<?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"
	>
<channel>
	<title>Comments on: Using Modsec2sguil for HTTP transaction logging</title>
	<atom:link href="http://www.inliniac.net/blog/2007/08/15/using-modsec2sguil-for-http-transaction-logging.html/feed" rel="self" type="application/rss+xml" />
	<link>http://www.inliniac.net/blog/2007/08/15/using-modsec2sguil-for-http-transaction-logging.html</link>
	<description>Everything inline.</description>
	<pubDate>Sat, 22 Nov 2008 09:19:56 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.6.3</generator>
		<item>
		<title>By: Ryan Barnett</title>
		<link>http://www.inliniac.net/blog/2007/08/15/using-modsec2sguil-for-http-transaction-logging.html#comment-7255</link>
		<dc:creator>Ryan Barnett</dc:creator>
		<pubDate>Thu, 16 Aug 2007 20:20:40 +0000</pubDate>
		<guid isPermaLink="false">http://www.inliniac.net/blog/2007/08/15/using-modsec2sguil-for-http-transaction-logging.html#comment-7255</guid>
		<description>This is a great idea from a forensic/IR perspective.  Any time a valid alert is identified and you start the IR process, the first thing that customers ask is "What else did this person do?"  Without the full audit trail of HTTP transactions, you have no visibility into what they did before or after the alert.  This will allow you to quickly do session reconstruction by simply querying for the IP address.  Better yet, you could query based on a sessionid/cookie value to uniquely rebuild sessions.  For this to work, however, it seems that the agent script would need to be able to parse the data into proper fields.  

It is true though that there is a performance hit when you set the modsecurity audit engine to on (meaning log everything) vs. what most people use which is the relevantonly setting (which will only generate an audit log entry when a rule matches).  There are some optimizations that could be made to lighten the load somewhat such as to filter out "static" content requests for images.

Even with some performance hurdles, this is exciting stuff :)</description>
		<content:encoded><![CDATA[<p>This is a great idea from a forensic/IR perspective.  Any time a valid alert is identified and you start the IR process, the first thing that customers ask is &#8220;What else did this person do?&#8221;  Without the full audit trail of HTTP transactions, you have no visibility into what they did before or after the alert.  This will allow you to quickly do session reconstruction by simply querying for the IP address.  Better yet, you could query based on a sessionid/cookie value to uniquely rebuild sessions.  For this to work, however, it seems that the agent script would need to be able to parse the data into proper fields.  </p>
<p>It is true though that there is a performance hit when you set the modsecurity audit engine to on (meaning log everything) vs. what most people use which is the relevantonly setting (which will only generate an audit log entry when a rule matches).  There are some optimizations that could be made to lighten the load somewhat such as to filter out &#8220;static&#8221; content requests for images.</p>
<p>Even with some performance hurdles, this is exciting stuff <img src='http://www.inliniac.net/blog/wp-includes/images/smilies/icon_smile.gif' alt=':)' class='wp-smiley' /></p>
]]></content:encoded>
	</item>
	<item>
		<title>By: VictorJ</title>
		<link>http://www.inliniac.net/blog/2007/08/15/using-modsec2sguil-for-http-transaction-logging.html#comment-7241</link>
		<dc:creator>VictorJ</dc:creator>
		<pubDate>Thu, 16 Aug 2007 01:46:07 +0000</pubDate>
		<guid isPermaLink="false">http://www.inliniac.net/blog/2007/08/15/using-modsec2sguil-for-http-transaction-logging.html#comment-7241</guid>
		<description>The extra information is indeed what I had in mind. What this approach would have in advance above the normal webserver logs, is that request headers, response headers and the post payload are also available in the event Sguil would receive.</description>
		<content:encoded><![CDATA[<p>The extra information is indeed what I had in mind. What this approach would have in advance above the normal webserver logs, is that request headers, response headers and the post payload are also available in the event Sguil would receive.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Richard Bejtlich</title>
		<link>http://www.inliniac.net/blog/2007/08/15/using-modsec2sguil-for-http-transaction-logging.html#comment-7228</link>
		<dc:creator>Richard Bejtlich</dc:creator>
		<pubDate>Wed, 15 Aug 2007 19:58:01 +0000</pubDate>
		<guid isPermaLink="false">http://www.inliniac.net/blog/2007/08/15/using-modsec2sguil-for-http-transaction-logging.html#comment-7228</guid>
		<description>Cool, so now Sguil can accept Web logs in addition to Web security events.  :)  I'm all for extra information if your setup can handle it.  Exporting Web logs via another method is usually preferred, I imagine?</description>
		<content:encoded><![CDATA[<p>Cool, so now Sguil can accept Web logs in addition to Web security events.  <img src='http://www.inliniac.net/blog/wp-includes/images/smilies/icon_smile.gif' alt=':)' class='wp-smiley' />  I&#8217;m all for extra information if your setup can handle it.  Exporting Web logs via another method is usually preferred, I imagine?</p>
]]></content:encoded>
	</item>
</channel>
</rss>
