Date: Wed, 27 Apr 2011 22:55:11 +0400 From: Lev Serebryakov <lev@FreeBSD.org> To: Miroslav Lachman <000.fbsd@quip.cz> Cc: freebsd-fs@freebsd.org Subject: Re: Why not just name the cam-ata devices the same as the old, names? Message-ID: <669365808.20110427225511@serebryakov.spb.ru> In-Reply-To: <4DB81F90.6020108@quip.cz> References: <20110427101728.49C801065709@hub.freebsd.org> <4DB7F61C.8060003@digsys.bg> <4DB81F90.6020108@quip.cz>
next in thread | previous in thread | raw e-mail | index | archive | help
Hello, Miroslav. You wrote 27 =E0=EF=F0=E5=EB=FF 2011 =E3., 17:52:16: >> What should be done, ideally before the 9.0 release is to find some sane >> resolution method of what happens when you happen to have two (for >> example) 'root' UFS labels during boot/mount time. > ...and that's the problem with labels in case of gmirror. > When you create gm0 of ada0 and ada1, then sometime in the future ada0 > timedout and is dropped by gmirror, on next reboot you will have same=20 > partitions available on device /dev/ada0 and /dev/mirror/gm0 so there=20 > will be two devices promoting same labels! I don't know the order of=20 > tasting devices, but there is a chance that system will boot (mount root > by label) from broken ada0 instead of gm0. Even without broken mirror, in case of perfectly valid reboot, glabel could pick-up label first (before gmirror) from one of (both are non-broken) mirror components, and after that gmirror builds itself from one true component (ada0, for example) and one label-provided (instead of ada1) :( So, UFS labels CAN NOT BE MIXED with gmirror... --=20 // Black Lion AKA Lev Serebryakov <lev@FreeBSD.org>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?669365808.20110427225511>