Comments on: Rawhide on Dell Mini 9 http://hircus.wordpress.com/2009/03/09/rawhide-on-dell-mini-9/ Michel explores computing and assorted gadgetries Mon, 07 Dec 2009 23:12:37 +0000 http://wordpress.com/ hourly 1 By: Michel http://hircus.wordpress.com/2009/03/09/rawhide-on-dell-mini-9/#comment-4800 Michel Thu, 02 Apr 2009 02:40:34 +0000 http://hircus.wordpress.com/?p=240#comment-4800 I've not done any benchmarks, but suffice to say it's *really* noticeable -- the Atom is not a very fast CPU (1.6 GHz, in-order, single-core design, with hyperthreading to ameliorate pipeline stalls), and I'm guessing Btrfs developers have been testing on machines with a bit more oomph. Last time I checked, Btrfs support in Fedora kernels is not always stable; until it is stable enough (or the Btrfs team bless certain kernel versions as being tested), there are probably other more important things to focus on. But feel free to forward this to the Btrfs mailing list! I’ve not done any benchmarks, but suffice to say it’s *really* noticeable — the Atom is not a very fast CPU (1.6 GHz, in-order, single-core design, with hyperthreading to ameliorate pipeline stalls), and I’m guessing Btrfs developers have been testing on machines with a bit more oomph.

Last time I checked, Btrfs support in Fedora kernels is not always stable; until it is stable enough (or the Btrfs team bless certain kernel versions as being tested), there are probably other more important things to focus on. But feel free to forward this to the Btrfs mailing list!

]]> By: Plouj http://hircus.wordpress.com/2009/03/09/rawhide-on-dell-mini-9/#comment-4757 Plouj Mon, 16 Mar 2009 19:30:41 +0000 http://hircus.wordpress.com/?p=240#comment-4757 Can you elaborate on the Btrfs checksumming and performance of RPM transactions? How did you determine this to be a problem? What kind of performance difference were you getting compared to ext3? Have you filed a bug about this with the Btrfs developers? It sounds like something that would impact and matter to a lot of (RPM) users. Thanks. Can you elaborate on the Btrfs checksumming and performance of RPM transactions? How did you determine this to be a problem? What kind of performance difference were you getting compared to ext3? Have you filed a bug about this with the Btrfs developers? It sounds like something that would impact and matter to a lot of (RPM) users.

Thanks.

]]>