<?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: autobench: Cannot find httperf in $PATH</title>
	<atom:link href="http://www.question-defense.com/2008/08/11/autobench-cannot-find-httperf-in-path/feed" rel="self" type="application/rss+xml" />
	<link>http://www.question-defense.com/2008/08/11/autobench-cannot-find-httperf-in-path</link>
	<description>Technical answers for technical questions</description>
	<lastBuildDate>Mon, 20 May 2013 08:02:09 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
		<item>
		<title>By: alex</title>
		<link>http://www.question-defense.com/2008/08/11/autobench-cannot-find-httperf-in-path#comment-17301</link>
		<dc:creator>alex</dc:creator>
		<pubDate>Mon, 09 Apr 2012 02:26:11 +0000</pubDate>
		<guid isPermaLink="false">http://question-defense.com/?p=5#comment-17301</guid>
		<description><![CDATA[Hello lyapma,

The autobench Perl script is not necessarily something that you can control your CPU with only in the fact that it will make LOAD rise during your tests. So as far as having the CPU use 50% for three minutes and/or then having it fall to 0% is not really its purpose. The script is made to stress test a web server to see what kind of load the server can handle and while this would include monitoring CPU usage/LOAD to see how your server does under stress it is not to have the CPU perform at specific levels for specific intervals. 

Thanks.
alex]]></description>
		<content:encoded><![CDATA[<p>Hello lyapma,</p>
<p>The autobench Perl script is not necessarily something that you can control your CPU with only in the fact that it will make LOAD rise during your tests. So as far as having the CPU use 50% for three minutes and/or then having it fall to 0% is not really its purpose. The script is made to stress test a web server to see what kind of load the server can handle and while this would include monitoring CPU usage/LOAD to see how your server does under stress it is not to have the CPU perform at specific levels for specific intervals. </p>
<p>Thanks.<br />
alex</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: lyapma</title>
		<link>http://www.question-defense.com/2008/08/11/autobench-cannot-find-httperf-in-path#comment-17020</link>
		<dc:creator>lyapma</dc:creator>
		<pubDate>Fri, 09 Mar 2012 02:58:03 +0000</pubDate>
		<guid isPermaLink="false">http://question-defense.com/?p=5#comment-17020</guid>
		<description><![CDATA[Hi
I&#039;m out of the topic, but i really need some helps....:(

I&#039;m currently using autobench to test a personal web server. I want to define a scenario with 3 main steps:

Step 1: generate an increasing request load. For example from low_rate=10 to high_rate=100 with a rate=10 ==&gt; Everything is OK here

Step 2: I want to be able to have a &quot;fix&quot; CPU load. For example if at the end of Step 1 my CPU load is around 50 I want my server to stay at a load of 50 for almost 3mn before decreasing ==&gt; is it possible to do it with autobench? If yes, how can I do it? Else, can you propose some useful tools to active this step? Step 3: I want to be able do

Step 3: I want to be able to generate load from the current CPU load to 0. For example, configure autobench like this: low_rate=100 high_rate=10 and rate=-10 (negative value) ==&gt; It is not working? Is it possible to do it with autobench? If yes how? Else, and you give me some advises to perform my experience? Sorry for my English!]]></description>
		<content:encoded><![CDATA[<p>Hi<br />
I&#8217;m out of the topic, but i really need some helps&#8230;.:(</p>
<p>I&#8217;m currently using autobench to test a personal web server. I want to define a scenario with 3 main steps:</p>
<p>Step 1: generate an increasing request load. For example from low_rate=10 to high_rate=100 with a rate=10 ==&gt; Everything is OK here</p>
<p>Step 2: I want to be able to have a &#8220;fix&#8221; CPU load. For example if at the end of Step 1 my CPU load is around 50 I want my server to stay at a load of 50 for almost 3mn before decreasing ==&gt; is it possible to do it with autobench? If yes, how can I do it? Else, can you propose some useful tools to active this step? Step 3: I want to be able do</p>
<p>Step 3: I want to be able to generate load from the current CPU load to 0. For example, configure autobench like this: low_rate=100 high_rate=10 and rate=-10 (negative value) ==&gt; It is not working? Is it possible to do it with autobench? If yes how? Else, and you give me some advises to perform my experience? Sorry for my English!</p>
]]></content:encoded>
	</item>
</channel>
</rss>
