All

Cost Reduction of RMB 30 Million! YMatrix Helps Customer Save 100 Servers

With “less storage, less reading, less data movement, less computation, and fewer systems” as its core principles, YMatrix builds a resource efficiency framework designed for enterprise-grade analytical scenarios.

#Technical Discussion

2026-06-18

Why Can WAL Data Become “Corrupted Out of Thin Air”?——An Incident Investigation Spanning the Boundary Between Databases and Hardware

A rare WAL CRC32C anomaly investigation traced data corruption from the database layer down to NIC DMA and PCIe hardware transmission paths, exposing reliability challenges in modern data infrastructure.

#Technical Discussion

2026-05-15

AI Era Database Infrastructure: Exploring Vectorized Execution in PostgreSQL

YMatrix mxvector builds an analytical execution path more suitable for OLAP/HTAP scenarios without rewriting the PostgreSQL executor.

#Technical Discussion

2026-05-07

The PostgreSQL Streaming Model

PostgreSQL is Already a Streaming System — Domino Make It Explicit

#Technical Discussion

2026-04-15

In the AI Era, Parallel Query Is Evolving from “Faster Scanning” to “Faster Computing”

How MARS3 Bucket Enables Parallel Gains to Extend into Computation

#Technical Discussion

2026-04-09

In-depth Analysis (Part 3) — In the AI Era, Databases Are Entering the “Unified Storage Era”

Introduction to the MARS3 Storage Engine

#Technical Discussion

2026-04-02

In-depth Analysis (Part 2) — In the AI Era, Databases Are Entering the “Unified Storage Era”

Introduction to the MARS3 Storage Engine

#Technical Discussion

2026-04-02

In-depth Analysis (Part 1) — In the AI Era, Databases Are Entering the “Unified Storage Era”

Introduction to the MARS3 Storage Engine

#Technical Discussion

2026-04-02

Previous
All
Cost Reduction of RMB 30 Million! YMatrix Helps Customer Save 100 Servers
With “less storage, less reading, less data movement, less computation, and fewer systems” as its core principles, YMatrix builds a resource efficiency framework designed for enterprise-grade analytical scenarios.
#Technical Discussion
2026-06-18
Why Can WAL Data Become “Corrupted Out of Thin Air”?——An Incident Investigation Spanning the Boundary Between Databases and Hardware
A rare WAL CRC32C anomaly investigation traced data corruption from the database layer down to NIC DMA and PCIe hardware transmission paths, exposing reliability challenges in modern data infrastructure.
#Technical Discussion
2026-05-15
AI Era Database Infrastructure: Exploring Vectorized Execution in PostgreSQL
YMatrix mxvector builds an analytical execution path more suitable for OLAP/HTAP scenarios without rewriting the PostgreSQL executor.
#Technical Discussion
2026-05-07
The PostgreSQL Streaming Model
PostgreSQL is Already a Streaming System — Domino Make It Explicit
#Technical Discussion
2026-04-15
In the AI Era, Parallel Query Is Evolving from “Faster Scanning” to “Faster Computing”
How MARS3 Bucket Enables Parallel Gains to Extend into Computation
#Technical Discussion
2026-04-09
In-depth Analysis (Part 3) — In the AI Era, Databases Are Entering the “Unified Storage Era”
Introduction to the MARS3 Storage Engine
#Technical Discussion
2026-04-02
In-depth Analysis (Part 2) — In the AI Era, Databases Are Entering the “Unified Storage Era”
Introduction to the MARS3 Storage Engine
#Technical Discussion
2026-04-02
In-depth Analysis (Part 1) — In the AI Era, Databases Are Entering the “Unified Storage Era”
Introduction to the MARS3 Storage Engine
#Technical Discussion
2026-04-02
Previous