<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Fast on cd ~</title><link>https://www.rottenbytes.info/tags/fast/</link><description>Recent content in Fast on cd ~</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Wed, 23 Jun 2010 12:19:34 +0000</lastBuildDate><atom:link href="https://www.rottenbytes.info/tags/fast/index.xml" rel="self" type="application/rss+xml"/><item><title>A new brick in my infrastructure</title><link>https://www.rottenbytes.info/2010/06/23/a-new-brick-in-my-infrastructure/</link><pubDate>Wed, 23 Jun 2010 12:19:34 +0000</pubDate><guid>https://www.rottenbytes.info/2010/06/23/a-new-brick-in-my-infrastructure/</guid><description>&lt;p&gt;&lt;a href="http://www.rottenbytes.info/wp-content/uploads/2009/02/eth0.jpg"&gt;&lt;img class="size-full wp-image-62 alignleft" title="eth0" src="http://www.rottenbytes.info/wp-content/uploads/2009/02/eth0.jpg" alt="eth0" width="80" height="103" /&gt;&lt;/a&gt;If you read this blog you may know I daily use puppet at $WORK. Puppet is made to maintain configuration on machines, but not for one shot actions. For a couple of month I used to work with fabric for this but It had a few drawbacks, mainly because you need to maintain the list of hosts you want to act on and that it hates dead hosts, even if &lt;a href="http://ghantoos.org/"&gt;ghantoos&lt;/a&gt; dropped a link showing how to &lt;a href="http://stackoverflow.com/questions/1956777/how-to-make-fabric-ignore-offline-hosts-in-the-env-hosts-list"&gt;get rid of this&lt;/a&gt;. So I replaced it with &lt;a href="http://code.google.com/p/mcollective/"&gt;mcollective&lt;/a&gt; : not exactly the same (it uses an agent instead of SSH) but it dynamicaly knows which hosts are up, allows to filter on facts (from facter, the puppet companion).&lt;/p&gt;</description></item><item><title>Enlarge your munin</title><link>https://www.rottenbytes.info/2009/02/25/enlarge-your-munin/</link><pubDate>Wed, 25 Feb 2009 16:10:53 +0000</pubDate><guid>https://www.rottenbytes.info/2009/02/25/enlarge-your-munin/</guid><description>&lt;p&gt;&lt;img class="alignleft size-full wp-image-62" title="eth0" src="http://www.rottenbytes.info/wp-content/uploads/2009/02/eth0.jpg" alt="eth0" width="80" height="103" /&gt;Au boulot j’ai remplacé &lt;a href="http://www.cacti.net"&gt;cacti&lt;/a&gt; par &lt;a href="http://munin.projects.linpro.no/"&gt;munin&lt;/a&gt;. Munin c’est choupi, la conf est en flat file mais ça a deux défauts, ou un selon le point de vue : le &lt;a href="http://munin.projects.linpro.no/wiki/CgiHowto"&gt;munin-cgi-graph&lt;/a&gt; est tout simplement moisi car d’une lenteur sans nom, et générer les images lors du poll c’est long. Sauf quand on tombe sur ce petit &lt;a href="http://blog.apokalyptik.com/2009/01/23/making-munin-graph-take-advantage-of-multiple-cpuscores/"&gt;hack&lt;/a&gt; convivial qui permet de lancer en parallèle la génération pour X nodes. Malheureusement, l’exemple donné en lien n’est valable que pour la version du trunk (au moment ou j’écris ceci) et pas pour la version 1.2.6 livrée avec lenny. J’ai donc modifié mon munin-graph en “portant” le hack disponible. Résultat : 2min30s au lieu de 9min pour la génération de 80 machines (dont certaines avec &lt;strong&gt;beaucoup&lt;/strong&gt; d’infos)&lt;/p&gt;</description></item></channel></rss>