<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>Join-Order on Kent Yao</title>
    <link>https://yaooqinn.github.io/zh/tags/join-order/</link>
    <description>Recent content in Join-Order on Kent Yao</description>
    <generator>Hugo -- 0.157.0</generator>
    <language>zh-cn</language>
    <lastBuildDate>Mon, 25 May 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://yaooqinn.github.io/zh/tags/join-order/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>LLM 做 Join Order：从 Databricks 的实验,看 Apache Spark 视角下的三层落地阶梯</title>
      <link>https://yaooqinn.github.io/zh/posts/spark/llm-for-join-order-an-apache-spark-perspective/</link>
      <pubDate>Mon, 25 May 2026 00:00:00 +0000</pubDate>
      <guid>https://yaooqinn.github.io/zh/posts/spark/llm-for-join-order-an-apache-spark-perspective/</guid>
      <description>Databricks 与 UPenn 把 LLM agent 当成离线 join-order 调优师,在 JOB 113 条查询上拿到 P90 -41% / 几何均值 1.288× 的提速,甚至超过&amp;quot;完美基数估计&amp;quot;。从 Apache Spark 一线视角看,这件事说明了什么、又没说明什么。</description>
    </item>
  </channel>
</rss>
