<?xml version="1.0" encoding="utf-8"?>
<rss xmlns:atom="http://www.w3.org/2005/Atom" version="2.0"><channel><title>Disqus - Latest Comments for madssj</title><link>http://disqus.com/by/madssj/</link><description></description><atom:link href="http://disqus.com/madssj/comments.rss" rel="self"></atom:link><language>en</language><lastBuildDate>Wed, 03 Sep 2014 04:52:54 -0000</lastBuildDate><item><title>Re: 71 TiB DIY NAS based on ZFS on Linux</title><link>http://louwrentius.com/71-tib-diy-nas-based-on-zfs-on-linux.html#comment-1571624495</link><description>&lt;p&gt;I'm just saying, that there's a huge difference between the two. As we've already covered in another comment, the difference between using and not using the buffer cache on a 16 GB memory system, with a 1 TiB file will be negligible.&lt;/p&gt;&lt;p&gt;conv=sync won't affect the speed of what you're writing, but the output of what you're writing.&lt;/p&gt;&lt;p&gt;For any size file that'll fit into the buffer cache, the results will the difference between writing to memory and writing to disk.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">madssj</dc:creator><pubDate>Wed, 03 Sep 2014 04:52:54 -0000</pubDate></item><item><title>Re: 71 TiB DIY NAS based on ZFS on Linux</title><link>http://louwrentius.com/71-tib-diy-nas-based-on-zfs-on-linux.html#comment-1571611087</link><description>&lt;p&gt;Agreed.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">madssj</dc:creator><pubDate>Wed, 03 Sep 2014 04:29:51 -0000</pubDate></item><item><title>Re: 71 TiB DIY NAS based on ZFS on Linux</title><link>http://louwrentius.com/71-tib-diy-nas-based-on-zfs-on-linux.html#comment-1571573594</link><description>&lt;p&gt;The results would be ~1.5% wrong without it.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">madssj</dc:creator><pubDate>Wed, 03 Sep 2014 03:26:37 -0000</pubDate></item><item><title>Re: 71 TiB DIY NAS based on ZFS on Linux</title><link>http://louwrentius.com/71-tib-diy-nas-based-on-zfs-on-linux.html#comment-1569970086</link><description>&lt;p&gt;I'm actually using ZoL/Debian on another system, and it's been running very smoothly.&lt;/p&gt;&lt;p&gt;As for testing with dd, there's a huge difference between conv=sync, and conv=fsync. One pads the input, the other calls fsync to sync data and metadata to disk before exiting.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">madssj</dc:creator><pubDate>Tue, 02 Sep 2014 04:51:24 -0000</pubDate></item><item><title>Re: 71 TiB DIY NAS based on ZFS on Linux</title><link>http://louwrentius.com/71-tib-diy-nas-based-on-zfs-on-linux.html#comment-1568870439</link><description>&lt;p&gt;ZoL is *the* way to add ZFS to any Linux distribution.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">madssj</dc:creator><pubDate>Mon, 01 Sep 2014 10:45:05 -0000</pubDate></item><item><title>Re: 71 TiB DIY NAS based on ZFS on Linux</title><link>http://louwrentius.com/71-tib-diy-nas-based-on-zfs-on-linux.html#comment-1568655075</link><description>&lt;p&gt;I did a build based on the same chase, hba's and motherboard a year ago. It's running FreeBSD and with the hba's cross flashed to IT mode. It's been running very stable, although there's been a few disk casualties.&lt;/p&gt;&lt;p&gt;A few notes:&lt;/p&gt;&lt;p&gt;Remember to add conv=fsync to your dd run, otherwise you'll end up testing the speed of your buffer cache.&lt;/p&gt;&lt;p&gt;With that many disks in your pools, you'd probably want to use uuid-based assignment (/dev/disk/by-uuid) over location based (/dev/sdX).&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">madssj</dc:creator><pubDate>Mon, 01 Sep 2014 07:15:33 -0000</pubDate></item><item><title>Re: Extracting a single database from a pg_dumpall PostgreSQL dump</title><link>https://www.madssj.com/blog/2010/04/09/extracting-a-single-database-from-a-pg_dumpall-postgresql-dump/#comment-1165482417</link><description>&lt;p&gt;It worked wonderfully in 2010, the dump format might have changed. I'll add a link in the top for your updated gist - thanks!&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">madssj</dc:creator><pubDate>Mon, 16 Dec 2013 05:22:14 -0000</pubDate></item><item><title>Re: Extracting a single database from a pg_dumpall PostgreSQL dump</title><link>https://www.madssj.com/blog/2010/04/09/extracting-a-single-database-from-a-pg_dumpall-postgresql-dump/#comment-918429960</link><description>&lt;p&gt;Then it can't find a line that says "\connect dbname" in your dump file.&lt;/p&gt;&lt;p&gt;Feel free to tinker with the script.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">madssj</dc:creator><pubDate>Tue, 04 Jun 2013 04:23:42 -0000</pubDate></item><item><title>Re: And a fun weekend was had by all...</title><link>http://blog.reddit.com/2010/03/and-fun-weekend-was-had-by-all.html#comment-37505044</link><description>&lt;p&gt;&amp;gt; 5 new instances to run memcached. This allowed us to expand the size of the memcache in front of memcachedb to 6GB (up from 2GB).&lt;/p&gt;&lt;p&gt;Wait, what. Are you using m1.small instances for your caches or did you actually gain a couple of hundred megabytes of memory from those 5 instances, and from that were able to use 6GB to stick in front of memcachedb?&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">madssj</dc:creator><pubDate>Mon, 01 Mar 2010 16:08:34 -0000</pubDate></item><item><title>Re: Securing an HTTPS server</title><link>http://www.daemonology.net/blog/2009-09-28-securing-https.html#comment-17829823</link><description>&lt;p&gt;Remind me to get you a tinfoil hat that says "Thank you".&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">madssj</dc:creator><pubDate>Wed, 30 Sep 2009 03:28:44 -0000</pubDate></item><item><title>Re: Tarsnap mailing lists</title><link>http://www.daemonology.net/blog/2009-07-18-tarsnap-mailing-lists.html#comment-12997017</link><description>&lt;p&gt;Oh, I've been looking forward to this day!&lt;/p&gt;&lt;p&gt;Thanks a bunch Colin.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">madssj</dc:creator><pubDate>Tue, 21 Jul 2009 10:22:51 -0000</pubDate></item><item><title>Re: A call for schwag</title><link>http://www.daemonology.net/blog/2009-07-14-a-call-for-schwag.html#comment-12679618</link><description>&lt;p&gt;Not to say that this is not a good idea, but it kind of raises the question about who to send stuff or money.&lt;/p&gt;&lt;p&gt;For example, I use Debian a lot, now say I want to thank someone for that. Who sould I thank? I could thank Linus for writing Linux in the first place, or Allan Cox for contributing a whole lot of code to Linux and the community as a whole, or I could thank some random Debian developer. The project itself could probably also use some money (don't know about Debian actually, but FreeBSD has the FreeBSD Foundation - Debian probably has something similar).&lt;/p&gt;&lt;p&gt;Perhaps this would be easier if the larger OSS projects had some kind of "show your appriciation" page with wishlists and such. Spending an hour to track down some random guy will make a lot of people not do it.&lt;/p&gt;&lt;p&gt;Now, I'm off to find the author of ionice and the author of the kernel code that makes it possible for ionice to work in Linux.&lt;/p&gt;&lt;p&gt;Thanks for the enlightment.&lt;/p&gt;&lt;p&gt;Also:&lt;/p&gt;&lt;p&gt;Dear Mr. Developer Colin.&lt;/p&gt;&lt;p&gt;Thanks for all your work on FreeBSD - if you ever find yourself in Copenhagen, I'll happliy supply you with beer and a fistfull of stickers that say "Konstellation".&lt;/p&gt;&lt;p&gt;Cheers.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">madssj</dc:creator><pubDate>Wed, 15 Jul 2009 05:11:31 -0000</pubDate></item><item><title>Re: bitquabit - Zombie Operating Systems and ASP.NET MVC</title><link>http://blog.bitquabit.com/2009/06/12/zombie-operating-systems-and-aspnet-mvc/#comment-10821399</link><description>&lt;p&gt;Two words. Backwards compatibility.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">madssj</dc:creator><pubDate>Fri, 12 Jun 2009 19:54:34 -0000</pubDate></item></channel></rss>