<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Standard Edition on Rudi Bruchez</title><link>https://www.pachadata.com/en/tags/standard-edition/</link><description>Recent content in Standard Edition on Rudi Bruchez</description><generator>Hugo</generator><language>en</language><lastBuildDate>Sat, 21 Nov 2026 00:05:20 +0100</lastBuildDate><atom:link href="https://www.pachadata.com/en/tags/standard-edition/index.xml" rel="self" type="application/rss+xml"/><item><title>A partitioned columnstore with no downtime on Standard edition</title><link>https://www.pachadata.com/en/blog/2026/11/20/partitioned-columnstore-standard-edition/</link><pubDate>Fri, 20 Nov 2026 00:00:00 +0000</pubDate><guid>https://www.pachadata.com/en/blog/2026/11/20/partitioned-columnstore-standard-edition/</guid><description>&lt;p&gt;&lt;em&gt;Co-written with Claude, who wrote the migration scripts with me and replayed the method in the lab.&lt;/em&gt;&lt;/p&gt;&#10;&lt;div class="alert alert-success" role="alert"&gt;&lt;div class="h4 alert-heading" role="heading"&gt;TL;DR&lt;/div&gt;&#10;&lt;p&gt;On Standard edition, every structural operation on a table (partitioning it, rebuilding it, creating an index) blocks all writes to it for as long as it runs. To move a 385 million row monitoring table, written ten times per second, to a columnstore partitioned by year, I built the new table alongside: a trigger copied the new rows into a partition of their own, the history was copied in batches, and the cutover took 2.2 seconds. The table went from 50.4 GB to 9.3 GB.&lt;/p&gt;&#10;&lt;p&gt;But the columnstore alone was slower than the old table on four of the five dashboard queries, which all filter on a site, and more than twenty times slower on one of them. It took an extra B-tree index, built offline, and measuring before the cutover. With it, three of the five queries went from 0.4 to 0.7 s to under 0.1 s.&lt;/p&gt;&#10;&lt;/div&gt;</description></item></channel></rss>