LWN: Comments on "btrfs and NILFS" https://lwn.net/Articles/238923/ This is a special feed containing comments posted to the individual LWN article titled "btrfs and NILFS". en-us Sun, 26 Oct 2025 21:30:52 +0000 Sun, 26 Oct 2025 21:30:52 +0000 https://www.rssboard.org/rss-specification lwn@lwn.net NILFS looks quite good https://lwn.net/Articles/394368/ https://lwn.net/Articles/394368/ nilfsguy <div class="FormattedComment"> I just used NILFS (it's actually NILFS2) as filesystem for my USB-stick and it looks quite good - it's fast and didn't have problems (yet) with file corruption (opposed to BtrFS). I am using the kernel 2.6.34.<br> Here the measurements I did and some tips to use it as rootFS:<br> <a rel="nofollow" href="http://www.blah-blah.ch/Mra/Nilfs2performance">http://www.blah-blah.ch/Mra/Nilfs2performance</a><br> <p> </div> Wed, 30 Jun 2010 20:01:52 +0000 btrfs sounds mightly cool. https://lwn.net/Articles/239659/ https://lwn.net/Articles/239659/ dlang how much of this is a limitation of the technology (like the inability to shrink XFS) and how much is just a need for better userspace tools (like easily being able to resize extX)<br> <p> don't mix one with the other.<br> Sun, 24 Jun 2007 22:16:24 +0000 btrfs and NILFS https://lwn.net/Articles/239654/ https://lwn.net/Articles/239654/ k8to Of course, a codebase is not just a set of features. Maybe some NIH is going on here, but maybe not.<br> Sun, 24 Jun 2007 21:22:15 +0000 btrfs sounds mightly cool. https://lwn.net/Articles/239533/ https://lwn.net/Articles/239533/ Tomasu The part I like about ZFS is the way you can dynamically allocate <br> physical volume space to any logical volume (aka: filesystem) at any <br> time.<br> <p> Nothing else does that as far as I know. All you get is LVM2, EVMS, <br> or "mdraid" none of which can dynamically resize the volume and <br> underlying filesystem on the fly, and EASILY. resizing an ext partition <br> is imo, too hard, and you _can't_ shrink an XFS filesyste. Function isn't <br> supported.<br> Sat, 23 Jun 2007 06:26:31 +0000 btrfs sounds mightly cool. https://lwn.net/Articles/239406/ https://lwn.net/Articles/239406/ aglet I wondered about the ZFS "license fuss" -- there's a good precis here: <a href="http://kerneltrap.org/node/8066">http://kerneltrap.org/node/8066</a><br> Fri, 22 Jun 2007 13:13:57 +0000 btrfs and NILFS https://lwn.net/Articles/239223/ https://lwn.net/Articles/239223/ Tet I admit I haven't looked into it in any depth, but from skimming the btrfs feature list, it looks like they're mostly just reinventing vxfs, partial support for which is already in the mainline kernel.<br> Thu, 21 Jun 2007 13:47:27 +0000 btrfs and NILFS https://lwn.net/Articles/239212/ https://lwn.net/Articles/239212/ i3839 There's also Chunkfs, though that seems more like a research project than a real filesystem at the moment.<br> <p> Thu, 21 Jun 2007 13:23:09 +0000 btrfs sounds mightly cool. https://lwn.net/Articles/239203/ https://lwn.net/Articles/239203/ dion Wow, it certainly sounds like btrfs got most of the features I've been wanting from ZFS (checksumming &amp; snapshots), but with less of the license fuss.<br> <p> All that's needed to gain coolness parity with ZFS is something like raid-z, unfortunately I don't see how that can be done without implementing it in the fs itself.<br> <p> <p> <p> Thu, 21 Jun 2007 12:16:45 +0000