Date: Fri, 4 May 2007 22:31:26 +0300 From: deniswork@uzvik.kiev.ua To: =?ISO-8859-1?Q?Arne_W=F6rner?= <arne_woerner@yahoo.com> Cc: freebsd-geom@freebsd.org Subject: Re[2]: graid5 after-reboot problem Message-ID: <171980743.20070504223126@uzvik.kiev.ua> In-Reply-To: <386758.54024.qm@web30314.mail.mud.yahoo.com> References: <1015033466.20070504204946@uzvik.kiev.ua> <386758.54024.qm@web30314.mail.mud.yahoo.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> That is strange... > Obviously ur file system is damaged... > I do not know, how that happened... > Did u abort the device rebuild process? Yes, i make $graid5 configure -R disk this stops sync. than I make graid5 remove ... and disclable foreach disck. than i create new array, and all goes fine. I rebooted server few times, all was fine. I start coping my data, than power reboot happens... in logs i see: May 4 20:37:48 82 kernel: GEOM_RAID5: data: device created (stripesize=3D1= 31072). May 4 20:37:48 82 kernel: GEOM_RAID5: data: ad10(3): disk attached. May 4 20:37:48 82 kernel: GEOM_RAID5: data: ad8(2): disk attached. May 4 20:37:48 82 kernel: GEOM_RAID5: data: ad6(1): disk attached. May 4 20:37:48 82 kernel: GEOM_RAID5: data: ad4(0): disk attached. May 4 20:37:48 82 kernel: GEOM_RAID5: data: activated. May 4 20:37:48 82 kernel: WARNING: R/W mount of /usr/home/esc/daemon/uploa= d/dl2 denied. Filesystem is not clean - run fsck May 4 20:37:56 82 kernel: WARNING: R/W mount of /usr/home/esc/daemon/uploa= d/dl2 denied. Filesystem is not clean - run fsck May 4 20:40:04 82 last message repeated 4 times May 4 20:45:29 82 last message repeated 2 times I start scan_ffs for raid ... it take much time .. i will tell results... but i think, this will not help , coz mounted(in reas-only) data is damaged .... > If ur box can do a "graid5 remove" without crashing u could try this: > Try to do that fsck in degraded mode (in the hope that just one disk cont= ains > silly data)... U should start with removing disk#0... Then do the fsck...= But > before u put back disk#0 u should remove disk#1 in order to avoid a furth= er > rebuild process, which might damage the data further... > Maybe u could try sysutils/scan_ffs (from the ports) in order to repair t= he > disklabel? And then do a fsck again? > I mount my 3 graid5 devices every day without such problems... Even > "power-failures" (when my CPU gets too hot, the PSU shuts the box down :-= ] ) r > no real problem... > -Arne
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?171980743.20070504223126>