<?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: lockdown and the di-semi-default route</title>
	<atom:link href="http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/feed/" rel="self" type="application/rss+xml" />
	<link>http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/</link>
	<description>a lowercase manifesto</description>
	<lastBuildDate>Thu, 15 Oct 2009 09:25:50 +0000</lastBuildDate>
	<generator>http://wordpress.org/?v=2.8.6</generator>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
		<item>
		<title>By: Ilya</title>
		<link>http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/comment-page-1/#comment-388</link>
		<dc:creator>Ilya</dc:creator>
		<pubDate>Mon, 01 Oct 2007 14:47:32 +0000</pubDate>
		<guid isPermaLink="false">http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/#comment-388</guid>
		<description>Regarding finding the route through which the VPN is reached, you should probably check what &quot;ip route get 1.2.3.4&quot; does and do the same. It might work just as you do, by manually evaluating the routing rules in userspace, but perhaps it does something smarter, taking the more-complex routing options into account.</description>
		<content:encoded><![CDATA[<p>Regarding finding the route through which the VPN is reached, you should probably check what &#8220;ip route get 1.2.3.4&#8243; does and do the same. It might work just as you do, by manually evaluating the routing rules in userspace, but perhaps it does something smarter, taking the more-complex routing options into account.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: M Welinder</title>
		<link>http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/comment-page-1/#comment-386</link>
		<dc:creator>M Welinder</dc:creator>
		<pubDate>Mon, 01 Oct 2007 13:26:36 +0000</pubDate>
		<guid isPermaLink="false">http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/#comment-386</guid>
		<description>If you want to isolate a process, you also need to make
sure it cannot attach as a debugger to any process
outside the jail.</description>
		<content:encoded><![CDATA[<p>If you want to isolate a process, you also need to make<br />
sure it cannot attach as a debugger to any process<br />
outside the jail.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: blah</title>
		<link>http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/comment-page-1/#comment-384</link>
		<dc:creator>blah</dc:creator>
		<pubDate>Mon, 01 Oct 2007 12:15:21 +0000</pubDate>
		<guid isPermaLink="false">http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/#comment-384</guid>
		<description>desrt: Also, on systems like Debian you can set the metric for each interface explicitly by using the &quot;metric&quot; stanza in /etc/network/interfaces.

And there is this nifty tool called &quot;ifmetric&quot; which modifies all current routes on a specific interface to a certain value.</description>
		<content:encoded><![CDATA[<p>desrt: Also, on systems like Debian you can set the metric for each interface explicitly by using the &#8220;metric&#8221; stanza in /etc/network/interfaces.</p>
<p>And there is this nifty tool called &#8220;ifmetric&#8221; which modifies all current routes on a specific interface to a certain value.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: blah</title>
		<link>http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/comment-page-1/#comment-383</link>
		<dc:creator>blah</dc:creator>
		<pubDate>Mon, 01 Oct 2007 12:13:25 +0000</pubDate>
		<guid isPermaLink="false">http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/#comment-383</guid>
		<description>desrt: On Linux routing metrics can be negative!</description>
		<content:encoded><![CDATA[<p>desrt: On Linux routing metrics can be negative!</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Stef Walter</title>
		<link>http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/comment-page-1/#comment-377</link>
		<dc:creator>Stef Walter</dc:creator>
		<pubDate>Mon, 01 Oct 2007 05:52:45 +0000</pubDate>
		<guid isPermaLink="false">http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/#comment-377</guid>
		<description>di-semi-default route I never thought of that.  Learned something handy today.</description>
		<content:encoded><![CDATA[<p>di-semi-default route I never thought of that.  Learned something handy today.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: desrt</title>
		<link>http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/comment-page-1/#comment-374</link>
		<dc:creator>desrt</dc:creator>
		<pubDate>Mon, 01 Oct 2007 00:32:08 +0000</pubDate>
		<guid isPermaLink="false">http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/#comment-374</guid>
		<description>blah: the problem with using route metrics is that in almost every case ever the default route already has metric 0 :(</description>
		<content:encoded><![CDATA[<p>blah: the problem with using route metrics is that in almost every case ever the default route already has metric 0 :(</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: blah</title>
		<link>http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/comment-page-1/#comment-373</link>
		<dc:creator>blah</dc:creator>
		<pubDate>Mon, 01 Oct 2007 00:27:45 +0000</pubDate>
		<guid isPermaLink="false">http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/#comment-373</guid>
		<description>Dude, that routing stuff you are suggesting  is pure crack!

Use route metrics for this! That way you may have multiple default routes, and can specify which one is the higher priority one. It&#039;s easy-to-use and actually invinted for cases like this one.</description>
		<content:encoded><![CDATA[<p>Dude, that routing stuff you are suggesting  is pure crack!</p>
<p>Use route metrics for this! That way you may have multiple default routes, and can specify which one is the higher priority one. It&#8217;s easy-to-use and actually invinted for cases like this one.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: blah</title>
		<link>http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/comment-page-1/#comment-372</link>
		<dc:creator>blah</dc:creator>
		<pubDate>Mon, 01 Oct 2007 00:24:13 +0000</pubDate>
		<guid isPermaLink="false">http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/#comment-372</guid>
		<description>You might also want to set setrlimit(RLIMIT_FSIZE) to 0. That way, a process can no longer create any files on disk anymore.

Also, consider setting RLIMIT_NOFILE to your highest fd+1. That way, the process may not create any further file descriptors.

Just paranoia, though.</description>
		<content:encoded><![CDATA[<p>You might also want to set setrlimit(RLIMIT_FSIZE) to 0. That way, a process can no longer create any files on disk anymore.</p>
<p>Also, consider setting RLIMIT_NOFILE to your highest fd+1. That way, the process may not create any further file descriptors.</p>
<p>Just paranoia, though.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Matthew W. S. Bell</title>
		<link>http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/comment-page-1/#comment-371</link>
		<dc:creator>Matthew W. S. Bell</dc:creator>
		<pubDate>Sun, 30 Sep 2007 23:31:14 +0000</pubDate>
		<guid isPermaLink="false">http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/#comment-371</guid>
		<description>This feels like badness to me</description>
		<content:encoded><![CDATA[<p>This feels like badness to me</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: desrt</title>
		<link>http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/comment-page-1/#comment-370</link>
		<dc:creator>desrt</dc:creator>
		<pubDate>Sun, 30 Sep 2007 21:51:11 +0000</pubDate>
		<guid isPermaLink="false">http://blogs.gnome.org/desrt/2007/09/30/lockdown-and-the-di-semi-default-route/#comment-370</guid>
		<description>craig: i didn&#039;t consider this problem.

i think the only reasonable thing to do here is to query the route table to see what route will match the ip of the vpn server.  if it is the default route, then add another rule -- if not, then do nothing.

thanks for the note.  i was so excited about the other part that i didn&#039;t give this part due consideration.  i now have something to think about :)</description>
		<content:encoded><![CDATA[<p>craig: i didn&#8217;t consider this problem.</p>
<p>i think the only reasonable thing to do here is to query the route table to see what route will match the ip of the vpn server.  if it is the default route, then add another rule &#8212; if not, then do nothing.</p>
<p>thanks for the note.  i was so excited about the other part that i didn&#8217;t give this part due consideration.  i now have something to think about :)</p>
]]></content:encoded>
	</item>
</channel>
</rss>
