Date: Fri, 07 Apr 2006 09:49:50 +0200 From: "=?iso-8859-15?Q?Jos=E9?= M. =?iso-8859-15?Q?Fandi=F1o?=" <freebsd4@fadesa.es> To: freebsd-stable@freebsd.org Subject: Re: GEOM_RAID3: Device datos is broken, too few valid components Message-ID: <4436199E.3E019AD4@fadesa.es> References: <4433DB2B.7753D2C3@fadesa.es> <20060406153115.GC5578@garage.freebsd.pl> <44354FB1.6163DF7A@fadesa.es>
next in thread | previous in thread | raw e-mail | index | archive | help
"Jos=E9 M. Fandi=F1o" wrote:
> =
> Pawel Jakub Dawidek wrote:
> > +> Unfortunately, the metadata structure of my data partition (a geo=
m raid3
> > +> array with tree components ) seems to be corrupted by this hard =
lock,
> > +> the following message is scrolled constantly on the screen:
> > +>
> > +> GEOM_RAID3: Device datos created (id=3D3217021940).
> > +> GEOM_RAID3: Device datos: provider ad6s2 detected.
> > +> GEOM_RAID3: Device datos: provider ad5s2 detected.
> > +> GEOM_RAID3: Device datos: provider ad4s2 detected.
> > +> GEOM_RAID3: Component ad6s2 (device datos) broken, skipping.
> > +> GEOM_RAID3: Component ad4s2 (device datos) broken, skipping.
> > +> GEOM_RAID3: Device datos is broken, too few valid components.
> > +> GEOM_RAID3: Device datos destroyed.
> > +>
> > +> Checking the search engine results it isn't a very usual problem, =
the advice
> > +> in the returned hits is rerunning "graid label -h datos ad4s2 ad5s=
2 ad6s2",
> > +> but before of erasing all my data I would like to ask to list memb=
ers.
> > +>
> > +> How dangerous is running the mentioned command in this context?
> >
> > You should be safe as long as the order of slices you give here is th=
e
> > same as it was when device was initially labeled.
> =
> I don't remember the exact order (my history file is only 2000 lines
> long :) however I'm going to duplicate two disks, because they are
> the minimum necessary to reconstruct the raid3, and I will do some test=
s
> over them, so I can maintain intact the original disks.
Pawel,
there is a 'dump' parameter for graid3 (it isn't mentioned in the manual=
page but you can see it if you run graid3 in help mode)
would it be sufficient follow the order indicated by the 'no' field?
something like:
graid3 label -h datos ad5s2 ad6s2 ad4s2
# graid3 dump ad4s2
Metadata on ad4s2:
magic: GEOM::RAID3
version: 4
name: datos
id: 3217021940
no: 2
all: 3
genid: 2
syncid: 24
mediasize: 311491352576
sectorsize: 1024
syncoffset: 0
mflags: ROUND-ROBIN
dflags: NONE
hcprovider: =
provsize: 155745676800
MD5 hash: 3592c58adc68a476ce5253d2629f5c77
# graid3 dump ad4s=08 =08=08 =085s2
Metadata on ad5s2:
magic: GEOM::RAID3
version: 4
name: datos
id: 3217021940
no: 0
all: 3
genid: 0
syncid: 24
mediasize: 311491352576
sectorsize: 1024
syncoffset: 0
mflags: ROUND-ROBIN
dflags: NONE
hcprovider: =
provsize: 155745676800
MD5 hash: 5dc9f22fb87c2d14c80cbd6b84dc4496
# graid3 dump ad6s2
Metadata on ad6s2:
magic: GEOM::RAID3
version: 4
name: datos
id: 3217021940
no: 1
all: 3
genid: 2
syncid: 24
mediasize: 311491352576
sectorsize: 1024
syncoffset: 0
mflags: ROUND-ROBIN
dflags: DIRTY
hcprovider: =
provsize: 155745676800
MD5 hash: 75f9c48df82fcbd75b1b5d6abc4c763c
-- =
-----BEGIN GEEK CODE BLOCK-----
Version: 3.1
GCS/IT d- s+:+() a32 C+++ UBL+++$ P+ L+++ E--- W++ N+ o++ K- w---
O+ M+ V- PS+ PE+ Y++ PGP+>+++ t+ 5 X+$ R- tv-- b+++ DI D++>+++
G++ e- h+(++) !r !z
------END GEEK CODE BLOCK------
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4436199E.3E019AD4>