<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
		>
<channel>
	<title>Comments on: Performance Testing: An OSE Case Study</title>
	<atom:link href="http://www.nickgerner.com/2010/01/performance-testing-case-study-ose/feed/" rel="self" type="application/rss+xml" />
	<link>http://www.nickgerner.com/2010/01/performance-testing-case-study-ose/</link>
	<description>Software Engineering and Entrepreneurship</description>
	<lastBuildDate>Thu, 14 Apr 2011 09:29:54 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.1</generator>
	<item>
		<title>By: Nick Gerner</title>
		<link>http://www.nickgerner.com/2010/01/performance-testing-case-study-ose/comment-page-1/#comment-38</link>
		<dc:creator>Nick Gerner</dc:creator>
		<pubDate>Thu, 18 Feb 2010 18:57:47 +0000</pubDate>
		<guid isPermaLink="false">http://www.nickgerner.com/?p=22#comment-38</guid>
		<description>All the measurements I&#039;m presenting here are about page views/second for the most expensive page we&#039;ve got (the first page of a report).  The expectation should be driven by marketing/product concerns: how many visitors are you expecting to get?  What is their behavior?  I discuss this a little bit in the &quot;Performance Targets&quot; section.

From the engineering side, this is probably driven by cost.  What&#039;s the cost of a page-view?  Hopefully you&#039;ve got a scalable architecture so you can say something like:

It&#039;ll cost us $300/month to push 20 pages/second, plus ops concerns.  If we want to push 40 pages/second, that&#039;ll cost us $600/month because we&#039;ll need to add another box.

I don&#039;t think I&#039;ve got a rule of thumb in general.  Performance varies from application to application so much that it&#039;s hard to say.  OSE is pretty data intensive: one report issues five or six API requests for a few thousand rows of data, hits the database a handful of times, and queues up other work for background processing.  On the other hand, there are plenty of &lt;a href=&quot;http://www.rubyenterpriseedition.com/comparisons.html&quot; rel=&quot;nofollow&quot;&gt;benchmarks&lt;/a&gt; that suggest a naive ruby app that does basically nothing can do ~500 page-views/second on low-end commodity hardware.  

I guess there you go: per box you should get at least 10 or 20 page-views/second at the low-end.  But it&#039;s unrealistic to expect 500 or more page-views/second from any real app on the kind of hardware you&#039;re likely to see in production.</description>
		<content:encoded><![CDATA[<p>All the measurements I&#8217;m presenting here are about page views/second for the most expensive page we&#8217;ve got (the first page of a report).  The expectation should be driven by marketing/product concerns: how many visitors are you expecting to get?  What is their behavior?  I discuss this a little bit in the &#8220;Performance Targets&#8221; section.</p>
<p>From the engineering side, this is probably driven by cost.  What&#8217;s the cost of a page-view?  Hopefully you&#8217;ve got a scalable architecture so you can say something like:</p>
<p>It&#8217;ll cost us $300/month to push 20 pages/second, plus ops concerns.  If we want to push 40 pages/second, that&#8217;ll cost us $600/month because we&#8217;ll need to add another box.</p>
<p>I don&#8217;t think I&#8217;ve got a rule of thumb in general.  Performance varies from application to application so much that it&#8217;s hard to say.  OSE is pretty data intensive: one report issues five or six API requests for a few thousand rows of data, hits the database a handful of times, and queues up other work for background processing.  On the other hand, there are plenty of <a href="http://www.rubyenterpriseedition.com/comparisons.html" rel="nofollow">benchmarks</a> that suggest a naive ruby app that does basically nothing can do ~500 page-views/second on low-end commodity hardware.  </p>
<p>I guess there you go: per box you should get at least 10 or 20 page-views/second at the low-end.  But it&#8217;s unrealistic to expect 500 or more page-views/second from any real app on the kind of hardware you&#8217;re likely to see in production.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Calin Brancus</title>
		<link>http://www.nickgerner.com/2010/01/performance-testing-case-study-ose/comment-page-1/#comment-37</link>
		<dc:creator>Calin Brancus</dc:creator>
		<pubDate>Thu, 18 Feb 2010 18:08:09 +0000</pubDate>
		<guid isPermaLink="false">http://www.nickgerner.com/?p=22#comment-37</guid>
		<description>Can you provide a relationship between the throughput (pages/second) and the pageviews/second - metric that is widely used for web performance.
Also what would be the throughput expectation - how many pages/second should a web site handle ?</description>
		<content:encoded><![CDATA[<p>Can you provide a relationship between the throughput (pages/second) and the pageviews/second &#8211; metric that is widely used for web performance.<br />
Also what would be the throughput expectation &#8211; how many pages/second should a web site handle ?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Tweets that mention Performance Testing: An OSE Case Study « Nick Gerner -- Topsy.com</title>
		<link>http://www.nickgerner.com/2010/01/performance-testing-case-study-ose/comment-page-1/#comment-10</link>
		<dc:creator>Tweets that mention Performance Testing: An OSE Case Study « Nick Gerner -- Topsy.com</dc:creator>
		<pubDate>Mon, 25 Jan 2010 21:16:08 +0000</pubDate>
		<guid isPermaLink="false">http://www.nickgerner.com/?p=22#comment-10</guid>
		<description>[...] This post was mentioned on Twitter by Rand Fishkin and Nick Gerner, jason nadaf. jason nadaf said: how to launch it, the tech side: RT @randfish RT @gerner just posted OSE perf testing case study: http://bit.ly/5fWLzX [...]</description>
		<content:encoded><![CDATA[<p>[...] This post was mentioned on Twitter by Rand Fishkin and Nick Gerner, jason nadaf. jason nadaf said: how to launch it, the tech side: RT @randfish RT @gerner just posted OSE perf testing case study: <a href="http://bit.ly/5fWLzX" rel="nofollow">http://bit.ly/5fWLzX</a> [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Chris Arkwright</title>
		<link>http://www.nickgerner.com/2010/01/performance-testing-case-study-ose/comment-page-1/#comment-9</link>
		<dc:creator>Chris Arkwright</dc:creator>
		<pubDate>Mon, 25 Jan 2010 07:01:49 +0000</pubDate>
		<guid isPermaLink="false">http://www.nickgerner.com/?p=22#comment-9</guid>
		<description>This is an awesome in-depth look at the complexity and quality of testing done to ensure a successful launch of OSE... which, by the way, is amazing. Nice write-up Nick. My compliments to the chef! :P

Cheers,
Chris</description>
		<content:encoded><![CDATA[<p>This is an awesome in-depth look at the complexity and quality of testing done to ensure a successful launch of OSE&#8230; which, by the way, is amazing. Nice write-up Nick. My compliments to the chef! <img src='http://www.nickgerner.com/wp-includes/images/smilies/icon_razz.gif' alt=':P' class='wp-smiley' /> </p>
<p>Cheers,<br />
Chris</p>
]]></content:encoded>
	</item>
</channel>
</rss>

<!-- Dynamic Page Served (once) in 0.221 seconds -->

