<?xml version="1.0" encoding="utf-8" ?>

<rss version="2.0" 
   xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
   xmlns:admin="http://webns.net/mvcb/"
   xmlns:dc="http://purl.org/dc/elements/1.1/"
   xmlns:slash="http://purl.org/rss/1.0/modules/slash/"
   xmlns:wfw="http://wellformedweb.org/CommentAPI/"
   xmlns:content="http://purl.org/rss/1.0/modules/content/"
   >
<channel>
    
    <title>Open-Source, Physik &amp; Politik - Kommentare</title>
    <link>http://sebastian.marsching.com/blog/</link>
    <description>Open-Source, Physik &amp; Politik - Weblog von Sebastian Marsching</description>
    <dc:language>de</dc:language>
    <generator>Serendipity 1.5.5 - http://www.s9y.org/</generator>
    <pubDate>Wed, 08 Feb 2012 07:41:38 GMT</pubDate>

    <image>
        <url>http://sebastian.marsching.com/blog/templates/bulletproof/img/s9y_banner_small.png</url>
        <title>RSS: Open-Source, Physik &amp; Politik - Kommentare - Open-Source, Physik &amp; Politik - Weblog von Sebastian Marsching</title>
        <link>http://sebastian.marsching.com/blog/</link>
        <width>100</width>
        <height>21</height>
    </image>

<item>
    <title>Matt Brown: KVM and Graceful Shutdown on Ubuntu</title>
    <link>http://sebastian.marsching.com/blog/archives/112-KVM-and-Graceful-Shutdown-on-Ubuntu.html#c2750</link>
            <category></category>
    
    <comments>http://sebastian.marsching.com/blog/archives/112-KVM-and-Graceful-Shutdown-on-Ubuntu.html#comments</comments>
    <wfw:comment>http://sebastian.marsching.com/blog/wfwcomment.php?cid=112</wfw:comment>

    

    <author>nospam@example.com (Matt Brown)</author>
    <content:encoded>
    Thanks! I was struggling with a solution to this. Turns out the sendsigs part was tripping me up. Thanks for the complete solution, works like a charm. 
    </content:encoded>

    <pubDate>Wed, 28 Sep 2011 22:57:42 +0200</pubDate>
    <guid isPermaLink="false">http://sebastian.marsching.com/blog/archives/112-guid.html#c2750</guid>
    
</item>
<item>
    <title>Sebastian Marsching: KVM and Graceful Shutdown on Ubuntu</title>
    <link>http://sebastian.marsching.com/blog/archives/112-KVM-and-Graceful-Shutdown-on-Ubuntu.html#c2747</link>
            <category></category>
    
    <comments>http://sebastian.marsching.com/blog/archives/112-KVM-and-Graceful-Shutdown-on-Ubuntu.html#comments</comments>
    <wfw:comment>http://sebastian.marsching.com/blog/wfwcomment.php?cid=112</wfw:comment>

    

    <author>nospam@example.com (Sebastian Marsching)</author>
    <content:encoded>
    Although upstart and SysV-Init can work together (in fact in Ubuntu they do), making them work is not trivial and I would not recommend installing Upstart on a distribution which does not provide it by default.

However, the good news is, that you can use nearly the same approach with SysV-Init as well:
Before Ubuntu moved libvirt to upstart, I just modified /etc/init.d/libvirt-bin in order to call the shutdown-domains script on the stop action. You use the same /usr/local/bin/libvirt-shutdown-domains script, and add the section from the Upstart job between &quot;script&quot; and &quot;end script&quot; to the stop section of /etc/init.d/libvirt-bin. As this effectively blocks the shutdown process, you do not need the /etc/init.d/await-libvirt-bin-stopped script, which you need when using Upstart. Just make sure, that the code you add to /etc/init.d/libvirt-bin is before libvirtd is stopped, because otherwise the Python script cannot work. 
    </content:encoded>

    <pubDate>Mon, 15 Aug 2011 22:13:50 +0200</pubDate>
    <guid isPermaLink="false">http://sebastian.marsching.com/blog/archives/112-guid.html#c2747</guid>
    
</item>
<item>
    <title>Diego Bendlin: KVM and Graceful Shutdown on Ubuntu</title>
    <link>http://sebastian.marsching.com/blog/archives/112-KVM-and-Graceful-Shutdown-on-Ubuntu.html#c2746</link>
            <category></category>
    
    <comments>http://sebastian.marsching.com/blog/archives/112-KVM-and-Graceful-Shutdown-on-Ubuntu.html#comments</comments>
    <wfw:comment>http://sebastian.marsching.com/blog/wfwcomment.php?cid=112</wfw:comment>

    

    <author>nospam@example.com (Diego Bendlin)</author>
    <content:encoded>
    Sebastian,
I&#039;m trying to implement your scripts as a workarround on this libvirt issue, but i&#039;m using debian squeeze.

It seems that upstart is not installed on my box, afaik upstart is a replacement for sysV scripts, is it safe to install upstart without breaking my system?

Thanks in advance!

Regards,

Diego Bendlin 
    </content:encoded>

    <pubDate>Mon, 01 Aug 2011 18:34:44 +0200</pubDate>
    <guid isPermaLink="false">http://sebastian.marsching.com/blog/archives/112-guid.html#c2746</guid>
    
</item>
<item>
    <title>Sebastian Marsching: KVM and Graceful Shutdown on Ubuntu</title>
    <link>http://sebastian.marsching.com/blog/archives/112-KVM-and-Graceful-Shutdown-on-Ubuntu.html#c2745</link>
            <category></category>
    
    <comments>http://sebastian.marsching.com/blog/archives/112-KVM-and-Graceful-Shutdown-on-Ubuntu.html#comments</comments>
    <wfw:comment>http://sebastian.marsching.com/blog/wfwcomment.php?cid=112</wfw:comment>

    

    <author>nospam@example.com (Sebastian Marsching)</author>
    <content:encoded>
    In fact the problem is not as simple as it might seem: If a simple sequential approach, like the one you suggest, is used, the shutdown process can take a very, very long time. So you rather need a dependency-tree, which is harder to implement.

The script I wrote is mainly intended for emergency-shutdown situations (like the UPS becoming low on battery). Therefore, the main goal is to get all virtual machines into a consistent state as soon as possible.

I think there is no real need for a specific shutdown order. To use your example, if you have an application server and a webserver running the frontend, if the application server is shutdown first, the web frontend will stop working. However, if the web server is shutdown first, the web frontend will not be available either. But the state of the whole system will always be consistent, independent of the shutdown order. If you need to maximize the availability of certain services, I suggest using a Linux-HA setup.

For example, I am using a Heartbeat/Pacemaker setup consisting of two virtual machines running on two different host systems. Pacemaker ensures that the services running in the seconds virtual machine take over, if the first virtual machine fails for some reasons. 
    </content:encoded>

    <pubDate>Sun, 17 Jul 2011 17:16:04 +0200</pubDate>
    <guid isPermaLink="false">http://sebastian.marsching.com/blog/archives/112-guid.html#c2745</guid>
    
</item>
<item>
    <title>Franck: KVM and Graceful Shutdown on Ubuntu</title>
    <link>http://sebastian.marsching.com/blog/archives/112-KVM-and-Graceful-Shutdown-on-Ubuntu.html#c2744</link>
            <category></category>
    
    <comments>http://sebastian.marsching.com/blog/archives/112-KVM-and-Graceful-Shutdown-on-Ubuntu.html#comments</comments>
    <wfw:comment>http://sebastian.marsching.com/blog/wfwcomment.php?cid=112</wfw:comment>

    

    <author>nospam@example.com (Franck)</author>
    <content:encoded>
    Hello,

A real problem to solve ! Unbelievable existing on a server system (ubuntu).

Another problem with shutting down VM is the order of shutting down (app server first please).

It is evident that stopping the host&#039;s VMs is a task for the virtual manager.

vmware (in 2006) had a simple &quot;autostop.order = x&quot; field. May not solve every case.

in libvirt world, the discussion is &#039;it does not solve every case so it is not implemented&#039; !

A workaround could be autostart links to xml files prefixed (10- as init scripts) and your script following that order for killing VM, then remaining VM.

Franck 
    </content:encoded>

    <pubDate>Wed, 13 Jul 2011 13:30:05 +0200</pubDate>
    <guid isPermaLink="false">http://sebastian.marsching.com/blog/archives/112-guid.html#c2744</guid>
    
</item>
<item>
    <title>Sebastian Marsching: Fighting SPAM Bounces with Exim (Part I)</title>
    <link>http://sebastian.marsching.com/blog/archives/97-Fighting-SPAM-Bounces-with-Exim-Part-I.html#c2742</link>
            <category></category>
    
    <comments>http://sebastian.marsching.com/blog/archives/97-Fighting-SPAM-Bounces-with-Exim-Part-I.html#comments</comments>
    <wfw:comment>http://sebastian.marsching.com/blog/wfwcomment.php?cid=97</wfw:comment>

    

    <author>nospam@example.com (Sebastian Marsching)</author>
    <content:encoded>
    You are right, if you are takling about &lt;strong&gt;sender&lt;/strong&gt; callout verification.

However, in my article I was talking about &lt;strong&gt;recipient&lt;/strong&gt; callout verification.

The latter one is very useful to limit the number of bounce mails produced. Doing callout verification on recipient addresses will enable you to report errors on wrong adresses during the SMTP dialog. If it is not enabled, the secondary MX will accept e-mails for all localparts of a valid domain and generate bounce mails when it tries to deliver this e-mails to the primary MX and detects that some of these localparts are invalid in fact.

So, in a short summary: Sender callout verification bad, recipient callout verification good &lt;img src=&quot;http://sebastian.marsching.com/blog/templates/default/img/emoticons/smile.png&quot; alt=&quot;:-)&quot; style=&quot;display: inline; vertical-align: bottom;&quot; class=&quot;emoticon&quot; /&gt; 
    </content:encoded>

    <pubDate>Wed, 24 Feb 2010 16:06:28 +0100</pubDate>
    <guid isPermaLink="false">http://sebastian.marsching.com/blog/archives/97-guid.html#c2742</guid>
    
</item>
<item>
    <title>James: Fighting SPAM Bounces with Exim (Part I)</title>
    <link>http://sebastian.marsching.com/blog/archives/97-Fighting-SPAM-Bounces-with-Exim-Part-I.html#c2741</link>
            <category></category>
    
    <comments>http://sebastian.marsching.com/blog/archives/97-Fighting-SPAM-Bounces-with-Exim-Part-I.html#comments</comments>
    <wfw:comment>http://sebastian.marsching.com/blog/wfwcomment.php?cid=97</wfw:comment>

    

    <author>nospam@example.com (James)</author>
    <content:encoded>
    Please note that sender callout verification has some very serious problems and is depreciated nowadays. Please read about the problems in detail (google it). In short:

1. It opens up your server for mounting distributed denial of service attacks on third parties.

2. Some legitimate mail servers blacklist you as a spammer if you use sender callout, as they feel you waste their resources.

3. You will refuse legitimate mail from any server that uses itself sender callout. Simply because your callout will be answered with a callout from them which you will again answer with a callout -- ergo you get a loop and the mail bounces. This is a fundamental flaw of callout verification!

In short: Don&#039;t use callout verification! 
    </content:encoded>

    <pubDate>Mon, 22 Feb 2010 09:08:56 +0100</pubDate>
    <guid isPermaLink="false">http://sebastian.marsching.com/blog/archives/97-guid.html#c2741</guid>
    
</item>
<item>
    <title>Kanto: LHC erfolgreich gestartet</title>
    <link>http://sebastian.marsching.com/blog/archives/93-LHC-erfolgreich-gestartet.html#c2740</link>
            <category></category>
    
    <comments>http://sebastian.marsching.com/blog/archives/93-LHC-erfolgreich-gestartet.html#comments</comments>
    <wfw:comment>http://sebastian.marsching.com/blog/wfwcomment.php?cid=93</wfw:comment>

    

    <author>nospam@example.com (Kanto)</author>
    <content:encoded>
    Bald gehts los. Noch 107 Tage? Nach dieser Seite:
www.lhc-facts.ch 
    </content:encoded>

    <pubDate>Sat, 01 Aug 2009 21:24:17 +0200</pubDate>
    <guid isPermaLink="false">http://sebastian.marsching.com/blog/archives/93-guid.html#c2740</guid>
    
</item>
<item>
    <title>Tina: Schockierend</title>
    <link>http://sebastian.marsching.com/blog/archives/17-Schockierend.html#c2739</link>
            <category></category>
    
    <comments>http://sebastian.marsching.com/blog/archives/17-Schockierend.html#comments</comments>
    <wfw:comment>http://sebastian.marsching.com/blog/wfwcomment.php?cid=17</wfw:comment>

    

    <author>nospam@example.com (Tina)</author>
    <content:encoded>
    Beim Lesen meiner Page muss man nun nichts mehr in den Editor kopieren, denn ich habe die &quot;Augenkrebsgefahr&quot; gebannt! ;o) 
    </content:encoded>

    <pubDate>Sat, 18 Apr 2009 06:50:57 +0200</pubDate>
    <guid isPermaLink="false">http://sebastian.marsching.com/blog/archives/17-guid.html#c2739</guid>
    
</item>
<item>
    <title>Sebastian: Making Xen work on Ubuntu 8.04 Hardy Heron</title>
    <link>http://sebastian.marsching.com/blog/archives/89-Making-Xen-work-on-Ubuntu-8.04-Hardy-Heron.html#c2738</link>
            <category></category>
    
    <comments>http://sebastian.marsching.com/blog/archives/89-Making-Xen-work-on-Ubuntu-8.04-Hardy-Heron.html#comments</comments>
    <wfw:comment>http://sebastian.marsching.com/blog/wfwcomment.php?cid=89</wfw:comment>

    

    <author>nospam@example.com (Sebastian)</author>
    <content:encoded>
    Natürlich, solange auf die Original-Quelle verwiesen und vorher gefragt wird, habe ich keine Einwände.

Das oben beschriebene Vorgehen hat mir übrigens auch geholfen, nachdem Xen nach einem &quot;aptitude dist-upgrade&quot; plötzlich gestreikt hat. Nachdem ich dann manuell auf Xen 3.3 geupdated hatte, funktionierte wieder alles... 
    </content:encoded>

    <pubDate>Sun, 08 Feb 2009 22:11:50 +0100</pubDate>
    <guid isPermaLink="false">http://sebastian.marsching.com/blog/archives/89-guid.html#c2738</guid>
    
</item>
<item>
    <title>Marek Koch: Making Xen work on Ubuntu 8.04 Hardy Heron</title>
    <link>http://sebastian.marsching.com/blog/archives/89-Making-Xen-work-on-Ubuntu-8.04-Hardy-Heron.html#c2737</link>
            <category></category>
    
    <comments>http://sebastian.marsching.com/blog/archives/89-Making-Xen-work-on-Ubuntu-8.04-Hardy-Heron.html#comments</comments>
    <wfw:comment>http://sebastian.marsching.com/blog/wfwcomment.php?cid=89</wfw:comment>

    

    <author>nospam@example.com (Marek Koch)</author>
    <content:encoded>
    Wow, ich hoffe das es alles so klappt =)
Aber im Prinzip erst ma genau so eine Upgradeanleitung habe ich gesucht. Bei mir ist nämlich zur Zeit das Problem das eine domU einfach nicht starten will und auch in den Xen Logs nicht erkennbar ist, warum die nicht starten will. Macht ziemlich gleich nachm Start 100% Last und das war&#039;s dann. xm shutdown oder destroy sind dann meine einzigen Freunde.
Würde diese Upgrade Anleitung mit Verweis auf diese Originalquelle und Autor gerne zitieren. Darf ich?

Gruß Marek Koch 
    </content:encoded>

    <pubDate>Thu, 29 Jan 2009 10:01:43 +0100</pubDate>
    <guid isPermaLink="false">http://sebastian.marsching.com/blog/archives/89-guid.html#c2737</guid>
    
</item>
<item>
    <title>Mao-B: Wohl kein April-Scherz...</title>
    <link>http://sebastian.marsching.com/blog/archives/84-Wohl-kein-April-Scherz....html#c2732</link>
            <category></category>
    
    <comments>http://sebastian.marsching.com/blog/archives/84-Wohl-kein-April-Scherz....html#comments</comments>
    <wfw:comment>http://sebastian.marsching.com/blog/wfwcomment.php?cid=84</wfw:comment>

    

    <author>nospam@example.com (Mao-B)</author>
    <content:encoded>
    ...vielleicht meinte er ja das schwarze Loch im Forschungsetat oder so... &lt;img src=&quot;http://sebastian.marsching.com/blog/templates/default/img/emoticons/wink.png&quot; alt=&quot;;-)&quot; style=&quot;display: inline; vertical-align: bottom;&quot; class=&quot;emoticon&quot; /&gt; 
    </content:encoded>

    <pubDate>Sat, 19 Apr 2008 22:34:58 +0200</pubDate>
    <guid isPermaLink="false">http://sebastian.marsching.com/blog/archives/84-guid.html#c2732</guid>
    
</item>
<item>
    <title>michael: Warum noch ein Weblog?</title>
    <link>http://sebastian.marsching.com/blog/archives/1-Warum-noch-ein-Weblog.html#c157</link>
            <category></category>
    
    <comments>http://sebastian.marsching.com/blog/archives/1-Warum-noch-ein-Weblog.html#comments</comments>
    <wfw:comment>http://sebastian.marsching.com/blog/wfwcomment.php?cid=1</wfw:comment>

    

    <author>nospam@example.com (michael)</author>
    <content:encoded>
    hab auch ein blog. aber warum? keine ahnung... &lt;strong&gt;g&lt;/strong&gt; 
    </content:encoded>

    <pubDate>Sun, 22 Oct 2006 07:38:35 +0200</pubDate>
    <guid isPermaLink="false">http://sebastian.marsching.com/blog/archives/1-guid.html#c157</guid>
    
</item>
<item>
    <title>Daniel: Playstation 3 vs. Nintendo Wii</title>
    <link>http://sebastian.marsching.com/blog/archives/56-Playstation-3-vs.-Nintendo-Wii.html#c117</link>
            <category></category>
    
    <comments>http://sebastian.marsching.com/blog/archives/56-Playstation-3-vs.-Nintendo-Wii.html#comments</comments>
    <wfw:comment>http://sebastian.marsching.com/blog/wfwcomment.php?cid=56</wfw:comment>

    

    <author>nospam@example.com (Daniel)</author>
    <content:encoded>
    Hi,

ich hatte bisher alles was es von Sega gab (bis aufs Master System), danach die PS und PS2 und jetz liebäugel ich mit der PS3.

Microsofts XBOX ist doch für PC-User schon fast wieder ein Gimmick - die meisten Games gibts auch für PC und für alles außer zocken nehm ich ebenfalls lieber den PC.

Bleibt Nintendo - aber! Egal wie kniffig die Hardware, der Preis oder die Werbung ist, es bleiben Spiele fürs Nintendo. Bunt, knuffig, fluffig, die gleichen kleinen Klemptnerhelden und rosa bekleideten blondhaarigen Prinzessinen. 

Die Konsolenentscheidung würde ich nicht an der Technik festmachen (zumal komischer weise die Spiele nach 2-3 Jahren auf einmal gigantisch aussehen im Vgl. zu den ersten Spielen auf einer Konsole) - sondern an der Zielgruppe! Und bei Resident Evil, Metal Gear Solid, Wipeout und Final Fantasy... hmmm...

PS3 - zumindest für mich, denn PS1 und PS2 laufen auch noch auf der neuen Konsole!

So, musste ich irgendwo loswerden XD

Bye,
Dani 
    </content:encoded>

    <pubDate>Thu, 24 Aug 2006 17:16:00 +0200</pubDate>
    <guid isPermaLink="false">http://sebastian.marsching.com/blog/archives/56-guid.html#c117</guid>
    
</item>
<item>
    <title>Drky: World of Warcraft als Lernsoftware?</title>
    <link>http://sebastian.marsching.com/blog/archives/54-World-of-Warcraft-als-Lernsoftware.html#c90</link>
            <category></category>
    
    <comments>http://sebastian.marsching.com/blog/archives/54-World-of-Warcraft-als-Lernsoftware.html#comments</comments>
    <wfw:comment>http://sebastian.marsching.com/blog/wfwcomment.php?cid=54</wfw:comment>

    

    <author>nospam@example.com (Drky)</author>
    <content:encoded>
    Es gibt tatsächlich Leute, die das, was dort geschrieben wird, auch lesen. Aber leider gibt es immer wieder Leute, die meinen sich mit ihrer Intelligenz in sämtlichen Foren kundtun zu müssen.

Nun ich finde diese Erkenntnis über WoW (selber Zocker &lt;img src=&quot;http://sebastian.marsching.com/blog/templates/default/img/emoticons/wink.png&quot; alt=&quot;;-)&quot; style=&quot;display: inline; vertical-align: bottom;&quot; class=&quot;emoticon&quot; /&gt; schon interessant. Allerdings glaube ich nicht, daß die emotionale Entwicklung wirklich gefördert wird, da man Spiel nie so perönlich kennenlernt wie in der Realität. Und dort zählen ebensolche und nicht in einer slebstgebastelten Computerwelt. 
    </content:encoded>

    <pubDate>Mon, 31 Jul 2006 16:44:22 +0200</pubDate>
    <guid isPermaLink="false">http://sebastian.marsching.com/blog/archives/54-guid.html#c90</guid>
    
</item>

</channel>
</rss>
