<?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>Charts on Insignificant Bits</title>
    <link>https://bobcopeland.com/blog/tags/charts/</link>
    <description>Recent content in Charts on Insignificant Bits</description>
    <generator>Hugo</generator>
    <language>en-US</language>
    <lastBuildDate>Tue, 12 Feb 2013 14:12:34 +0000</lastBuildDate>
    <atom:link href="https://bobcopeland.com/blog/tags/charts/feed.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Cooking with gnuplot</title>
      <link>https://bobcopeland.com/blog/2013/02/cooking-with-gnuplot/</link>
      <pubDate>Tue, 12 Feb 2013 14:12:34 +0000</pubDate>
       <guid isPermaLink="false">http://bobcopeland.com/blog/?p=1922</guid> 
      <description>&lt;p&gt;Over the winter holidays I was put in charge of cooking one (of several) of the family dinners. At my house, a Christmas dinner can mean only one thing: prime rib is on the menu. The local grocery store had a great deal on rib roasts, so I bought a whole one. All seven ribs, 25 pounds of it. When it came time to cook this beast, I did plenty of reading, and settled on &lt;a href=&#34;http://www.seriouseats.com/2009/12/the-food-lab-how-to-cook-roast-a-perfect-prime-rib.html&#34;&gt;this seriouseats recipe&lt;/a&gt;. I guessed at about six hours to slow-roast the behemoth. But after a few hours of roasting, I decided it would be nice to know whether it would finish in time for the guests, or whether we would have to invent some pre-meal activities to stall.&lt;/p&gt;</description>
    </item>
    <item>
      <title>Packet Timing</title>
      <link>https://bobcopeland.com/blog/2012/05/packet-timing/</link>
      <pubDate>Mon, 21 May 2012 12:16:53 +0000</pubDate>
       <guid isPermaLink="false">http://bobcopeland.com/blog/?p=1827</guid> 
      <description>&lt;p&gt;Continuing with the charting theme, I wrote a quick python script yesterday to generate timing diagrams from pcap files while learning matplotlib. It&amp;rsquo;s just at the proof-of-concept stage but could be useful to debug things like power-saving in wireless. Surely something like this exists somewhere already, right? I couldn&amp;rsquo;t find it in the Google.&lt;/p&gt;
&lt;p&gt;&lt;img loading=&#34;lazy&#34; src=&#34;https://bobcopeland.com/blog/images/lj/timing.png&#34;&gt;&lt;/p&gt;
&lt;p&gt;Along the way I had to write a python radiotap parser; that might actually be useful to others so I&amp;rsquo;ll try to put it on github at some point.&lt;/p&gt;</description>
    </item>
    <item>
      <title>Graphing HBase Splits</title>
      <link>https://bobcopeland.com/blog/2012/04/graphing-hbase-splits/</link>
      <pubDate>Wed, 25 Apr 2012 12:58:01 +0000</pubDate>
       <guid isPermaLink="false">http://bobcopeland.com/blog/?p=1813</guid> 
      <description>&lt;p&gt;I was asked for a blog post on this topic. I also do birthdays!&lt;/p&gt;
&lt;p&gt;HBase achieves balance by splitting regions when they reach a certain size, and by evenly distributing the number of regions among cluster machines. However, the balancer will not run in some cases (e.g. if there are regions stuck in transition), and balancing the number of regions alone may not help if region sizes are not mostly the same size. If a region server is hosting more regions than the others, requests to that server experience higher latency, and batch (map-reduce) jobs take longer to complete due to parallel skew.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
