<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Durability on Pi Stack</title>
    <link>https://www.pistack.xyz/tags/durability/</link>
    <description>Recent content in Durability on Pi Stack</description>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Sun, 22 Nov 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://www.pistack.xyz/tags/durability/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>fsync() Is Not Durability: How Linux Storage Lies About Writes in 2026</title>
      <link>https://www.pistack.xyz/posts/2026-11-22-fsync-durability-lies-linux-storage-write-cache/</link>
      <pubDate>Sun, 22 Nov 2026 00:00:00 +0000</pubDate>
      <guid>https://www.pistack.xyz/posts/2026-11-22-fsync-durability-lies-linux-storage-write-cache/</guid>
      <description>&lt;p&gt;Your application called &lt;code&gt;fsync()&lt;/code&gt;. The kernel returned &lt;code&gt;0&lt;/code&gt;. Two seconds later the rack lost power, and the row your customer paid for was not in the database. No error was logged, no alert fired, and the backup was 40 minutes old. This is the single most common silent data-loss bug in production, and it is not a database bug — it is a misunderstanding of what &lt;code&gt;fsync()&lt;/code&gt; actually promises.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
