<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Tracing on Adam Vu</title><link>https://vutg.me/tags/tracing/</link><description>Recent content in Tracing on Adam Vu</description><generator>Hugo</generator><language>en</language><lastBuildDate>Wed, 19 Nov 2025 00:00:00 +0000</lastBuildDate><atom:link href="https://vutg.me/tags/tracing/index.xml" rel="self" type="application/rss+xml"/><item><title>Stop Logging, Start Observing: OpenTelemetry Metrics and Traces with Honeycomb</title><link>https://vutg.me/posts/stop-logging-start-observing-opentelemetry-metrics-and-traces-with-honeycomb/</link><pubDate>Wed, 19 Nov 2025 00:00:00 +0000</pubDate><guid>https://vutg.me/posts/stop-logging-start-observing-opentelemetry-metrics-and-traces-with-honeycomb/</guid><description>&lt;p&gt;Your service emits ten thousand log lines a minute. When it breaks at 2 AM, you grep through them, pattern-match by eye, and eventually find the one line that mattered. You&amp;rsquo;ve just done manually what a machine should have done automatically. This is the fundamental problem with logging as a primary observability strategy.&lt;/p&gt;
&lt;p&gt;This post argues that &lt;strong&gt;logs are the wrong primitive&lt;/strong&gt; for understanding system behavior at scale, explains the OpenTelemetry model that replaces them, and walks through a hands-on Python lab that exports telemetry to Honeycomb so you can see the difference yourself.&lt;/p&gt;</description></item></channel></rss>