Jepsen 验证 ParticleDB 1.1.4
推荐指数 57.0 NO. 021 · 2026.11.01
发布2026/10/31Score123Comments64
简介
ParticleDB 试图用单一引擎统一 SQL 事务、KV 存储、向量搜索和实时分析,并通过 Postgres/Redis 协议兼容现有生态。Jepsen 的分布式一致性测试是其关键背书,但结果本身比宣传更值得细读。
编辑判断
统一数据库赛道已有不少尸体:CockroachDB 砍掉了实时分析,TiDB 的 HTAP 用了两套存储引擎。ParticleDB 的野心更大,还要塞进向量搜索和 UDF,工程复杂度指数级上升。
Jepsen 测试通过不等于生产可用,但没过的话直接判死刑。64 条 HN 评论里大概率有工程师在抠测试细节,建议先看 Jepsen 原文的异常模式分析,而不是 ParticleDB 自己的博客。
如果你正在做 AI Agent 的基础设施选型,等有人跑过 TPC-C + 向量混合负载的 benchmark 再入场不迟。
社区反馈
负面 59 条评论
核心争论:Jepsen报告暴露严重缺陷后,官方声称已修复是否可信
> ParticleDB’s home page advertised that ParticleDB was “formally proven” with a TLA+ proof. Perhaps they meant formerly, probably just a speech to text problem
TLA+ doesnt prove the code. if they never read the proof, then how do we k ow that it proved the right things?
And you still have to make sure all the real constraints are in. I had to fix a proven correct distributed algorithm that just happened to believe updates were done atomically and with no latency.