Segment 1: HiddenMerit Morning Post · Issue 108

Focus on Database Frontiers, Practical Insights for DBAs September 20, 2026 ｜ 5 Selected Global Breaking News

01｜Chengdu Research Base of Giant Panda Breeding Procures Kingbase Database: Three Sets of KingbaseES V9 at RMB 45,000 Each

On September 11, the winning bid announcement for the Chengdu Research Base of Giant Panda Breeding’s foundational software procurement was released, with Shenzhen Aerospace Smart City System Technology Research Institute Co., Ltd. winning the bid for RMB 233,800. The procurement list shows: 3 sets of Kingbase Database Management System KingbaseES V9.0 at RMB 45,000 each; and 38 sets of Tencent Cloud Linux Server Operating System V3 at RMB 2,600 each.

DBA Perspective: A research institution procuring domestic databases marks the extension of Xinchuang replacement from “traditional main battlefields” such as government and finance to broader public service sectors including research and culture. As a national‑level research institution, the panda base’s database selection carries demonstrative significance. For DBAs, this means employment scenarios for domestic databases are no longer limited to traditional industries — demand from emerging scenarios such as research institutions and cultural venues is being released.

02｜Larry Ellison Cancels $750 Million Oracle Stock Sale: Shares Undervalued, Ellison “Changes His Mind”

On September 11, a regulatory filing by Oracle revealed that co‑founder Larry Ellison planned to sell up to 50 million shares of Oracle stock before the end of October, worth approximately $7.5 billion based on Friday’s closing price of $150.

Just one day later, on September 12, Oracle announced: Ellison has decided not to proceed with the sale plan, and has no other plans to sell Oracle stock.

Background: Oracle’s share price has fallen by more than half since the $300 billion contract with OpenAI was announced in September 2025. Sources say Ellison believes the company’s stock is undervalued. The 82‑year‑old Ellison remains Oracle’s largest individual shareholder with a 40% stake, but has not appeared at recent earnings calls.

DBA Perspective: Ellison’s “change of mind” on the stock sale is an intriguing signal — when a founder believes their own stock is undervalued, it often means stronger confidence in the company’s long‑term value. For DBAs, Oracle’s share price volatility reflects market concerns about its “AI gamble,” but Ellison’s stance shows the founder’s long‑term judgement on the AI database strategy remains unshaken.

03｜Dameng: AI Capabilities Not Charged Separately, All‑in‑One Positioned as “Second Growth Curve”

On September 17, Dameng held an investor exchange meeting in Wuhan. The company clearly responded: AI capabilities are currently not charged as a standalone module, but integrated into the overall product capability to enhance comprehensive competitiveness and customer value.

Dual AI Tracks: The company is advancing both “AI for DB” (intelligent operations, SQL optimisation, automatic index recommendations, fault self‑healing) and “DB for AI” (native vector retrieval, multi‑modal data processing).

All‑in‑One Strategy: Dameng explicitly incorporates its database all‑in‑one business into its core strategy, positioning it as a “second growth curve.” H1 all‑in‑one revenue was RMB 27.12 million, up 519% year‑on‑year.

DBA Perspective: Dameng’s strategy of “not charging separately for AI capabilities” essentially treats AI as the product’s “comprehensive competitiveness” rather than a “profit centre.” For DBAs, this means that when choosing Dameng products, AI capabilities are “standard” rather than “optional.” The all‑in‑one positioning as a second growth curve also confirms the trend of domestic databases moving from “software replacement” to “hardware‑software synergy.”

04｜MySQL 8.0 EOL Follow‑Up: Community Governance Model Goes Live, AI and MCP Integration Added to Roadmap

On September 12, MySQL officially announced that the MySQL community governance model is now available. The governance framework clarifies the role boundaries of contributors, committers, project leads, and the Steering Committee, with committers and project leads open to non‑Oracle individuals, and the Steering Committee initially comprising Oracle, AWS, and Google Cloud.

AI Direction: Heather VanCura, Oracle’s VP of External Standards and Community Participation, emphasised that AI has been listed as one of MySQL’s top priorities, with the team discussing native vector support and having already included MCP integration in the product roadmap.

DBA Perspective: The transparency of MySQL community governance is an important sign of the project’s maturation. The roadmap for MCP integration and native vector support also indicates MySQL is extending into AI scenarios — DBAs with MySQL skills need to start understanding vector retrieval and agent data access protocols.

05｜PostgreSQL Anonymizer 3.2 Released: Faster Anonymisation, Plus Fix for Earlier Privilege Vulnerability

On September 9, Dalibo released PostgreSQL Anonymizer 3.2, introducing a new fast pseudonymisation feature panel.

Release Highlights:

New fast pseudonymisation capability, improving data masking efficiency
The earlier PostgreSQL Anonymizer 3.2.0 already fixed a privilege escalation vulnerability (CVE-2026-83534) that allowed table owners to run arbitrary code with superuser privileges

Related Ecosystem Developments: Also released were pg_vault_tde v1.7.1 (providing transparent data encryption for PG 17/18), the stable release of PostgreSQL Migrator 1.0, Autobase 2.11, and others.

DBA Perspective: The release of PostgreSQL Anonymizer 3.2 is a “feature‑enhanced version” following the fix of a privilege vulnerability. For DBAs, the simultaneous advancement of security fixes and feature enhancements reflects the health of open‑source extensions. pg_vault_tde’s transparent data encryption capability also provides DBAs with more options in compliance scenarios.

💡 SQL Little Knowledge Point

This Issue’s Knowledge Point: What is Transparent Data Encryption (TDE)?

TDE (Transparent Data Encryption) is a technology that encrypts database files at the storage layer. It is transparent to applications, requiring no modification of application code.

[quads id="805"]

Dimension	Traditional Encryption	TDE
Encryption Location	Application layer or column level	Storage layer (page/file)
Application Awareness	Requires code modification	Fully transparent
Key Management	Application‑side management	Database or external KMS
Typical Scenarios	Sensitive column encryption	Compliance requirements (classified protection, PCI‑DSS)

pg_vault_tde provides transparent data encryption for PostgreSQL 17/18, using AES‑256‑GCM to encrypt each tuple. For DBAs, this means meeting compliance requirements without modifying application code.

HiddenMerit Team Production Slogan: 绩优隐于内，金石启新程 | Hidden deep. Merit bold. Forge ahead.

Segment 2: HiddenMerit In‑Depth · Issue 21

Focus on Database Frontiers, In‑Depth Analysis September 20, 2026 ｜ 1 In‑Depth Feature Article

Ellison “Changes His Mind” on $7.5 Billion Stock Sale: Oracle Founder’s AI Gamble and the Observation Window for DBAs

1. A Decision to “Change His Mind”

On September 11, an Oracle regulatory filing revealed a major piece of news: co‑founder Larry Ellison planned to sell up to 50 million shares of Oracle stock, worth approximately $7.5 billion at the then‑current closing price. The trading plan was established at the end of June and was to run through the end of October.

Just one day later, Oracle announced: Ellison has decided not to proceed with the plan, and has no other plans to sell Oracle stock.

The company declined to explain why. But sources say Ellison believes the company’s stock is undervalued.

This is a rare “public change of mind” — an 82‑year‑old founder changing his mind within 24 hours of a regulatory filing disclosure.

2. Oracle’s “AI Gamble” and Share Price Pressure

To understand Ellison’s decision, one must first understand Oracle’s current situation.

Share Price Pressure: Oracle’s share price has fallen by more than half since the $300 billion contract with OpenAI was announced in September 2025. The market is cautious about Oracle’s “all‑in on AI data centres” strategy — investors worry about continuously compressed margins.

Financial Pressure: Oracle’s capital expenditure last quarter reached $28.5 billion, far exceeding revenue. The company is financing through debt issuance and stock offerings, and S&P downgraded its credit rating in July to just one notch above junk.

Business Progress: Oracle Cloud Infrastructure revenue was $7.39 billion, exceeding expectations. During the quarter, it signed $30 billion in new data centre capacity contracts, with a total pipeline of $664 billion.

Ellison’s “change of mind” occurred against this backdrop: the market is full of doubts about Oracle’s AI strategy, the share price is under pressure, and the founder chose to “hold, not sell.”

3. Ellison’s “All‑In on AI” Logic

Ellison’s strategic choices over the past two years can be summarised in four words: all‑in on AI.

Key Milestones:

Early 2025: Ellison, alongside OpenAI’s Sam Altman and SoftBank’s Masayoshi Son, announced the $500 billion Stargate plan at the White House
September 2025: Oracle signed a $300 billion data centre capacity contract with OpenAI, briefly making Ellison the world’s richest person, surpassing Musk
2026: Oracle pushes full‑speed ahead on data centre construction while conducting large‑scale layoffs (tens of thousands), concentrating resources on AI infrastructure

Ellison’s logic is clear: AI companies’ demand for compute power is exponential, and Oracle has the capability to provide that compute. The $300 billion OpenAI contract is the best proof — a “hyperscale” commitment sufficient to support Oracle’s growth for years to come.

4. The Observation Window for DBAs

Ellison’s “change of mind” on the stock sale is a signal worth watching for DBAs. It offers at least three observation windows:

Window 1: Oracle’s Strategic Resolve

The founder choosing to “hold, not sell” sends the signal that management is confident in the company’s long‑term value. For DBAs, this means Oracle’s AI database strategy will not pivot due to short‑term share price fluctuations. If you are evaluating Oracle as a long‑term technology stack, this is a positive signal.

Window 2: Oracle’s Product Priorities

When Oracle goes all‑in on AI data centres, the priority of traditional on‑premises databases may decline further. Ellison’s absence from earnings calls and co‑CEO Magouyrk leading the cloud business narrative both point to the same trend: Oracle’s future is in the cloud, in AI — not in traditional database licences.

For DBAs, this means: if you are still managing Oracle on‑premises instances, you need to start planning a migration path to the cloud or domestic databases.

Window 3: The “Bubble” Debate on AI Infrastructure

Ellison’s “change of mind” occurred at the height of the market debate over AI infrastructure ROI. Goldman Sachs analysts noted: “We have plenty of customers telling us that [Oracle] is too ‘spicy’ for them” — meaning Oracle’s AI promises are too aggressive, with risks too high.

For DBAs, the lesson from this debate is: the landing of AI databases takes time, and DBA skill development also needs “phased” planning — first mastering vector retrieval and agent data access, then progressively deepening into GPU databases and knowledge modelling.

5. The Bigger Picture

Ellison’s “change of mind” on the stock sale is essentially a story about “conviction.”

Against the backdrop of market doubts about Oracle’s AI strategy, a halved share price, and a credit rating downgrade, the 82‑year‑old founder chose “not to sell.” This doesn’t necessarily mean Oracle’s AI strategy will succeed, but it at least shows: Ellison himself believes that what Oracle is doing is worth persisting with.

For DBAs, the value of this signal is not “whether Oracle stock is worth buying,” but rather: as the database industry undergoes a paradigm shift from “traditional architecture” to “AI‑native,” choosing which technology path to follow and which database ecosystem to bet on requires independent judgement more than ever before.

Ellison chose to “hold Oracle.” DBAs also need to choose “what skills to hold.”