https://n.cqs.es/en/tags/paper-notes/Recent content in Paper-Notes on Costa Life & Tech LabHugoenWed, 22 Jul 2026 08:55:00 +0000https://n.cqs.es/en/posts/2026/2026-2-fletch-switch-metadata-cache/Wed, 22 Jul 2026 08:55:00 +0000https://n.cqs.es/en/posts/2026/2026-2-fletch-switch-metadata-cache/<h2 class="group head-tag" id="-paper-walkthrough">📄 Paper Walkthrough<a data-title-of-head class="group-hover:after:content-['__#'] no-underline hover:text-[--hs-text-color-hover]" href="#-paper-walkthrough"></a></h2><p><strong>Fletch: File-System Metadata Caching in Programmable Switches</strong> Qingxiu Liu, Jiazhen Cai, Siyuan Sheng, Yuhui Chen, Lu Tang, Zhirong Shen, Patrick P. C. Lee arXiv:2510.08351v3 · May 2026 · 17 pages Code: <a href="https://github.com/adslabcuhk/fletch" target="_blank" rel="noopener noreferrer" >https://github.com/adslabcuhk/fletch</a > </p> <h2 class="group head-tag" id="tldr">TL;DR<a data-title-of-head class="group-hover:after:content-['__#'] no-underline hover:text-[--hs-text-color-hover]" href="#tldr"></a></h2><p>Fletch moves HDFS metadata caching from clients into the programmable switch&rsquo;s data plane, compiles it to a Tofino chip via P4, and serves metadata requests directly on the network&rsquo;s critical path. It drops cache-coherence complexity from O(N) to O(1) and uses in-switch centralization to absorb hot requests and fix load imbalance on the backend.</p>