<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Distributed Systems on Yiwei Shi</title><link>/tags/distributed-systems/</link><description>Recent content in Distributed Systems on Yiwei Shi</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Mon, 11 May 2026 22:52:17 -0400</lastBuildDate><atom:link href="/tags/distributed-systems/index.xml" rel="self" type="application/rss+xml"/><item><title>Modern Real-Time OLAP Systems: ClickHouse is Winning</title><link>/posts/20260511-readtime-olap/</link><pubDate>Mon, 11 May 2026 22:52:17 -0400</pubDate><guid>/posts/20260511-readtime-olap/</guid><description>&lt;p&gt;A comparison of ClickHouse, StarRocks, Apache Druid, and Apache Pinot — the four engines defining real-time analytics in 2026.&lt;/p&gt;
&lt;h2 id="why-real-time-olap-matters-now"&gt;Why Real-Time OLAP Matters Now&lt;/h2&gt;
&lt;p&gt;The old split was clean: OLTP for operations, batch warehouses for BI. Data freshness measured in hours was fine, because the workloads were retrospective.&lt;/p&gt;
&lt;p&gt;That model has broken. User-facing analytics, fraud detection, ad bidding, and RAG-style AI applications all need sub-second queries over billions of rows, with data that&amp;rsquo;s seconds old. Real-time OLAP databases close the gap between streaming ingestion (Kafka, Flink) and interactive analytical queries.&lt;/p&gt;</description></item></channel></rss>