Date: Fri, 27 Aug 2004 11:33:52 +0200 From: Stijn Hoop <stijn@win.tue.nl> To: freebsd-questions@freebsd.org Subject: Re: vinum rebuildparity, when? Message-ID: <20040827093352.GB965@pcwin002.win.tue.nl> In-Reply-To: <86pt5f1qdm.fsf@borg.borderworlds.dk> References: <20040825083123.GD17106@pcwin002.win.tue.nl> <86pt5f1qdm.fsf@borg.borderworlds.dk>
next in thread | previous in thread | raw e-mail | index | archive | help
--UlVJffcvxoiEqYs2 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable Hi, thanks for your response, I didn't notice it at first because it only went to the mailing list :) On Wed, Aug 25, 2004 at 12:08:53PM +0200, Christian Laursen wrote: > Stijn Hoop <stijn@win.tue.nl> writes: > > I was wondering about the vinum 'rebuildparity' command, especially the > > times when one needs to use this. >=20 > I run rebuildparity if checkparity finds any errors after unclean shutdow= ns. OK, that's what was adviced. > > The problem is that I can't find anything in the vinum docs about this > > command other than it's purpose. What I don't understand is the differe= nce > > between reviving a disk in a RAID-5 plex, and rebuilding the parity. >=20 > When reviving a disk the data on that disk is calculated from the data an= d the > parity on the other disks. Yes, but the parity should be recalculated at the same time, right? > I think rebuildparity only reads data and writes the parity calculated > from that but for all disks. OK, that would seem logical. > > When I start a degraded disk it starts to revive -- which led me to bel= ieve > > that vinum was also recalculating the parity. Evidently it wasn't. > > I'm therefore now updating my procedures to always run > > 'checkparity -v <plex>' after a disk crash. >=20 > That shouldn't be neccesary. Well, it appears to be. checkparity found some errors in the parity after a single disk crash & rebuild (ie degraded mode -> start subdisk -> revive process complete). If this is not the expected behaviour it means something about the controll= er has blown; we do get lots of unexpected read/write errors which always turn out to be false alarms upon further inspection. Maybe a controller has turned bad :( FWIW, the rebuildparity helped, the parity is now again correct. --Stijn --=20 "Harry, I'm going to let you in on a little secret. Every day, once a day, give yourself a present. Don't plan it, don't wait for it, just let it happen. Could be a new shirt at the men's store, a catnap in your office chair, or... two cups of good, hot, black coffee. Like this." -- Special Agent Dale Cooper, "Twin Peaks" --UlVJffcvxoiEqYs2 Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (FreeBSD) iD8DBQFBLwAAY3r/tLQmfWcRAiDxAJwJN6+dI/uqIPXjletvuj4gGr09ewCgta5c Ox87rJRe1bWGaJnvxDlmz6U= =xoIF -----END PGP SIGNATURE----- --UlVJffcvxoiEqYs2--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040827093352.GB965>