<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Idempotency on IT Comparison</title><link>https://comparison.metacog.co.kr/tags/idempotency/</link><description>Recent content in Idempotency on IT Comparison</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Sun, 06 Sep 2026 09:59:45 +0900</lastBuildDate><atom:link href="https://comparison.metacog.co.kr/tags/idempotency/index.xml" rel="self" type="application/rss+xml"/><item><title>At-Least-Once vs Exactly-Once: Delivery Guarantees Compared</title><link>https://comparison.metacog.co.kr/posts/2026-09-06-at-least-once-vs-exactly-once-delivery-guarantees-compared/</link><pubDate>Sun, 06 Sep 2026 09:59:45 +0900</pubDate><guid>https://comparison.metacog.co.kr/posts/2026-09-06-at-least-once-vs-exactly-once-delivery-guarantees-compared/</guid><description>&lt;h2 id="overview"&gt;Overview&lt;/h2&gt;
&lt;p&gt;Messaging and stream-processing systems must pick a delivery guarantee: does a message arrive at least one time (possibly more), or does its effect happen precisely once no matter how many retries occur? &lt;strong class="kw"&gt;At-least-once&lt;/strong&gt; favors simplicity and throughput by retrying until acknowledged, at the cost of possible duplicates; &lt;strong class="kw"&gt;exactly-once&lt;/strong&gt; layers on deduplication or transactional coordination so retries never produce a second effect. The choice matters because a duplicate side effect — a double charge, a double email, a double stock decrement — can be catastrophic or merely annoying depending on the domain.&lt;/p&gt;</description></item></channel></rss>