Skip site navigation (1)Skip section navigation (2)
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>