<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:wfw="http://wellformedweb.org/CommentAPI/"
	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:slash="http://purl.org/rss/1.0/modules/slash/"
	>

<channel>
	<title>Andreas Wundsam</title>
	<atom:link href="http://wundsam.net/feed/" rel="self" type="application/rss+xml" />
	<link>http://wundsam.net</link>
	<description>Musings on Networks, Systems and Software Engineering</description>
	<lastBuildDate>Thu, 20 May 2010 14:09:25 +0000</lastBuildDate>
	<generator>http://wordpress.org/?v=2.9.1</generator>
	<language>en</language>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
			<item>
		<title>Ubuntu 10.04 lucid requires 2.6.32 &#8211; breaks XenServer, Hardy, Lenny Xen Hosts</title>
		<link>http://wundsam.net/2010/05/20/ubuntu-10-04-lucid-requires-2-6-32-excludes-xenserver-hardylenny-xen/</link>
		<comments>http://wundsam.net/2010/05/20/ubuntu-10-04-lucid-requires-2-6-32-excludes-xenserver-hardylenny-xen/#comments</comments>
		<pubDate>Thu, 20 May 2010 13:18:59 +0000</pubDate>
		<dc:creator>Andi</dc:creator>
				<category><![CDATA[systems]]></category>

		<guid isPermaLink="false">http://wundsam.net/?p=55</guid>
		<description><![CDATA[If you&#8217;re thinking about running the recently released Ubuntu 10.04 LTS &#8220;Lucid Lynx&#8221; as virtualized guest for a XEN machine hosted by Citrix&#8217;s XENServer, Ubuntu&#8217;s own Hardy 8.04LTS release, or Debian&#8217;s Lenny, don&#8217;t!
As I found out the hard way yesterday, there is no way you can run a Lucid system with a kernel &#60;= 2.6.32. [...]]]></description>
			<content:encoded><![CDATA[<p>If you&#8217;re thinking about running the recently released Ubuntu 10.04 LTS &#8220;Lucid Lynx&#8221; as virtualized guest for a XEN machine hosted by Citrix&#8217;s XENServer, Ubuntu&#8217;s own Hardy 8.04LTS release, or Debian&#8217;s Lenny, don&#8217;t!</p>
<p>As I found out the hard way yesterday, there is no way you can run a Lucid system with a kernel &lt;= 2.6.32. The reason is that Ubuntu&#8217;s replacement for init, upstart, recently revamped for even faster Desktop startup times, specifically its mount_all component, call experimental system calls only introduced with this kernel version. A fact that&#8217;s conveniently missing from their release document boasting about Cloud integration features.</p>
<p>Apparently it has not crossed their mind  (or they don&#8217;t care) that this excludes Lucid from a whole army of useful server deployment scenarios. For instance, Citrix&#8217; frequently used Virtualization host XENServer comes with a 2.6.18 kernel with backported drivers. Debian&#8217;s recent lenny runs 2.6.26,  even Ubuntu&#8217;s own hardy, a very stable platform as a XEN Dom0 host, runs 2.6.24. While it is possible to run differing kernels in Dom0 and DomU that&#8217;s generally considered a bad idea because of frequent incompatibilities in the past.  There&#8217;s loads of other reasons you may need an older kernel to work besides paravirtualization, e.g., specialized hardware drivers.</p>
<pre>libudev: udev_monitor_new_from_netlink: error getting socket:  Invalid argument
mountall:mountall.c:2955: Assertion failed in main:  udev_monitor = udev_monitor_new_from_netlink (udev, "udev")</pre>
<p>Obviously the design decision by Ubuntu to reduce Desktops boot times by revamping and parallelizing upstart has jeopardized its usability in many server scenarios.</p>
<p>What&#8217;s also increasingly annoying about these mishaps is the non-communication by Ubuntu. Only in the<a href="https://bugs.launchpad.net/ubuntu/+source/mountall/+bug/516684"> launchpad discussion about this specific bug</a>, you&#8217;ll find a confirmation by <a href="https://launchpad.net/%7Escott">Scott  James Remna</a>t writes &#8220;Sorry, we simply do not support earlier kernel versions.  The plumbing  layer is now tightly integrated with the kernel..&#8221;. This seems like a pretty important decision for a distribution seriously aiming at the server platform. One that should be discussed with the users, and prominently annouced, instead of being hidden away in a Bug discussion.</p>
<p>Maybe administrators will have to come to terms with the observation that Ubuntu is a distribution aimed exclusively at the Desktop , aiming at imitating Apple&#8217;s ease-of-use for end users, and increasingly sacrificing its Debian-inherited server friendly roots in turn.</p>
]]></content:encoded>
			<wfw:commentRss>http://wundsam.net/2010/05/20/ubuntu-10-04-lucid-requires-2-6-32-excludes-xenserver-hardylenny-xen/feed/</wfw:commentRss>
		<slash:comments>1</slash:comments>
		</item>
		<item>
		<title>Google patents Map/Reduce</title>
		<link>http://wundsam.net/2010/01/20/google-patents-mapreduce/</link>
		<comments>http://wundsam.net/2010/01/20/google-patents-mapreduce/#comments</comments>
		<pubDate>Wed, 20 Jan 2010 13:52:19 +0000</pubDate>
		<dc:creator>Andi</dc:creator>
				<category><![CDATA[software]]></category>
		<category><![CDATA[algorithms]]></category>
		<category><![CDATA[google]]></category>

		<guid isPermaLink="false">http://www-new.wundsam.net/?p=45</guid>
		<description><![CDATA[Heise and others report today that big G has acquired a patent on the Map/Reduce algorithm,  a critical algorithmic building block to their high scalability as a distributed system built from commodity nodes. Map/Reduce is recognized as a major advance to organizing distributed computations these days, useful for many different application scenarios besides search.
From their perspective, [...]]]></description>
			<content:encoded><![CDATA[<p><a href="http://www.h-online.com/open/news/item/Google-patents-Map-Reduce-908602.html">Heise</a> and others report today that big <a href="http://www.google.com">G</a> has acquired a patent on the <a href="http://labs.google.com/papers/mapreduce.html">Map/Reduce algorithm</a>,  a critical algorithmic building block to their high scalability as a distributed system built from commodity nodes. Map/Reduce is recognized as a major advance to organizing distributed computations these days, useful for many different application scenarios besides search.</p>
<p>From their perspective, it  is understandable that Google would want to protect such a crucial piece of their success from abuse by patent trolls. However, given the ubiquitous usefulness of the approach, it have been great to see this approach formally opened for anyone to use and benefit from.</p>
<p>But, hey, that&#8217;s not the business world we live in these days.</p>
<p>So for the next 20 years, anyone using this great approach is formally at the goodwill of  Google. Another reason to pray that they won&#8217;t change their <a href="http://en.wikipedia.org/wiki/Don't_be_evil">motto</a> anytime soon.</p>
]]></content:encoded>
			<wfw:commentRss>http://wundsam.net/2010/01/20/google-patents-mapreduce/feed/</wfw:commentRss>
		<slash:comments>2</slash:comments>
		</item>
	</channel>
</rss>
