<?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/"
	xmlns:georss="http://www.georss.org/georss" xmlns:geo="http://www.w3.org/2003/01/geo/wgs84_pos#" xmlns:media="http://search.yahoo.com/mrss/"
		>
<channel>
	<title>Comments on: How EC2 changes the game in batch grid computing</title>
	<atom:link href="http://blog.rightscale.com/2008/04/05/how-ec2-changes-the-game-in-batch-grid-computing/feed/" rel="self" type="application/rss+xml" />
	<link>http://blog.rightscale.com/2008/04/05/how-ec2-changes-the-game-in-batch-grid-computing/</link>
	<description>Cloud Management News &#38; Conversations</description>
	<lastBuildDate>Thu, 02 Feb 2012 17:49:37 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.com/</generator>
	<item>
		<title>By: Stronghold Crusader</title>
		<link>http://blog.rightscale.com/2008/04/05/how-ec2-changes-the-game-in-batch-grid-computing/#comment-132</link>
		<dc:creator><![CDATA[Stronghold Crusader]]></dc:creator>
		<pubDate>Tue, 15 Apr 2008 16:57:12 +0000</pubDate>
		<guid isPermaLink="false">http://rightscale.wordpress.com/?p=50#comment-132</guid>
		<description><![CDATA[Good thread, i like these tips, its looks that i knew just small part.]]></description>
		<content:encoded><![CDATA[<p>Good thread, i like these tips, its looks that i knew just small part.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: TvE</title>
		<link>http://blog.rightscale.com/2008/04/05/how-ec2-changes-the-game-in-batch-grid-computing/#comment-80</link>
		<dc:creator><![CDATA[TvE]]></dc:creator>
		<pubDate>Tue, 08 Apr 2008 00:04:53 +0000</pubDate>
		<guid isPermaLink="false">http://rightscale.wordpress.com/?p=50#comment-80</guid>
		<description><![CDATA[Kevin, that&#039;s a good point. Right now what we do is keep instances alive until they&#039;re 55 minutes into the hour (to leave a few minutes to actually shut down). This works pretty well in a continuous usage scenario where a queue is getting work pretty much 24x7 and it&#039;s only a matter of adjusting the number of workers up and down with load throughout the day. But if you do one-shot batches where you fill the queue and then have it worked down to empty a lot more planning is needed to avoid having too many instances sitting idle for too long at the end. The key here is find the right trade-off between complexity (=fragility) and actual cost. Wasting an hour on 5-10% of the instances is probably not an issue, specially if the batch runs for several hours. Adding an hour on 100% of the instances may start to affect the bottom line.]]></description>
		<content:encoded><![CDATA[<p>Kevin, that&#8217;s a good point. Right now what we do is keep instances alive until they&#8217;re 55 minutes into the hour (to leave a few minutes to actually shut down). This works pretty well in a continuous usage scenario where a queue is getting work pretty much 24&#215;7 and it&#8217;s only a matter of adjusting the number of workers up and down with load throughout the day. But if you do one-shot batches where you fill the queue and then have it worked down to empty a lot more planning is needed to avoid having too many instances sitting idle for too long at the end. The key here is find the right trade-off between complexity (=fragility) and actual cost. Wasting an hour on 5-10% of the instances is probably not an issue, specially if the batch runs for several hours. Adding an hour on 100% of the instances may start to affect the bottom line.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Kevin</title>
		<link>http://blog.rightscale.com/2008/04/05/how-ec2-changes-the-game-in-batch-grid-computing/#comment-79</link>
		<dc:creator><![CDATA[Kevin]]></dc:creator>
		<pubDate>Mon, 07 Apr 2008 23:42:10 +0000</pubDate>
		<guid isPermaLink="false">http://rightscale.wordpress.com/?p=50#comment-79</guid>
		<description><![CDATA[For me, the issue is using as much as possible of the hour, since the whole hour has been bought on Amazon EC2. If you were to fire up lots of instances to finish the queue, and the average instance lifetime was 30 minutes then I&#039;m paying 100% more to Amazon than I need to. So, my point is can you do some clever scheduler stuff to make sure instances live for 59 minutes, or 1 hour 59 minutes, ad infinitum?]]></description>
		<content:encoded><![CDATA[<p>For me, the issue is using as much as possible of the hour, since the whole hour has been bought on Amazon EC2. If you were to fire up lots of instances to finish the queue, and the average instance lifetime was 30 minutes then I&#8217;m paying 100% more to Amazon than I need to. So, my point is can you do some clever scheduler stuff to make sure instances live for 59 minutes, or 1 hour 59 minutes, ad infinitum?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: items for 04.06.2008 &#171; Tzetze Fly</title>
		<link>http://blog.rightscale.com/2008/04/05/how-ec2-changes-the-game-in-batch-grid-computing/#comment-77</link>
		<dc:creator><![CDATA[items for 04.06.2008 &#171; Tzetze Fly]]></dc:creator>
		<pubDate>Mon, 07 Apr 2008 15:52:19 +0000</pubDate>
		<guid isPermaLink="false">http://rightscale.wordpress.com/?p=50#comment-77</guid>
		<description><![CDATA[[...] How EC2 changes the game in batch grid computing - RightScale BlogEnter Amazon EC2. If user A enqueues a job needing 500 nodes for 10 hours and user B a job needing 800 nodes for 5 hours what do you do? Very simple: you check the balance in their account and then start 500 instances for user A and 800 instances for user B. Done. No priorities, no scheduling, just pure compute fun! [...]]]></description>
		<content:encoded><![CDATA[<p>[...] How EC2 changes the game in batch grid computing &#8211; RightScale BlogEnter Amazon EC2. If user A enqueues a job needing 500 nodes for 10 hours and user B a job needing 800 nodes for 5 hours what do you do? Very simple: you check the balance in their account and then start 500 instances for user A and 800 instances for user B. Done. No priorities, no scheduling, just pure compute fun! [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Randy Nichols</title>
		<link>http://blog.rightscale.com/2008/04/05/how-ec2-changes-the-game-in-batch-grid-computing/#comment-65</link>
		<dc:creator><![CDATA[Randy Nichols]]></dc:creator>
		<pubDate>Sun, 06 Apr 2008 05:21:29 +0000</pubDate>
		<guid isPermaLink="false">http://rightscale.wordpress.com/?p=50#comment-65</guid>
		<description><![CDATA[Nice Blog.  I like the layout you used.  Did you make that yourself?

- Randy Nichols.]]></description>
		<content:encoded><![CDATA[<p>Nice Blog.  I like the layout you used.  Did you make that yourself?</p>
<p>- Randy Nichols.</p>
]]></content:encoded>
	</item>
</channel>
</rss>

