Date: Mon, 11 Nov 1996 19:34:13 -0500 (EST) From: Dan Janowski <danj@3skel.com> To: freebsd-current@FreeBSD.ORG, roberto@keltia.freenix.fr Subject: Re: ufs is too slow? Message-ID: <199611120034.TAA15191@fnur.3skel.com>
next in thread | raw e-mail | index | archive | help
From: roberto@keltia.freenix.fr (Ollivier Robert) > According to Dan Janowski: > > is not xfs, they are both better performers than ufs/ffs (which > > Xfs is pretty fragile from what I've heard. SGI's XFS is actually pretty robust. We run 16 and 32GB file systems with it, and the performance is good. It can handle an FS of some-number Terabytes. It is really a journal filesystem, not a log filesystem. More like the way that relational databases work, commmiting of changes are done completely or not at all. There is no fsck (ever see an fsck run on multiple 32GB filesystems? It is really ugly), but some kind of consistancy check is done, but only doing it on the journal is necessary. > > are both REALLY old, I think ufs dates from the 50's and ffs > > from the 70's). > > They appeared in 1983 in a paper that you'll find in > /usr/src/share/doc/papers/smm/05.fastfs. Although the Berkeley FFS is a significant step up from that which preceded it, the FFS is largely of the same basis. Log or journal filesystems don't have the same kind of meta-data or data access. LFS was compared to more of a garbage-collection mechanism, and XFS as above. Dan
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199611120034.TAA15191>