<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Opentelemetry on Jorge's notebook</title><link>https://jorgesnotebook.com/tags/opentelemetry/</link><description>Recent content in Opentelemetry on Jorge's notebook</description><generator>Hugo</generator><language>en-GB</language><lastBuildDate>Sat, 18 Apr 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://jorgesnotebook.com/tags/opentelemetry/index.xml" rel="self" type="application/rss+xml"/><item><title>Designing an OpenTelemetry Pipeline for Scaling (Not Observability)</title><link>https://jorgesnotebook.com/posts/designing-otel-pipeline-for-scaling/</link><pubDate>Sat, 18 Apr 2026 00:00:00 +0000</pubDate><guid>https://jorgesnotebook.com/posts/designing-otel-pipeline-for-scaling/</guid><description>&lt;p&gt;We&amp;rsquo;re using Grafana Cloud for observability — Mimir, Tempo, Loki, the whole thing. It&amp;rsquo;s great, no complaints. Before all this we were only scaling on CPU and memory with the normal HPA. Which is fine until it isn&amp;rsquo;t — CPU and memory are lagging indicators, by the time they spike you&amp;rsquo;re already in trouble. So I started looking at KEDA because I wanted to scale on something that actually meant something, like queue depth. Nobody asked for it, I just thought it was the right call.&lt;/p&gt;</description></item></channel></rss>