#DistributedSQL

Relevant results can still be wrong if they belong to a different project version. Search must filter by access & version constraints BEFORE scoring relevance. Governance precedes retrieval. #DistributedSQL #Security #OceanBase

Distinguish evidence from guesses. Agents digging deeper need to track which findings are verified. Strong enough evidence supports change. Weak signals require more gathering. Clarity prevents cascading errors. #SRE #DistributedSQL #OceanBase

Don't model everything. Start with high-value use cases & expand via feedback. Complete graphs upfront stall progress. Concrete scenarios drive adoption. #DistributedSQL #Tech #OceanBase

Stateless APIs work for inference. Stateful storage works for engineering workflows. The gap is durable project context. Bridging this requires ACID guarantees on agent memory writes. #DistributedSQL #Backend #OceanBase

Eliminate the context loading latency spike at session start. Snapshot restores in milliseconds. Warm starts beat full context windows for interactive speed. #DistributedSQL #LLM #OceanBase

Independent verification prevents agent drift. A Verifier checks delivery without executing code. Leaders manage goals, Workers execute. Separating validation from execution ensures correctness. #AgenticAI #DistributedSQL #OceanBase

Error codes need keyword search. UI bugs need semantic search. Both need project/version filters. Separate queries fragment context. Hybrid search combines full-text, vector, & structured filters. How it works: #DistributedSQL #AI #OceanBase

Multi-agent systems need context routing. Leaders manage goals, Workers execute, Verifiers check. Don't dump the full log on every node. #DistributedSQL #OceanBase

Business entities rarely match single tables. A "customer" spans profiles, orders, & logs. Ontology assembles fragmented records into one logical object for agents. How models this: #DistributedSQL #AI #OceanBase

Context durability determines agent autonomy. Without it, you just have a fancy autocomplete tool for single sessions. #DistributedSQL #Tech #OceanBase

How works revised version 👀

HTAP usage in Learn #HTAP database with 3 SQLs 👇👇👇

Feel free to run a #distributedsql on a 4.8 billion-row of table with #HTAP #Database #TiDB, it's smart enough to know whether to use row index or vector calculation for the same query according to the amount of data: `SELECT COUNT(*)...` #100DaysOfCode

🌱A database is not built in a day. 🏗️ During the years of building #TiDB, we formed 5️⃣ principles that guide TiDB through a focused yet flexible evolution to an ideal #database. 👀 Hear the details from our CEO @ngaut1👉 #DistributedSQL

.@MarkusWinand (SQL ambassador @ModernSQL) goes deep on SQL’s history and developments in 21st century versions of the SQL spec in this paper: If you like to talk nerdy about #SQL (we do) then this is right up your alley. #distributedSQL #DevCommunity

Vitessの開発元PlanetScaleによる、コンセンサスアルゴリズムの解説blog。Distributed durabilityやAvailabilityは良く言われる観点だが、Automationも考えた場合にRaft/Paxosベースのシステムが本当にベストか？という疑問から出発している。 #distributedsql

やはりこれを解説しておこう。CockroachDBが原子時計なしで分散SQLを如何に構築するかについて。同期化された時刻は分散システムの聖杯で、Spanner以前のPercolatorでは単一の時間ソースと通信が性能を制約し、可用性の向上も難しかった。 #distributedsql

An "Awesome" project for digging deeper into the database, see #Database #DistributedSQL #tidb #tikv

Most semantic layers sit outside the database. Tight integration reduces latency & keeps context consistent with transactional data. Co-location is key for real-time agent reliability. #DistributedSQL #SystemDesign #OceanBase

🎉 TiDB is officially 10 years old! 🎉 This milestone wouldn’t be possible without our incredible team members, dedicated users, and vibrant open-source community. Thank you for being part of our journey. 💙 #PingCAP #TiDB #DistributedSQL #OpenSource

#Database performance hit a wall with the success of the internet in combination with explosive application development. 💥 At the heart of the issue is elasticity — and the solution is #DistributedSQL! @probablyrealrob explains why in @thenewstack:

🚨Download Distributed SQL Databases for Dummies - a @Yugabyte Special Edition for FREE!🚨 Discover why #distributedsql is the ideal scalable, highly available, secure, & flexible #cloudnative solution for a wide-range of use cases! ⬇️