📊 HiddenMerit Daily · Issue 4

Focus on Database Frontiers, Practical Insights for DBAs April 30, 2026 ｜ 5 Selected Global Hot Topics

---

01｜Dameng DM9 Debuts at Digital China Summit: Domestic Databases Move from “Running Alongside” to “Leading”

The 9th Digital China Construction Summit was held grandly in Fuzhou from April 29 to 30. Dameng Data showcased its newly released DM9 database and full-stack product matrix at the on-site experience zone. Just before the summit, Dameng officially released its next-generation core product — Dameng Database Management System DM9 — at the 2026 China Database Technology and Industry Conference.

Five Core Breakthroughs:

Market Position: According to the April domestic database popularity ranking from modb.pro, Dameng’s score exceeded the 700 mark, forming a “dual brilliance” pattern with OceanBase. Dameng successfully completed support for the 2026 National Two Sessions information platform, and has gone live in core systems (HIS, LIS, PACS, and 40+ other business systems) of multiple tier-3 hospitals across China.

DBA Perspective This is a milestone event in the development journey of domestic databases. DM9 has several points worthy of deep attention from DBAs:

1. “Centralized & distributed integration” solves the selection dilemma: Previously, DBAs had to choose between “simple and easy-to-use centralized” and “flexibly scalable distributed” architectures. DM9 claims to offer both in a single system. If this capability truly matures, it will significantly reduce the decision cost of technology selection. It is recommended to pay attention to real-world production scaling drill cases.

2. Financial-grade availability with shared storage cluster: In real-world tests, query services recovered within 3 seconds and update services within 8 seconds from failure — this is approaching Oracle RAC levels. For DBAs in finance and government advancing domestic-tech replacement, this serves as a reference alternative solution.

3. Vector capabilities enter the core: DM9’s native vector search support marks that domestic databases are no longer lagging behind foreign vendors in AI integration. DBAs are advised to start learning the basic principles and tuning methods of vector indexing.

One sentence summary: The release of Dameng DM9 means that domestic databases now have the capability to compete head-to-head with mainstream international products in core transaction processing, AI integration, and other key areas.

---

02｜PostgreSQL 14 Approaching EOL: Only 5 Months Remaining Until End of Support

With PostgreSQL 19 expected to be released in September 2026, PostgreSQL 14 will officially reach End of Life (EOL) at that time. The EOL timeline is as follows:

Migration Recommendations:

• Direct upgrade to PG 17/18: Current LTS release line, longer support period

• Use pg_upgrade: Major version upgrade tool, be aware of statistics retention issues (resolved in PG 18)

DBA Perspective The wave of attention from MySQL 8.0 EOL has not yet dissipated, and the countdown to PG 14’s end of service has already begun. For production environments still running PG 14, now is the critical window to initiate upgrade assessment.

Action Recommendations:

1. Assess upgrade complexity: Use pg_upgrade --check to pre-check compatibility issues

2. Pay attention to PG 18’s statistics retention feature: More stable execution plans after upgrade, reducing the risk of “upgrade equals incident”

3. If still running PG 13: Needs immediate action — this version has been EOL for over half a year

One sentence summary: A “wave of database version death” is approaching — MySQL 8.0 EOL today, PG 14 EOL in five months. DBAs need to establish regular version management mechanisms.

---

03｜Oracle Launches “Diamond Tier” High Availability: Failure Recovery < 3 Seconds

Oracle recently announced major enhancements to its AI Database, introducing “Platinum Tier” and “Diamond Tier” high availability service levels, designed to provide unprecedented availability and security assurance for critical workloads.

Core Capability Comparison:

Key Technologies:

• Data Guard failover/switchover: Achieves rapid failover while supporting seamless software updates

• No application modification required: Only upgrade Oracle Database and Exadata software to gain new capabilities

• Post-quantum encryption enhancements: Addresses “harvest now, decrypt later” attacks

Oracle Official Statement: Juan Loaiza, Executive Vice President of Oracle AI Database Technology, stated: “Oracle databases power over 90% of the world’s largest enterprises and tens of thousands of small and medium-sized businesses.” Holger Mueller, VP and Principal Analyst at Constellation Research, commented: “This clearly raises the bar for what truly mission-critical data architecture should look like in the AI era.”

DBA Perspective Oracle is effectively “democratizing” enterprise-grade high availability capabilities to a broader customer base:

1. What does <3 second recovery mean? For financial transaction scenarios, this is a nearly zero-perceived interruption experience. However, achieving this level of recovery typically requires specific hardware and network configurations. DBAs need to assess whether their existing infrastructure meets the requirements.

2. Data Guard enhancements are worth attention: The seamless software update feature addresses a long-standing operational pain point — previously, patching meant downtime windows; now business can remain unaware during updates.

3. Significance for small and medium businesses: Oracle explicitly positions this capability for “tens of thousands of small enterprises,” indicating that high availability is no longer exclusively for large enterprises. However, cost still needs evaluation.

One sentence summary: Oracle has pushed the availability competition to a new “sub-second” level. Whether other vendors will follow suit is worth watching.

---

04｜Google Cloud Defines “Agentic Data Cloud”: Database Becomes the Context Hub for AI Agents

At the Google Cloud Next ’26 conference, Google Cloud clearly released its “Agentic Data Cloud” strategy, redefining enterprise databases as the core infrastructure for AI agents, not just storage repositories.

Strategic Core:

• Database role transformation: Over the past 50 years, the core of databases has been storing and returning “exact results.” In the AI era, databases need to produce “best results”

• Context is key: Sailesh Krishnamurthy, Global VP of Google Cloud Database Engineering, stated: “Models are excellent, but they don’t know all the context… That context exists in data, and the core of data ultimately resides in database systems”

• Multi-modal query integration: Graph search, vector embeddings, semantic search, full-text search, and relational operations need to work together within the same system

Technology Announcements:

• Spanner Omni: A downloadable version of the globally distributed Spanner database that can run in on-premises environments or even on competitors’ clouds

• Gemini Migration Agent: Uses generative AI to accelerate database migration, dramatically shortening months of manual work

Industry Context (from Oracle and Google partnership): Oracle AI Database Agent for Gemini Enterprise is now available on Google Cloud Marketplace, allowing users to query Oracle data using natural language without writing SQL. Microsoft and Amazon are also moving in similar directions, but Oracle embeds its Agent in an office collaboration platform rather than targeting only developers.

DBA Perspective This marks another major database vendor clearly articulating its “AI + database” strategic positioning after Oracle. Points worth noting:

1. The concept of “context hub” is worth DBAs pondering: If AI agents need to access data distributed across multiple systems, the DBA’s job will expand from “managing individual databases” to “managing the data context network.” Permission management, data lineage, and cross-source querying will become new skill requirements.

2. The significance of Spanner Omni: Google’s move to bring cloud database technology down to on-premises and other clouds reflects the complexity of enterprise data distribution. For DBAs in multi-cloud or hybrid cloud deployments, this could become a new technology option.

3. AI-powered migration agents: Database migration is one of the most time-consuming tasks for DBAs. If AI can automate schema conversion, SQL rewriting, and application adaptation, DBAs can focus more energy on architectural design.

One sentence summary: Competition in the AI-era database space has shifted from “model capability” to “accessibility of data context” — database vendors are vying to be the “memory foundation” for AI agents.

---

05｜Alibaba Cloud RDS MySQL 8.0 End of Service Countdown: Deadline July 31

With MySQL 8.0 Community Edition officially reaching End of Life (EOL) today (April 30, 2026), Alibaba Cloud’s end-of-service timeline for RDS for MySQL 8.0 is also clearly defined: Standard support ends on July 31, 2026.

Key Timeline:

Extended Support Cost Reference (based on AWS RDS standard, Alibaba Cloud similar):

• Year 1: Additional 100% of current instance cost

• Year 2: Additional 200% of current instance cost

• Year 3: Additional 300% of current instance cost

Upgrade Path Recommendations:

1. Recommended path: Upgrade to MySQL 8.4 LTS (relatively smooth migration path)

2. Alternative path: Evaluate migration to cloud-native databases such as PolarDB or TiDB

3. Deferral option: Pay extended support fees to buy more time for planning (high cost)

DBA Perspective MySQL 8.0’s EOL has gone from “news” to “today.” For DBAs still running MySQL 8.0, a decision must be made now.

Action Checklist:

1. Immediately perform compatibility assessment: Focus on client compatibility for caching_sha2_password vs mysql_native_password

2. Develop a phased upgrade plan: Prioritize upgrading non-core environments, verify no issues, then proceed to production

3. Calculate extended support costs: If unable to complete upgrades by July 31, communicate budget needs with finance in advance

4. Utilize the buffer window provided by RDS: Alibaba Cloud RDS standard support ends 3 months after community EOL — use this time effectively

Special Note: Technical institutions like Lawrence Livermore National Laboratory have already required all MySQL instances to migrate to MariaDB by April 30. This reminds us that EOL is not just a technical issue but may also involve compliance and security audit requirements.

One sentence summary: The clock has run out for MySQL 8.0. If your databases are still on 8.0, today is the last call to initiate upgrades.

---

📅 Today’s Database Events Calendar

Today is April 30, 2026. Events worth paying attention to:

---

Summary of This Issue

📌 Editor’s Note Today is a special day — MySQL 8.0 Community Edition officially reaches EOL. For DBAs, this is both a moment to “bid farewell to a classic” and the starting point for “forced upgrades.” At the same time, Dameng DM9’s appearance at the Digital China Summit and Google’s “Agentic Data Cloud” strategy announcement tell us that the database industry is accelerating into the “AI-native” era.

Three Action Recommendations:

1. Make version management routine: Don’t wait until EOL to think about upgrades; establish regular version assessment mechanisms

2. Proactively build domestic-tech knowledge: Even if not currently using domestic databases, start learning and understanding them

3. Bring AI skills forward: Vector search, natural language querying, AI Agent permission models — these are no longer “future technologies” but “present continuous”

Welcome to leave comments: How is your MySQL 8.0 upgrade plan progressing? Are you choosing 8.4 LTS or moving to cloud-native? See you tomorrow!