件、运维服务拆分招标，而非采用一体化解决方案。
拆分招标后，数据库成为独立分包项目。
预算单独核算，为中小厂商提供了生存空间，也避免了头部厂商垄断市场，间接导致参与者数量激增。
其三。
开源技术降低创业门槛。
MySQL  PostgreSQL 等开源数据库的普及，大幅降低了数据库产品的研发门槛，无需从零搭建内核。
仅需浅层改造即可推出产品，使得无核心技术的团队能快速完成产品包装，参与市场竞争。
其次，政策补贴与订单红利驱动。
信创相关产业补贴、政务定向订单成为中小厂商的核心盈利来源。
此类厂商无需参与市场化竞争，仅需满足基础合规要求。
即可获得项目收益，形成重订单轻研发的经营模式。
从行业发展规律来看，第三梯队厂商普遍采用一锤子买卖的经营模式，项目交付后仅提供基础运维。
多数无后续技术升级能力，随着项目周期结束，大量厂商将面临淘汰或注销，后期大概率出现行业整合。
基于同源开源代码的中小厂商产品。
将逐步由少数具备运维能力的企业统一承接后续服务。
四分场景选型规则，政企与民企的差异化合规要求国内数据库市场。
按使用主体划分，形成两套完全独立的选型规则，无统一标准，核心差异在于合规性约束与市场化选择。
党政机关与国企合规优先，强制国产化党政机关、事业单位、央企及地方国企授信创政策、网络安全法数据安全法严格约束，数据库选型必须纳入国家信创产品目录，优先选用自主可控的国产数据库。
禁止直接采用国外商用数据库及未合规改造的开源数据库。
其中，基层单位以满足基础合规要求为主，选用第二、第三梯队产品即可。
核心政务、军工、大型央企对数据安全与系统稳定性要求较高。
优先选用第一梯队全自研产品。
此类选型不以业务性能需求为唯一标准，核心目标是实现技术自主可控，将财政资金转化为本土产业扶持资金。
民营企业无强制约束，市场化自主选择。
普通民营企业无强制性国产化与信创合规要求。
数据库选型完全依据自身业务规模、成本预算、技术需求决定。
中小微企业、轻量业务场景可直接选用 SQLite。
开源 MySQL 开源 PostgreSQL 等免费产品。
中型民企可根据需求选用商用数据库或国产改造数据库。
仅当民营企业承接政府项目、国企供应链业务时，需按照甲方要求选用信创目录内国产数据库，其余场景无任何限制。
五、行业未来趋势，泡沫出清，头部集中化。
当前167家厂商的混战格局，属于国产数据库产业发展的过渡阶段。
长期来看，行业将逐步走向规范化、集中化。
一方面，浅层改造无研发能力的中小厂商缺乏持续经营能力。
随着政策补贴收紧，项目验收标准提高，将批量退出市场，行业泡沫逐步出清。
另一方面，头部全自研厂商与优质开源改造厂商将通过技术迭代。
项目整合、并购重组，占据主流市场，最终形成少数头部全自研厂商加多家垂直领域专精厂商的稳定格局。
从产业核心价值来看，信创国产化的本质是摆脱国外技术垄断，保障国家数据安全，扶持本土数字技术产业。
当前行业乱象并非政策初衷，而是初期落地的阶段性问题。
随着行业标准完善，技术门槛提升。
国产数据库将逐步回归技术本质，从合规导向转向实力导向，实现真正意义上的自主可控与市场化发展。

修正脚本

国产数据库行业现状白描、格局、乱象与底层逻辑解析。
国产数据库行业自信创政策全面推进以来，迎来爆发式增长。
市场参与者短时间内激增至167家，形成头部厂商与中小厂商混战、自研产品与开源改造产品并存的复杂格局。
本文以中立第三方视角，对行业现状、形成根源、不同场景选型规则及未来趋势进行客观纪实与深度解析，还原产业真实面貌。
一、早期政务与国企数据库采购需求与选型错配的历史现状。
在信创国产化启动前，国内党政机关、事业单位及多数国企的核心数据管理系统，普遍采用 Oracle、IBM、DB2 等国外商用数据库。
这一选型现状与实际业务需求形成明显错配，是行业长期存在的既定事实。
从业务场景来看，基层党政机关、普通事业单位的核心数据需求集中于人事档案、工资报表、政务台账、公文管理等基础场景。
无高并发访问、海量交易处理、跨地域容灾等严苛要求，业务复杂度极低，轻量级本地数据库即可满足运行需求。
而大型央企、传统国企的核心业务多围绕 ERP 系统、生产物料管理、内部办公系统展开，交易频次与并发量远低于金融、互联网行业。
对数据库的高性能分布式能力要求有限。
在此背景下，大规模采用国外高端商用数据库，核心源于早期技术选型的惯性思维、行业采购惯例，以及国外厂商长期的市场布局。
此类采购产生的高额授权费、年度运维服务费，持续形成财政资金外流。
且核心技术与源码控制权掌握在外企手中，存在数据安全与供应链潜在风险。
这也是后续信创国产化政策推进的核心背景。
二、当前行业格局。
167家厂商分层，技术实力两极分化。
截至2026年，国内注册运营的国产数据库厂商已达167家。
看似竞争激烈，实则技术路径与实力呈现清晰分层，行业集中度极低，中小厂商占比远超头部企业。
第一梯队全自研核心厂商。
该梯队厂商数量极少，具备完全自主知识产权的数据库内核，无开源数据库底层代码依赖，研发周期长，技术投入大。
产品经过大规模高并发高可用场景验证。
代表厂商为蚂蚁集团旗下 OceanBase、华为 GaussDB，产品主要覆盖金融核心交易、央企核心系统、高端政务等对稳定性、安全性、分布式能力要求极高的场景，是行业内技术壁垒最高、市场认可度最强的阵营。
第二梯队开源深度改造合规厂商。
此类厂商基于 PostgreSQL PG 开源数据库进行深度二次开发。
PG 采用 BSD 类宽松开源协议。
允许商用闭源代码修改与品牌重塑，无开源传染限制，属于合规合法的技术路径。
厂商会重写存储引擎、事务管理、国产软硬件适配模块，产品满足信创测评要求，具备稳定的运维与迭代能力。
主要服务于地方政务、中型国企、传统行业数字化场景，是当前信创市场的主力参与者之一。
第三梯队，浅层改造与套壳厂商，该梯队占行业总量的百分之九十左右。
是167家厂商的主体构成。
多数厂商基于 MySQL 开源数据库进行浅层修改，无视 MySQL GPL v2 强开源协议要求。
仅更换界面、修改参数、重新包装品牌，谎称自主研发，规避合规审核。
部分厂商基于 PG 做表层优化。
无核心研发与长期迭代能力，多为中小型初创企业。
核心目标为承接地方信创项目与政策补贴，无持续技术投入规划。
三、行业乱象根源，政策体制与利益交织的客观成因。
167家厂商混战的行业乱象，并非技术层面的供需失衡。
而是政策落地、地方考核、招标机制多重因素共同作用的结果，属于产业发展初期的阶段性特征。
其一，信创政策与地方政绩考核导向。
信创国产化率是各级政务部门的硬性考核指标。
各地为完成指标，扶持本土数字经济企业，允许本土中小数据库厂商参与项目招标，形成一地一厂的分散格局。
行业准入门槛阶段性降低，大量中小厂商得以入场。
其二，项目拆分式招标机制，政务与国企数字化项目普遍将 ERP 系统、数据库、中间件、运维服务拆分招标，而非采用一体化解决方案。
拆分招标后，数据库成为独立分包项目。
预算单独核算，为中小厂商提供了生存空间，也避免了头部厂商垄断市场，间接导致参与者数量激增。
其三、开源技术降低创业门槛。
MySQL、PostgreSQL 等开源数据库的普及，大幅降低了数据库产品的研发门槛，无需从零搭建内核。
仅需浅层改造即可推出产品，使得无核心技术的团队能快速完成产品包装，参与市场竞争。
其四、政策补贴与订单红利驱动。
信创相关产业补贴、政务定向订单成为中小厂商的核心盈利来源。
此类厂商无需参与市场化竞争，仅需满足基础合规要求。
即可获得项目收益，形成重订单轻研发的经营模式。
从行业发展规律来看，第三梯队厂商普遍采用一锤子买卖的经营模式，项目交付后仅提供基础运维。
多数无后续技术升级能力，随着项目周期结束，大量厂商将面临淘汰或注销，后期大概率出现行业整合。
基于同源开源代码的中小厂商产品。
将逐步由少数具备运维能力的企业统一承接后续服务。
四、分场景选型规则，政企与民企的差异化合规要求。国内数据库市场，按使用主体划分，形成两套完全独立的选型规则，无统一标准，核心差异在于合规性约束与市场化选择。
党政机关与国企合规优先，强制国产化。党政机关、事业单位、央企及地方国企受信创政策、网络安全法、数据安全法严格约束，数据库选型必须纳入国家信创产品目录，优先选用自主可控的国产数据库。
禁止直接采用国外商用数据库及未合规改造的开源数据库。
其中，基层单位以满足基础合规要求为主，选用第二、第三梯队产品即可。
核心政务、军工、大型央企对数据安全与系统稳定性要求较高。
优先选用第一梯队全自研产品。
此类选型不以业务性能需求为唯一标准，核心目标是实现技术自主可控，将财政资金转化为本土产业扶持资金。
民营企业无强制约束，市场化自主选择。
普通民营企业无强制性国产化与信创合规要求。
数据库选型完全依据自身业务规模、成本预算、技术需求决定。
中小微企业、轻量业务场景可直接选用 SQLite、开源 MySQL、开源 PostgreSQL 等免费产品。
中型民企可根据需求选用商用数据库或国产改造数据库。
仅当民营企业承接政府项目、国企供应链业务时，需按照甲方要求选用信创目录内国产数据库，其余场景无任何限制。
五、行业未来趋势，泡沫出清，头部集中化。
当前167家厂商的混战格局，属于国产数据库产业发展的过渡阶段。
长期来看，行业将逐步走向规范化、集中化。
一方面，浅层改造无研发能力的中小厂商缺乏持续经营能力。
随着政策补贴收紧，项目验收标准提高，将批量退出市场，行业泡沫逐步出清。
另一方面，头部全自研厂商与优质开源改造厂商将通过技术迭代、项目整合、并购重组，占据主流市场，最终形成少数头部全自研厂商加多家垂直领域专精厂商的稳定格局。
从产业核心价值来看，信创国产化的本质是摆脱国外技术垄断，保障国家数据安全，扶持本土数字技术产业。
当前行业乱象并非政策初衷，而是初期落地的阶段性问题。
随着行业标准完善，技术门槛提升。
国产数据库将逐步回归技术本质，从合规导向转向实力导向，实现真正意义上的自主可控与市场化发展。

英文翻译

**White Paper on the Current State of the Domestic Database Industry: Landscape, Chaos, and Underlying Logic**

Since the comprehensive rollout of the Xinchuang policy, the domestic database industry has experienced explosive growth.  
The number of market participants surged to 167 in a short period, forming a complex landscape where leading vendors compete with small and medium-sized players, and self-developed products coexist with open-source-based modifications.  
From a neutral third-party perspective, this article provides an objective record and in-depth analysis of the industry's current state, root causes, selection rules for different scenarios, and future trends, restoring the true picture of the industry.

**I. Historical Status: Mismatch Between Procurement Needs and Selection in Early Government and State-Owned Enterprise Database Purchases**

Before the launch of Xinchuang localization, core data management systems in China's party and government organs, public institutions, and most state-owned enterprises commonly adopted foreign commercial databases such as Oracle, IBM DB2, and others.  
This selection status created an obvious mismatch with actual business needs, a long-standing fact in the industry.  
From a business scenario perspective, the core data needs of grassroots party and government organs and ordinary public institutions are concentrated on basic scenarios such as personnel files, salary reports, government ledgers, and document management.  
There are no stringent requirements for high-concurrency access, massive transaction processing, or cross-regional disaster recovery. Business complexity is extremely low, and lightweight local databases are sufficient to meet operational needs.  
Meanwhile, the core businesses of large central enterprises and traditional state-owned enterprises mainly revolve around ERP systems, production material management, and internal office systems, with transaction frequency and concurrency far lower than those in the financial and internet industries.  
The demand for high-performance distributed capabilities in databases is limited.  
Against this backdrop, the large-scale adoption of high-end foreign commercial databases stems primarily from the inertial thinking of early technology selection, industry procurement practices, and the long-term market presence of foreign vendors.  
Such procurements generate high licensing fees and annual maintenance service costs, leading to a continuous outflow of fiscal funds.  
Moreover, core technology and source code control remain in the hands of foreign companies, posing potential risks to data security and supply chain stability.  
This is also the core background driving the subsequent Xinchuang localization policy.

**II. Current Industry Landscape: 167 Vendors Stratified, Technological Strength Polarized**

As of 2026, the number of domestic database vendors registered and operating in China has reached 167.  
The landscape appears fiercely competitive, but in reality, technological paths and strengths are clearly stratified. Industry concentration is extremely low, with small and medium-sized vendors far outnumbering leading enterprises.

**First Tier: Full Self-Developed Core Vendors**  
This tier contains very few vendors. They possess fully independent intellectual property rights for their database kernels, with no dependency on underlying open-source database code. Their R&D cycles are long and technology investments are substantial. Their products have been verified in large-scale, high-concurrency, high-availability scenarios.  
Representative vendors include OceanBase under Ant Group and Huawei GaussDB. Their products mainly cover scenarios requiring extremely high stability, security, and distributed capabilities, such as core financial transactions, central enterprise core systems, and high-end government affairs. This is the camp with the highest technological barriers and strongest market recognition.

**Second Tier: Deep Open-Source Modification Compliant Vendors**  
These vendors conduct deep secondary development based on the PostgreSQL (PG) open-source database.  
PG adopts a BSD-like permissive open-source license, allowing commercial closed-source code modifications and rebranding, with no open-source copyleft restrictions. This is a compliant and legal technical path.  
Vendors rewrite storage engines, transaction management, and domestic software/hardware adaptation modules. Their products meet Xinchuang evaluation requirements and possess stable operation, maintenance, and iteration capabilities.  
They primarily serve local government affairs, medium-sized state-owned enterprises, and digitalization scenarios in traditional industries. They are among the main participants in the current Xinchuang market.

**Third Tier: Shallow Modification and Shell Vendors**  
This tier accounts for about 90% of the total number of vendors in the industry.  
It constitutes the main body of the 167 vendors.  
Most of these vendors make shallow modifications based on the MySQL open-source database, ignoring the strict requirements of the MySQL GPL v2 strong open-source license.  
They merely replace the interface, modify parameters, and rebrand the product, falsely claiming independent R&D while evading compliance review.  
Some vendors perform superficial optimizations based on PG.  
They lack core R&D capabilities and long-term iteration ability, and are mostly small or medium-sized startups.  
Their core goal is to undertake local Xinchuang projects and obtain policy subsidies, with no plans for sustained technology investment.

**III. Root Causes of Industry Chaos: Objective Factors Where Policy Systems and Interests Intertwine**

The industry chaos of 167 vendors in a melee is not due to a supply-demand imbalance at the technical level.  
Rather, it is the result of multiple factors including policy implementation, local government performance assessments, and bidding mechanisms—a characteristic of the early stage of industry development.

**First, Xinchuang Policy and Local Government Performance Assessment Orientation**  
The domestic substitution rate under Xinchuang is a hard performance indicator for government departments at all levels.  
To meet these targets and support local digital economy enterprises, local governments allow local small and medium-sized database vendors to participate in project bidding, forming a fragmented pattern of "one locality, one vendor."  
Industry entry barriers have been temporarily lowered, allowing a large number of small and medium-sized vendors to enter the market.

**Second, Project Split Bidding Mechanism**  
Digitalization projects for government affairs and state-owned enterprises generally split bidding for ERP systems, databases, middleware, and operation and maintenance services, rather than adopting integrated solutions.  
After splitting, the database becomes an independent sub-project with separate budget accounting. This provides survival space for small and medium-sized vendors and also prevents leading vendors from monopolizing the market, indirectly leading to a surge in the number of participants.

**Third, Open-Source Technology Lowering Entrepreneurship Barriers**  
The prevalence of open-source databases such as MySQL and PostgreSQL has significantly reduced the R&D threshold for database products, eliminating the need to build a kernel from scratch.  
Shallow modifications alone can produce a product, enabling teams without core technology to quickly package products and participate in market competition.

**Fourth, Policy Subsidies and Order Bonuses**  
Xinchuang-related industry subsidies and government-directed orders have become the main profit source for small and medium-sized vendors.  
Such vendors do not need to participate in market competition; they only need to meet basic compliance requirements to obtain project revenue, forming a business model that prioritizes orders over R&D.

From the perspective of industry development patterns, third-tier vendors generally adopt a "one-shot deal" business model. After project delivery, they only provide basic maintenance.  
Most lack subsequent technology upgrade capabilities. As project cycles end, a large number of these vendors will face elimination or deregistration. Industry consolidation is likely in the later stage.  
Products from small and medium-sized vendors based on the same open-source code will gradually be taken over by a few companies with maintenance capabilities for unified follow-up services.

**IV. Scenario-Specific Selection Rules: Differentiated Compliance Requirements for Government Enterprises and Private Enterprises**

In the domestic database market, two completely independent selection rules exist based on the type of user, with no unified standard. The core difference lies in compliance constraints versus market-oriented choices.

**Party and Government Organs and State-Owned Enterprises: Compliance First, Mandatory Domestic Substitution**  
Party and government organs, public institutions, central enterprises, and local state-owned enterprises are strictly bound by Xinchuang policies, the Cybersecurity Law, and the Data Security Law. Database selection must be included in the national Xinchuang product catalog, prioritizing independent and controllable domestic databases.  
Foreign commercial databases and non-compliant open-source databases are prohibited.  
Among them, grassroots units mainly need to meet basic compliance requirements and can choose second- or third-tier products.  
Core government affairs, military, and large central enterprises have higher requirements for data security and system stability, prioritizing first-tier full self-developed products.  
Such selection is not solely based on business performance needs; the core goal is to achieve technological self-control and convert fiscal funds into support for local industries.

**Private Enterprises: No Mandatory Constraints, Market-Oriented Autonomous Choice**  
Ordinary private enterprises have no mandatory requirements for domestic substitution or Xinchuang compliance.  
Database selection is entirely determined by their own business scale, cost budget, and technical needs.  
Small and micro enterprises and light-weight business scenarios can directly use free products such as SQLite, open-source MySQL, or open-source PostgreSQL.  
Medium-sized private enterprises can choose commercial databases or domestically modified databases according to their needs.  
Only when private enterprises undertake government projects or state-owned enterprise supply chain businesses are they required to use domestic databases from the Xinchuang catalog as per the client's requirements. In all other scenarios, there are no restrictions.

**V. Future Trends: Bubble Clearing, Head Concentration**

The current chaotic landscape of 167 vendors is a transitional phase in the development of the domestic database industry.  
In the long run, the industry will gradually move toward standardization and concentration.  
On one hand, shallow-modification vendors without R&D capabilities lack sustainable operation ability. As policy subsidies tighten and project acceptance standards rise, they will exit the market in batches, and the industry bubble will gradually clear.  
On the other hand, leading full self-developed vendors and high-quality open-source modification vendors will occupy the mainstream market through technology iteration, project integration, mergers and acquisitions, eventually forming a stable pattern of a few leading full self-developed vendors plus multiple specialized vendors in vertical fields.

From the perspective of core industrial value, the essence of Xinchuang localization is to break free from foreign technology monopoly, ensure national data security, and support the domestic digital technology industry.  
The current industry chaos is not the policy's original intention but a phased problem in early implementation.  
As industry standards improve and technology thresholds rise, domestic databases will gradually return to their technical essence, shifting from compliance orientation to capability orientation, achieving genuine independent control and market-oriented development.