<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>The Unified Logging Infrastructure for Data Analytics at Twitter · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/1208.4171/</link><description>Collected HN comments and submissions linking this paper.</description><language>en</language><lastBuildDate>Mon, 02 Nov 2026 11:16:22 GMT</lastBuildDate><atom:link href="https://hackerpaper.kaipfister.de/p/1208.4171/feed.xml" rel="self" type="application/rss+xml" /><item><title>The Unified Logging Infrastructure for Data Analytics at Twitter — mentioned by cdavid</title><link>https://news.ycombinator.com/item?id=31547616</link><guid isPermaLink="false">urn:hackerpapers:mention:1208.4171:31547616</guid><pubDate>Sun, 29 May 2022 09:02:08 GMT</pubDate><description>&lt;p&gt;I agree certain orders of magnitude are harder to guess accurately. My claim for 1k qps being conservative is mostly based on * my current company, but I can&amp;#x27;t share more precise numbers. So I guess not very convincing :) * however, at ~ 300 millions listings in 2021 for airbnb, that means i.e. ~10 listing per second. 1k qps implies ~ 100 queries per listing, which would be an extremely good ratio. Every time you change a parameter (price range, map range, type of place), that&amp;#x27;s a new query. &amp;gt; Sure, I think 30 bytes is reasonable for a click. When I said 15 I was squeezing a bit much. But…&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/1208.4171/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.DB</category></item></channel></rss>
