<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0"><channel><title><![CDATA[本地知识库更新，是重建全量还是增量？]]></title><description><![CDATA[<p dir="auto">本地知识库一周更新一次，全量重建要 4 小时。增量更新怎么做才不乱？</p>
]]></description><link>https://localaihub.com/topic/69/本地知识库更新-是重建全量还是增量</link><generator>RSS for Node</generator><lastBuildDate>Wed, 03 Jun 2026 19:23:07 GMT</lastBuildDate><atom:link href="https://localaihub.com/topic/69.rss" rel="self" type="application/rss+xml"/><pubDate>Sun, 03 May 2026 20:44:00 GMT</pubDate><ttl>60</ttl><item><title><![CDATA[Reply to 本地知识库更新，是重建全量还是增量？ on Mon, 04 May 2026 18:37:00 GMT]]></title><description><![CDATA[<p dir="auto">旧帖补后续：加 hash 后全量 4 小时变成增量 18 分钟，主要时间花在 PDF 解析。</p>
]]></description><link>https://localaihub.com/post/318</link><guid isPermaLink="true">https://localaihub.com/post/318</guid><dc:creator><![CDATA[小谢]]></dc:creator><pubDate>Mon, 04 May 2026 18:37:00 GMT</pubDate></item><item><title><![CDATA[Reply to 本地知识库更新，是重建全量还是增量？ on Mon, 04 May 2026 15:48:00 GMT]]></title><description><![CDATA[<p dir="auto">先补版本表。字段不用复杂：doc_id、hash、mtime、status、index_version、acl_version。</p>
]]></description><link>https://localaihub.com/post/317</link><guid isPermaLink="true">https://localaihub.com/post/317</guid><dc:creator><![CDATA[nora]]></dc:creator><pubDate>Mon, 04 May 2026 15:48:00 GMT</pubDate></item><item><title><![CDATA[Reply to 本地知识库更新，是重建全量还是增量？ on Mon, 04 May 2026 15:17:00 GMT]]></title><description><![CDATA[<p dir="auto">我们现在完全靠 rsync 后全量扫，确实没有版本表。</p>
]]></description><link>https://localaihub.com/post/316</link><guid isPermaLink="true">https://localaihub.com/post/316</guid><dc:creator><![CDATA[小谢]]></dc:creator><pubDate>Mon, 04 May 2026 15:17:00 GMT</pubDate></item><item><title><![CDATA[Reply to 本地知识库更新，是重建全量还是增量？ on Mon, 04 May 2026 14:46:00 GMT]]></title><description><![CDATA[<p dir="auto">删除也要有墓碑记录。否则后面同步断了，不知道是没扫到还是删了。</p>
]]></description><link>https://localaihub.com/post/315</link><guid isPermaLink="true">https://localaihub.com/post/315</guid><dc:creator><![CDATA[小潘同学]]></dc:creator><pubDate>Mon, 04 May 2026 14:46:00 GMT</pubDate></item><item><title><![CDATA[Reply to 本地知识库更新，是重建全量还是增量？ on Mon, 04 May 2026 14:24:00 GMT]]></title><description><![CDATA[<p dir="auto">用内容 hash + 稳定 doc_id。path 只是 metadata，不要当唯一身份。</p>
]]></description><link>https://localaihub.com/post/314</link><guid isPermaLink="true">https://localaihub.com/post/314</guid><dc:creator><![CDATA[阿白]]></dc:creator><pubDate>Mon, 04 May 2026 14:24:00 GMT</pubDate></item><item><title><![CDATA[Reply to 本地知识库更新，是重建全量还是增量？ on Mon, 04 May 2026 11:56:00 GMT]]></title><description><![CDATA[<p dir="auto">本地文件夹还要处理改名。path 变了但内容没变，不应该当新文档重复入库。</p>
]]></description><link>https://localaihub.com/post/313</link><guid isPermaLink="true">https://localaihub.com/post/313</guid><dc:creator><![CDATA[米饭]]></dc:creator><pubDate>Mon, 04 May 2026 11:56:00 GMT</pubDate></item><item><title><![CDATA[Reply to 本地知识库更新，是重建全量还是增量？ on Mon, 04 May 2026 11:44:00 GMT]]></title><description><![CDATA[<p dir="auto">看数据量。生产里 2 倍临时空间通常比半夜修脏数据便宜。</p>
]]></description><link>https://localaihub.com/post/312</link><guid isPermaLink="true">https://localaihub.com/post/312</guid><dc:creator><![CDATA[小路灯]]></dc:creator><pubDate>Mon, 04 May 2026 11:44:00 GMT</pubDate></item><item><title><![CDATA[Reply to 本地知识库更新，是重建全量还是增量？ on Mon, 04 May 2026 11:14:00 GMT]]></title><description><![CDATA[<p dir="auto">但向量库里建两份很占空间。</p>
]]></description><link>https://localaihub.com/post/311</link><guid isPermaLink="true">https://localaihub.com/post/311</guid><dc:creator><![CDATA[小周]]></dc:creator><pubDate>Mon, 04 May 2026 11:14:00 GMT</pubDate></item><item><title><![CDATA[Reply to 本地知识库更新，是重建全量还是增量？ on Mon, 04 May 2026 08:15:00 GMT]]></title><description><![CDATA[<p dir="auto">推荐蓝绿索引。新版本建好后切指针，失败就回滚旧版本。</p>
]]></description><link>https://localaihub.com/post/310</link><guid isPermaLink="true">https://localaihub.com/post/310</guid><dc:creator><![CDATA[阿航]]></dc:creator><pubDate>Mon, 04 May 2026 08:15:00 GMT</pubDate></item><item><title><![CDATA[Reply to 本地知识库更新，是重建全量还是增量？ on Mon, 04 May 2026 06:02:00 GMT]]></title><description><![CDATA[<p dir="auto">可以用 ingestion pipeline 记录转换步骤和缓存，但还是要有自己的版本表。</p>
]]></description><link>https://localaihub.com/post/309</link><guid isPermaLink="true">https://localaihub.com/post/309</guid><dc:creator><![CDATA[MingK]]></dc:creator><pubDate>Mon, 04 May 2026 06:02:00 GMT</pubDate></item><item><title><![CDATA[Reply to 本地知识库更新，是重建全量还是增量？ on Mon, 04 May 2026 03:41:00 GMT]]></title><description><![CDATA[<p dir="auto">这是最常见脏库。增量更新不是只加，是加、改、删、失效都要管。</p>
]]></description><link>https://localaihub.com/post/308</link><guid isPermaLink="true">https://localaihub.com/post/308</guid><dc:creator><![CDATA[林小北]]></dc:creator><pubDate>Mon, 04 May 2026 03:41:00 GMT</pubDate></item><item><title><![CDATA[Reply to 本地知识库更新，是重建全量还是增量？ on Mon, 04 May 2026 01:41:00 GMT]]></title><description><![CDATA[<p dir="auto">我们之前只追加新 chunk，不删旧 chunk。结果用户问制度，新旧版本一起回来。</p>
]]></description><link>https://localaihub.com/post/307</link><guid isPermaLink="true">https://localaihub.com/post/307</guid><dc:creator><![CDATA[小陈在改bug]]></dc:creator><pubDate>Mon, 04 May 2026 01:41:00 GMT</pubDate></item><item><title><![CDATA[Reply to 本地知识库更新，是重建全量还是增量？ on Sun, 03 May 2026 22:50:00 GMT]]></title><description><![CDATA[<p dir="auto">还要区分 doc_id 和 chunk_id。文档变了，旧 chunk 要能删干净。</p>
]]></description><link>https://localaihub.com/post/306</link><guid isPermaLink="true">https://localaihub.com/post/306</guid><dc:creator><![CDATA[nora]]></dc:creator><pubDate>Sun, 03 May 2026 22:50:00 GMT</pubDate></item><item><title><![CDATA[Reply to 本地知识库更新，是重建全量还是增量？ on Sun, 03 May 2026 22:06:00 GMT]]></title><description><![CDATA[<p dir="auto">先给文档算 hash。文件没变就别重嵌入，变了再处理。</p>
]]></description><link>https://localaihub.com/post/305</link><guid isPermaLink="true">https://localaihub.com/post/305</guid><dc:creator><![CDATA[rootless]]></dc:creator><pubDate>Sun, 03 May 2026 22:06:00 GMT</pubDate></item></channel></rss>