<?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: Can Enterprise IT Adapt? Sure, But Why Start Now?</title>
	<atom:link href="http://storagemojo.com/2006/08/04/can-enterprise-it-adapt-sure-but-why-start-now/feed/" rel="self" type="application/rss+xml" />
	<link>http://storagemojo.com/2006/08/04/can-enterprise-it-adapt-sure-but-why-start-now/</link>
	<description>Data storage info &#38; analysis</description>
	<lastBuildDate>Tue, 15 May 2012 23:44:19 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.3.2</generator>
	<item>
		<title>By: Robert Pearson</title>
		<link>http://storagemojo.com/2006/08/04/can-enterprise-it-adapt-sure-but-why-start-now/comment-page-1/#comment-4953</link>
		<dc:creator>Robert Pearson</dc:creator>
		<pubDate>Mon, 07 Aug 2006 07:46:28 +0000</pubDate>
		<guid isPermaLink="false">http://storagemojo.com/?p=210#comment-4953</guid>
		<description>To me, the main thing your IDC series highlighted is that none of the known IDCs used &quot;COTS&quot; Storage. 
From the reading I have done it appears that is because COTS Storage does not scale well. In this case both vertically and horizontally. And certainly not cost effectively. More like cost prohibitively. 

My interests have moved from the IDC to scalable, cost effective Storage. Like you talk about in your &quot; White Box Arrays Are Here. w00t!&quot; at http://storagemojo.com/?p=160 
As I read around the Storage Blogs I find more chatter about the lack of specs for RAID-6 and work-a-rounds for this. 
If these are the perceived problems I have some solutions. 

Let&#039;s talk about really cheap disks. What can we do with really cheap disks. For one thing we can &quot;Roll our own Storage&quot;. If disks are really cheap and capacities are doubling rapidly why do I need RAID? Let&#039;s just use JBOD and find some fast interconnect controllers and use duplicate or triplicate JBOD. The I/O &quot;stripe size&quot; then becomes the JBOD size. All I need is an Operating System or just a File System like ZFS that is Storage aware and Storage centric. GFS (Global File System) would do it but ZFS is newer technology. 

I proposed to the NetApps guys one time that the Storage cache could provide Content Information automatically. They looked like they were having heart attacks. They did not even want to discuss it! I don&#039;t care if it is in cache or not. It should be in the Storage box. The box should be able to provide this Information. 

Then there was Gene Amdahl&#039;s Storage box design. The Sun Fire X4500 has a rough similarity to Gene&#039;s box. 

A real interesting development is the VMware &quot;Software Lifecycle Management&quot;. 
http://www.vmware.com/news/releases/software_lifecycle.html
Once you have this &quot;All things are possible&quot;. 

Will the &quot;Tiger Woods&quot; of Storage ever appear? Will he raise the bar so that we can be the best we can be? 
I had hoped Jon Toigo would be the man. Jon seems to feel he has. I don&#039;t see it yet.</description>
		<content:encoded><![CDATA[<p>To me, the main thing your IDC series highlighted is that none of the known IDCs used &#8220;COTS&#8221; Storage.<br />
From the reading I have done it appears that is because COTS Storage does not scale well. In this case both vertically and horizontally. And certainly not cost effectively. More like cost prohibitively. </p>
<p>My interests have moved from the IDC to scalable, cost effective Storage. Like you talk about in your &#8221; White Box Arrays Are Here. w00t!&#8221; at <a href="http://storagemojo.com/?p=160" rel="nofollow">http://storagemojo.com/?p=160</a><br />
As I read around the Storage Blogs I find more chatter about the lack of specs for RAID-6 and work-a-rounds for this.<br />
If these are the perceived problems I have some solutions. </p>
<p>Let&#8217;s talk about really cheap disks. What can we do with really cheap disks. For one thing we can &#8220;Roll our own Storage&#8221;. If disks are really cheap and capacities are doubling rapidly why do I need RAID? Let&#8217;s just use JBOD and find some fast interconnect controllers and use duplicate or triplicate JBOD. The I/O &#8220;stripe size&#8221; then becomes the JBOD size. All I need is an Operating System or just a File System like ZFS that is Storage aware and Storage centric. GFS (Global File System) would do it but ZFS is newer technology. </p>
<p>I proposed to the NetApps guys one time that the Storage cache could provide Content Information automatically. They looked like they were having heart attacks. They did not even want to discuss it! I don&#8217;t care if it is in cache or not. It should be in the Storage box. The box should be able to provide this Information. </p>
<p>Then there was Gene Amdahl&#8217;s Storage box design. The Sun Fire X4500 has a rough similarity to Gene&#8217;s box. </p>
<p>A real interesting development is the VMware &#8220;Software Lifecycle Management&#8221;.<br />
<a href="http://www.vmware.com/news/releases/software_lifecycle.html" rel="nofollow">http://www.vmware.com/news/releases/software_lifecycle.html</a><br />
Once you have this &#8220;All things are possible&#8221;. </p>
<p>Will the &#8220;Tiger Woods&#8221; of Storage ever appear? Will he raise the bar so that we can be the best we can be?<br />
I had hoped Jon Toigo would be the man. Jon seems to feel he has. I don&#8217;t see it yet.</p>
]]></content:encoded>
	</item>
</channel>
</rss>

