<?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: The Blade is Dead! Long Live the Rack!</title>
	<atom:link href="http://blog.nigelpoulton.com/the-blade-is-dead-long-live-the-rack/feed/" rel="self" type="application/rss+xml" />
	<link>http://blog.nigelpoulton.com/the-blade-is-dead-long-live-the-rack/#utm_source=feed&amp;utm_medium=feed&amp;utm_campaign=feed</link>
	<description>with nigel poulton</description>
	<lastBuildDate>Tue, 24 Aug 2010 18:10:47 +0000</lastBuildDate>
	<generator>http://wordpress.org/?v=2.8.6</generator>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
		<item>
		<title>By: Nigel Poulton</title>
		<link>http://blog.nigelpoulton.com/the-blade-is-dead-long-live-the-rack/comment-page-1/#comment-701</link>
		<dc:creator>Nigel Poulton</dc:creator>
		<pubDate>Wed, 20 Jan 2010 22:44:20 +0000</pubDate>
		<guid isPermaLink="false">http://blog.nigelpoulton.com/the-blade-is-dead-long-live-the-rack/#comment-701</guid>
		<description>Cam,

I was thinking about what you said about management in your comment above and had an aha! moment.  I misunderstood what you were getting at.

Yes I think an IOV solution makes managing the RAN and your I/O environment far simpler.  It offers a flexibility that will be a great addition to virtual server environments.  I/O flexibility has been lagging and IOV solutions like those offered by Xsigo will bring it up to speed.

Unified fabric technologies like FCoE (DCB) can still play a part though... but more than likely outside of the RAN and in the network core....

Nigel</description>
		<content:encoded><![CDATA[<p>Cam,</p>
<p>I was thinking about what you said about management in your comment above and had an aha! moment.  I misunderstood what you were getting at.</p>
<p>Yes I think an IOV solution makes managing the RAN and your I/O environment far simpler.  It offers a flexibility that will be a great addition to virtual server environments.  I/O flexibility has been lagging and IOV solutions like those offered by Xsigo will bring it up to speed.</p>
<p>Unified fabric technologies like FCoE (DCB) can still play a part though&#8230; but more than likely outside of the RAN and in the network core&#8230;.</p>
<p>Nigel</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: John Obeto</title>
		<link>http://blog.nigelpoulton.com/the-blade-is-dead-long-live-the-rack/comment-page-1/#comment-693</link>
		<dc:creator>John Obeto</dc:creator>
		<pubDate>Mon, 18 Jan 2010 23:14:30 +0000</pubDate>
		<guid isPermaLink="false">http://blog.nigelpoulton.com/the-blade-is-dead-long-live-the-rack/#comment-693</guid>
		<description>Hello Nigel,

As you rightfully point out, the FCoE RAN solution is basically an evolutionary step from where we are today.
In a fully-converged architecture (thanks for the moniker, HP), a revolutionary improvement from today&#8217;s infrastructure &lt;b&gt;&lt;i&gt;must &lt;/i&gt;&lt;/b&gt;&lt;span&gt;&#160;&lt;/span&gt;be delivered.
As to your tweet, there is room in there for Microsoft.
If we take the fact that Microsoft is not a hardware vendor as the prime factor, then it behooves Microsoft to make sure that their server software designs take these sort of advances into consideration.
They will play in that space, no matter what it is.</description>
		<content:encoded><![CDATA[<p>Hello Nigel,</p>
<p>As you rightfully point out, the FCoE RAN solution is basically an evolutionary step from where we are today.<br />
In a fully-converged architecture (thanks for the moniker, HP), a revolutionary improvement from today&rsquo;s infrastructure <b><i>must </i></b><span>&nbsp;</span>be delivered.<br />
As to your tweet, there is room in there for Microsoft.<br />
If we take the fact that Microsoft is not a hardware vendor as the prime factor, then it behooves Microsoft to make sure that their server software designs take these sort of advances into consideration.<br />
They will play in that space, no matter what it is.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Nigel Poulton</title>
		<link>http://blog.nigelpoulton.com/the-blade-is-dead-long-live-the-rack/comment-page-1/#comment-692</link>
		<dc:creator>Nigel Poulton</dc:creator>
		<pubDate>Mon, 18 Jan 2010 21:19:11 +0000</pubDate>
		<guid isPermaLink="false">http://blog.nigelpoulton.com/the-blade-is-dead-long-live-the-rack/#comment-692</guid>
		<description>&lt;p&gt;J Metz,&lt;/p&gt;
&lt;p&gt;Im planning on writing more about it and a collaborativ post with a couple of other guys.&lt;/p&gt;
&lt;p&gt;A coupoe of companies might include NextIO and VirtenSys.&lt;/p&gt;
&lt;p&gt;Nigel&lt;/p&gt;</description>
		<content:encoded><![CDATA[<p>J Metz,</p>
<p>Im planning on writing more about it and a collaborativ post with a couple of other guys.</p>
<p>A coupoe of companies might include NextIO and VirtenSys.</p>
<p>Nigel</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Nigel Poulton</title>
		<link>http://blog.nigelpoulton.com/the-blade-is-dead-long-live-the-rack/comment-page-1/#comment-691</link>
		<dc:creator>Nigel Poulton</dc:creator>
		<pubDate>Mon, 18 Jan 2010 21:17:09 +0000</pubDate>
		<guid isPermaLink="false">http://blog.nigelpoulton.com/the-blade-is-dead-long-live-the-rack/#comment-691</guid>
		<description>Hi Cam,

Interestingly, when configuring the HP Blade MAtrix solution we dropped a couple of the software components from the &quot;Matrix&quot; solution and suddenly it wasn&#039;t &quot;Matrix&quot; any more.. so went from something like a 2 week &quot;time to ship&quot; to 12 weeks!  Just for taking a couple software licenses out!  Anyway, Im digressing..

As for management - if Im being honest, I try not  to get into management discussion as I would probably rather pull my own teeth out ;-)  Its not a very sexy topic - although I think iPhone type apps are a fresh idea.  On FCoE and CEE/DCB management Im hoping things like ETS, DCBX and may be even some vendor value-add management tools might make things like dynamic b/width management etc simpler.  If you have a single cable for LAN and SAN and manage it all through software then surely that has to be better than the old way of separate networks (cards, cables ports...) for LAN and SAN and adding more cards cables and ports if more b/width is needed....?

Debugging, firmware management and things like that are definitely a concern though in FCoE networks.

Nigel</description>
		<content:encoded><![CDATA[<p>Hi Cam,</p>
<p>Interestingly, when configuring the HP Blade MAtrix solution we dropped a couple of the software components from the &#8220;Matrix&#8221; solution and suddenly it wasn&#8217;t &#8220;Matrix&#8221; any more.. so went from something like a 2 week &#8220;time to ship&#8221; to 12 weeks!  Just for taking a couple software licenses out!  Anyway, Im digressing..</p>
<p>As for management &#8211; if Im being honest, I try not  to get into management discussion as I would probably rather pull my own teeth out <img src='http://blog.nigelpoulton.com/wp-includes/images/smilies/icon_wink.gif' alt=';-)' class='wp-smiley' />   Its not a very sexy topic &#8211; although I think iPhone type apps are a fresh idea.  On FCoE and CEE/DCB management Im hoping things like ETS, DCBX and may be even some vendor value-add management tools might make things like dynamic b/width management etc simpler.  If you have a single cable for LAN and SAN and manage it all through software then surely that has to be better than the old way of separate networks (cards, cables ports&#8230;) for LAN and SAN and adding more cards cables and ports if more b/width is needed&#8230;.?</p>
<p>Debugging, firmware management and things like that are definitely a concern though in FCoE networks.</p>
<p>Nigel</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: J Michel Metz</title>
		<link>http://blog.nigelpoulton.com/the-blade-is-dead-long-live-the-rack/comment-page-1/#comment-686</link>
		<dc:creator>J Michel Metz</dc:creator>
		<pubDate>Fri, 15 Jan 2010 19:52:45 +0000</pubDate>
		<guid isPermaLink="false">http://blog.nigelpoulton.com/the-blade-is-dead-long-live-the-rack/#comment-686</guid>
		<description>This is truly eye-opening stuff. I&#039;ve never thought about running PCIe cables within a rack. I&#039;d be interested to hear more about it. Got any pointers about where to look?</description>
		<content:encoded><![CDATA[<p>This is truly eye-opening stuff. I&#39;ve never thought about running PCIe cables within a rack. I&#39;d be interested to hear more about it. Got any pointers about where to look?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Cam Ford</title>
		<link>http://blog.nigelpoulton.com/the-blade-is-dead-long-live-the-rack/comment-page-1/#comment-685</link>
		<dc:creator>Cam Ford</dc:creator>
		<pubDate>Fri, 15 Jan 2010 19:15:35 +0000</pubDate>
		<guid isPermaLink="false">http://blog.nigelpoulton.com/the-blade-is-dead-long-live-the-rack/#comment-685</guid>
		<description>Hi Nigel,

Interesting article.  You have clearly addressed the Hardware side of the RAN, but what about the management model.  The big question I would ask is......does converged fabrics.....either FCoE/CEE or IOV approach make management easier or more difficult.  How does the RAN change the existing management model in the datacenter?  As you have pointed out, the HP Blade Matrix is just a bundle of what everyone already does today, so there is not much gain in management simplicity with this approach.....only a simplified SKU for ordering....assuming that the stock bundle meets your needs.

Now, I will also make the case that FCoE convegence actually makes management more difficult.  There is a lot of comfort in having separate FC and Ethernet networks.....from a division of responsibility perspective, from a debug perspective, and from a provisioning and management perspective.  On the other hand, I believe that the IOV approach actually simplifies the management model quite a bit.  It readically simplfies provisioning, allows for granular control of expensive FC and Ethernet resources, and provides greater flexibility in change managment....and at he same time keeps the division of responsibility roughy the same.....but you know my view.</description>
		<content:encoded><![CDATA[<p>Hi Nigel,</p>
<p>Interesting article.  You have clearly addressed the Hardware side of the RAN, but what about the management model.  The big question I would ask is&#8230;&#8230;does converged fabrics&#8230;..either FCoE/CEE or IOV approach make management easier or more difficult.  How does the RAN change the existing management model in the datacenter?  As you have pointed out, the HP Blade Matrix is just a bundle of what everyone already does today, so there is not much gain in management simplicity with this approach&#8230;..only a simplified SKU for ordering&#8230;.assuming that the stock bundle meets your needs.</p>
<p>Now, I will also make the case that FCoE convegence actually makes management more difficult.  There is a lot of comfort in having separate FC and Ethernet networks&#8230;..from a division of responsibility perspective, from a debug perspective, and from a provisioning and management perspective.  On the other hand, I believe that the IOV approach actually simplifies the management model quite a bit.  It readically simplfies provisioning, allows for granular control of expensive FC and Ethernet resources, and provides greater flexibility in change managment&#8230;.and at he same time keeps the division of responsibility roughy the same&#8230;..but you know my view.</p>
]]></content:encoded>
	</item>
</channel>
</rss>
