2012/4/29 tsuna <tsuna...@gmail.com>: > On Fri, Apr 20, 2012 at 8:09 AM, Christian Brunner > <christ...@brunner-muc.de> wrote: >> After running ceph on XFS for some time, I decided to try btrfs again. >> Performance with the current "for-linux-min" branch and big metadata >> is much better. > > I've heard that although performance from btrfs is better at first, it > degrades over time due to metadata fragmentation, whereas XFS' > performance starts off a little worse, but remains stable even after > weeks of heavy utilization. Would be curious to hear your (or > others') feedback on that topic.
Metadata fragmentation was a big problem (for us) in the past. With the "big metatdata feature" (mkfs.btrfs -l 64k -n 64k) these problems seem to be solved. We do not use it in production yet, but my stress test didn't show any degradation. The only remaining issues I've seen are these warnings. Regards, Christian -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
