Date: Wed, 02 May 2007 11:37:08 -0500 From: Eric Anderson <anderson@freebsd.org> To: Ivan Voras <ivoras@fer.hr> Cc: freebsd-current@freebsd.org Subject: Re: gjournal + gmirror question Message-ID: <4638BE34.3070709@freebsd.org> In-Reply-To: <f1a8bn$l9m$1@sea.gmane.org> References: <200705021453.20372.h.schmalzbauer@omnisec.de> <f1a3r8$2gg$1@sea.gmane.org> <46389DED.5060509@freebsd.org> <200705021631.02167.h.schmalzbauer@omnisec.de> <f1a8bn$l9m$1@sea.gmane.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On 05/02/07 09:48, Ivan Voras wrote: > Harald Schmalzbauer wrote: > >> Because it's a large fs with standard newfs parameters, so many inodes to >> check... >> Or does fsck recognize that it's a gjournaled UFS and skip unnessecarry >> checks? >> That's what my initinal interest was about. > > Judging from > http://www.freebsd.org/cgi/cvsweb.cgi/src/sbin/fsck_ffs/gjournal.c you > might be right, but I don't see it documented in man pages. IIRC Pawel > did at one time explain that fsck will only delete stale/unreferenced > inodes and so be quicker than usual. > > If you have the time for benchmarks, they would be appreciated :) > Well, there would be less of a point to using gjournaling if you still had to fsck the whole thing, right? With gjournaling, you can just run an 'fsck -p' on it, which takes less than 1 second on my 10Tb fs. So, no need to benchmark. Eric
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4638BE34.3070709>