<?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: New mechanism: kcollect	</title>
	<atom:link href="https://www.dragonflydigest.com/2017/08/07/new-mechanism-kcollect/feed/" rel="self" type="application/rss+xml" />
	<link>https://www.dragonflydigest.com/2017/08/07/new-mechanism-kcollect/</link>
	<description>A running description of activity related to DragonFly BSD.</description>
	<lastBuildDate>Tue, 08 Aug 2017 19:19:56 +0000</lastBuildDate>
	<sy:updatePeriod>
	hourly	</sy:updatePeriod>
	<sy:updateFrequency>
	1	</sy:updateFrequency>
	<generator>https://wordpress.org/?v=6.9.4</generator>
	<item>
		<title>
		By: SolarFlame		</title>
		<link>https://www.dragonflydigest.com/2017/08/07/new-mechanism-kcollect/comment-page-1/#comment-486679</link>

		<dc:creator><![CDATA[SolarFlame]]></dc:creator>
		<pubDate>Tue, 08 Aug 2017 19:19:56 +0000</pubDate>
		<guid isPermaLink="false">https://www.dragonflydigest.com/?p=20061#comment-486679</guid>

					<description><![CDATA[@Jenkins

Yes people do…

http://dtrace.org/blogs/brendan/2012/08/09/10-performance-wins/

https://milek.blogspot.de/2012/10/running-openafs-on-solaris-11-x86-zfs.html

and the all-time favourite question do I need to kldload aesni to make openssl AES go faster?!

https://crc32c.blogspot.com/2017/05/freebsd-aesni4-and-openssl.html

This list could go on and on…]]></description>
			<content:encoded><![CDATA[<p>@Jenkins</p>
<p>Yes people do…</p>
<p><a href="http://dtrace.org/blogs/brendan/2012/08/09/10-performance-wins/" rel="nofollow ugc">http://dtrace.org/blogs/brendan/2012/08/09/10-performance-wins/</a></p>
<p><a href="https://milek.blogspot.de/2012/10/running-openafs-on-solaris-11-x86-zfs.html" rel="nofollow ugc">https://milek.blogspot.de/2012/10/running-openafs-on-solaris-11-x86-zfs.html</a></p>
<p>and the all-time favourite question do I need to kldload aesni to make openssl AES go faster?!</p>
<p><a href="https://crc32c.blogspot.com/2017/05/freebsd-aesni4-and-openssl.html" rel="nofollow ugc">https://crc32c.blogspot.com/2017/05/freebsd-aesni4-and-openssl.html</a></p>
<p>This list could go on and on…</p>
]]></content:encoded>
		
			</item>
		<item>
		<title>
		By: Kacy		</title>
		<link>https://www.dragonflydigest.com/2017/08/07/new-mechanism-kcollect/comment-page-1/#comment-486678</link>

		<dc:creator><![CDATA[Kacy]]></dc:creator>
		<pubDate>Tue, 08 Aug 2017 18:28:33 +0000</pubDate>
		<guid isPermaLink="false">https://www.dragonflydigest.com/?p=20061#comment-486678</guid>

					<description><![CDATA[On the topic of system load, a Netflix engineer has a great blog post today on what system load even means and how to interrupt kernel system data. 

http://www.brendangregg.com/blog/2017-08-08/linux-load-averages.html]]></description>
			<content:encoded><![CDATA[<p>On the topic of system load, a Netflix engineer has a great blog post today on what system load even means and how to interrupt kernel system data. </p>
<p><a href="http://www.brendangregg.com/blog/2017-08-08/linux-load-averages.html" rel="nofollow ugc">http://www.brendangregg.com/blog/2017-08-08/linux-load-averages.html</a></p>
]]></content:encoded>
		
			</item>
		<item>
		<title>
		By: Justin Sherrill		</title>
		<link>https://www.dragonflydigest.com/2017/08/07/new-mechanism-kcollect/comment-page-1/#comment-486677</link>

		<dc:creator><![CDATA[Justin Sherrill]]></dc:creator>
		<pubDate>Tue, 08 Aug 2017 17:26:13 +0000</pubDate>
		<guid isPermaLink="false">https://www.dragonflydigest.com/?p=20061#comment-486677</guid>

					<description><![CDATA[For what it&#039;s worth, I think DTrace and kcollect are solving two different problems.  kcollect is very light and was written in an afternoon, as far as I can tell, to assist with problem diagnosis in as automatic a way as possible - autocollection, graph generation, etc.  DTrace is very capable but it&#039;s a porting effort and a different approach.

I saw this without directly interacting with either right now, so I could be wrong.]]></description>
			<content:encoded><![CDATA[<p>For what it&#8217;s worth, I think DTrace and kcollect are solving two different problems.  kcollect is very light and was written in an afternoon, as far as I can tell, to assist with problem diagnosis in as automatic a way as possible &#8211; autocollection, graph generation, etc.  DTrace is very capable but it&#8217;s a porting effort and a different approach.</p>
<p>I saw this without directly interacting with either right now, so I could be wrong.</p>
]]></content:encoded>
		
			</item>
		<item>
		<title>
		By: Jenkins		</title>
		<link>https://www.dragonflydigest.com/2017/08/07/new-mechanism-kcollect/comment-page-1/#comment-486676</link>

		<dc:creator><![CDATA[Jenkins]]></dc:creator>
		<pubDate>Tue, 08 Aug 2017 16:47:23 +0000</pubDate>
		<guid isPermaLink="false">https://www.dragonflydigest.com/?p=20061#comment-486676</guid>

					<description><![CDATA[Serious question. Do people find DTrace useful and a good tool?]]></description>
			<content:encoded><![CDATA[<p>Serious question. Do people find DTrace useful and a good tool?</p>
]]></content:encoded>
		
			</item>
		<item>
		<title>
		By: SolarFlame		</title>
		<link>https://www.dragonflydigest.com/2017/08/07/new-mechanism-kcollect/comment-page-1/#comment-486675</link>

		<dc:creator><![CDATA[SolarFlame]]></dc:creator>
		<pubDate>Tue, 08 Aug 2017 14:28:21 +0000</pubDate>
		<guid isPermaLink="false">https://www.dragonflydigest.com/?p=20061#comment-486675</guid>

					<description><![CDATA[@Pedro

DTrace probes aren&#039;t free…

http://dtrace.org/blogs/brendan/2011/02/18/dtrace-pid-provider-overhead/]]></description>
			<content:encoded><![CDATA[<p>@Pedro</p>
<p>DTrace probes aren&#8217;t free…</p>
<p><a href="http://dtrace.org/blogs/brendan/2011/02/18/dtrace-pid-provider-overhead/" rel="nofollow ugc">http://dtrace.org/blogs/brendan/2011/02/18/dtrace-pid-provider-overhead/</a></p>
]]></content:encoded>
		
			</item>
		<item>
		<title>
		By: Pedro		</title>
		<link>https://www.dragonflydigest.com/2017/08/07/new-mechanism-kcollect/comment-page-1/#comment-486674</link>

		<dc:creator><![CDATA[Pedro]]></dc:creator>
		<pubDate>Tue, 08 Aug 2017 14:17:41 +0000</pubDate>
		<guid isPermaLink="false">https://www.dragonflydigest.com/?p=20061#comment-486674</guid>

					<description><![CDATA[@SolarFlame

Dtrace is also lockless. Yes, you have to collect the data but you can do the same with much better detail and no performance impact.]]></description>
			<content:encoded><![CDATA[<p>@SolarFlame</p>
<p>Dtrace is also lockless. Yes, you have to collect the data but you can do the same with much better detail and no performance impact.</p>
]]></content:encoded>
		
			</item>
		<item>
		<title>
		By: mer		</title>
		<link>https://www.dragonflydigest.com/2017/08/07/new-mechanism-kcollect/comment-page-1/#comment-486673</link>

		<dc:creator><![CDATA[mer]]></dc:creator>
		<pubDate>Tue, 08 Aug 2017 14:06:50 +0000</pubDate>
		<guid isPermaLink="false">https://www.dragonflydigest.com/?p=20061#comment-486673</guid>

					<description><![CDATA[sysctl&#039;s may be already collecting data, but it sounds like kcollect is a &quot;periodic snapshot&quot; of some sysctls, aggregated and presented via a plottable interface.  Basically, saving you the trouble of writing a userland that periodically polls a bunch of sysctls and doing the same thing.  A quick glance at commit messages at least one mentioned &quot;lockless&quot;, which not all the sysctl interfaces are.
DTrace is good, but again, requires you to do something in userland to collect and present the data.]]></description>
			<content:encoded><![CDATA[<p>sysctl&#8217;s may be already collecting data, but it sounds like kcollect is a &#8220;periodic snapshot&#8221; of some sysctls, aggregated and presented via a plottable interface.  Basically, saving you the trouble of writing a userland that periodically polls a bunch of sysctls and doing the same thing.  A quick glance at commit messages at least one mentioned &#8220;lockless&#8221;, which not all the sysctl interfaces are.<br />
DTrace is good, but again, requires you to do something in userland to collect and present the data.</p>
]]></content:encoded>
		
			</item>
		<item>
		<title>
		By: SolarFlame		</title>
		<link>https://www.dragonflydigest.com/2017/08/07/new-mechanism-kcollect/comment-page-1/#comment-486672</link>

		<dc:creator><![CDATA[SolarFlame]]></dc:creator>
		<pubDate>Tue, 08 Aug 2017 10:40:12 +0000</pubDate>
		<guid isPermaLink="false">https://www.dragonflydigest.com/?p=20061#comment-486672</guid>

					<description><![CDATA[@Pedro

those probes only fire when you&#039;re tracing for them, kcollect runs all the time.

I think it&#039;s better to compare it with Solaris&#039; kstats. OTOH DFly is already collecting stats via sysctl&#039;s, so why another mechanism like kcollect?]]></description>
			<content:encoded><![CDATA[<p>@Pedro</p>
<p>those probes only fire when you&#8217;re tracing for them, kcollect runs all the time.</p>
<p>I think it&#8217;s better to compare it with Solaris&#8217; kstats. OTOH DFly is already collecting stats via sysctl&#8217;s, so why another mechanism like kcollect?</p>
]]></content:encoded>
		
			</item>
		<item>
		<title>
		By: Pedro		</title>
		<link>https://www.dragonflydigest.com/2017/08/07/new-mechanism-kcollect/comment-page-1/#comment-486671</link>

		<dc:creator><![CDATA[Pedro]]></dc:creator>
		<pubDate>Tue, 08 Aug 2017 05:47:42 +0000</pubDate>
		<guid isPermaLink="false">https://www.dragonflydigest.com/?p=20061#comment-486671</guid>

					<description><![CDATA[DTrace has specific probes and is more fine-grained: it lets you know exactly what the kernel is doing. More importantly, it carries its own language and the interface is already popular in Mac, Solaris, FreeBSD, and to some extent linux and NetBSD. 

Brendan Gregg is famous for his approach measuring performance and his flamegraphs.]]></description>
			<content:encoded><![CDATA[<p>DTrace has specific probes and is more fine-grained: it lets you know exactly what the kernel is doing. More importantly, it carries its own language and the interface is already popular in Mac, Solaris, FreeBSD, and to some extent linux and NetBSD. </p>
<p>Brendan Gregg is famous for his approach measuring performance and his flamegraphs.</p>
]]></content:encoded>
		
			</item>
		<item>
		<title>
		By: Justin Sherrill		</title>
		<link>https://www.dragonflydigest.com/2017/08/07/new-mechanism-kcollect/comment-page-1/#comment-486670</link>

		<dc:creator><![CDATA[Justin Sherrill]]></dc:creator>
		<pubDate>Tue, 08 Aug 2017 03:46:51 +0000</pubDate>
		<guid isPermaLink="false">https://www.dragonflydigest.com/?p=20061#comment-486670</guid>

					<description><![CDATA[Does DTrace do the same thing?]]></description>
			<content:encoded><![CDATA[<p>Does DTrace do the same thing?</p>
]]></content:encoded>
		
			</item>
		<item>
		<title>
		By: Anonymous		</title>
		<link>https://www.dragonflydigest.com/2017/08/07/new-mechanism-kcollect/comment-page-1/#comment-486669</link>

		<dc:creator><![CDATA[Anonymous]]></dc:creator>
		<pubDate>Tue, 08 Aug 2017 03:21:53 +0000</pubDate>
		<guid isPermaLink="false">https://www.dragonflydigest.com/?p=20061#comment-486669</guid>

					<description><![CDATA[Pedro. Can you explain more. I&#039;m not following.]]></description>
			<content:encoded><![CDATA[<p>Pedro. Can you explain more. I&#8217;m not following.</p>
]]></content:encoded>
		
			</item>
		<item>
		<title>
		By: Pedro		</title>
		<link>https://www.dragonflydigest.com/2017/08/07/new-mechanism-kcollect/comment-page-1/#comment-486668</link>

		<dc:creator><![CDATA[Pedro]]></dc:creator>
		<pubDate>Tue, 08 Aug 2017 03:10:16 +0000</pubDate>
		<guid isPermaLink="false">https://www.dragonflydigest.com/?p=20061#comment-486668</guid>

					<description><![CDATA[Sounds misguided ...  ignorance of the DTrace way.]]></description>
			<content:encoded><![CDATA[<p>Sounds misguided &#8230;  ignorance of the DTrace way.</p>
]]></content:encoded>
		
			</item>
		<item>
		<title>
		By: New mechanism: kcollect &#8211; FreshBSD.com		</title>
		<link>https://www.dragonflydigest.com/2017/08/07/new-mechanism-kcollect/comment-page-1/#comment-486667</link>

		<dc:creator><![CDATA[New mechanism: kcollect &#8211; FreshBSD.com]]></dc:creator>
		<pubDate>Tue, 08 Aug 2017 01:28:00 +0000</pubDate>
		<guid isPermaLink="false">https://www.dragonflydigest.com/?p=20061#comment-486667</guid>

					<description><![CDATA[[&#8230;] Go to Source Author: Justin Sherrill [&#8230;]]]></description>
			<content:encoded><![CDATA[<p>[&#8230;] Go to Source Author: Justin Sherrill [&#8230;]</p>
]]></content:encoded>
		
			</item>
		<item>
		<title>
		By: Anonymous		</title>
		<link>https://www.dragonflydigest.com/2017/08/07/new-mechanism-kcollect/comment-page-1/#comment-486666</link>

		<dc:creator><![CDATA[Anonymous]]></dc:creator>
		<pubDate>Tue, 08 Aug 2017 00:51:36 +0000</pubDate>
		<guid isPermaLink="false">https://www.dragonflydigest.com/?p=20061#comment-486666</guid>

					<description><![CDATA[I wonder why Matt thought it was necessary to add this functionality.]]></description>
			<content:encoded><![CDATA[<p>I wonder why Matt thought it was necessary to add this functionality.</p>
]]></content:encoded>
		
			</item>
	</channel>
</rss>
