Re: RAID 6, 6 device array - all devices lost superblock

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Tried with the /dev/mapper/sdx devices.

root@superior:/mnt/backup# mdadm --create /dev/md0 --level=raid6 -n 6
/dev/mapper/sdb  /dev/mapper/sdc /dev/mapper/sdd /dev/mapper/sde
/dev/mapper/sdf /dev/mapper/sdg
mdadm: partition table exists on /dev/mapper/sdb
mdadm: partition table exists on /dev/mapper/sdc
mdadm: partition table exists on /dev/mapper/sdc but will be lost or
       meaningless after creating array
mdadm: partition table exists on /dev/mapper/sdd
mdadm: partition table exists on /dev/mapper/sdd but will be lost or
       meaningless after creating array
mdadm: partition table exists on /dev/mapper/sde
mdadm: partition table exists on /dev/mapper/sde but will be lost or
       meaningless after creating array
mdadm: partition table exists on /dev/mapper/sdf
mdadm: partition table exists on /dev/mapper/sdf but will be lost or
       meaningless after creating array
mdadm: partition table exists on /dev/mapper/sdg
mdadm: partition table exists on /dev/mapper/sdg but will be lost or
       meaningless after creating array
Continue creating array? n
mdadm: create aborted.
root@superior:/mnt/backup#

Chickened out and aborted the create.
Are those expected messages for the mess I am in?

And the victory conditions would be a mountable file system that passes a fsck?


On Mon, Aug 29, 2022 at 7:53 PM Peter Sanders <plsander@xxxxxxxxx> wrote:
>
> Couple more questions.
>
> Mdadm -create ... Do I use the /dev/sdx or /dev/mapper/sdx name for
> the overlayed device?
>
> And reset the mapping between each create attempt by doing:
> remove the loop-device/overlay association
>    dmsetup remove on all devices
> remove the overlay files
>   rm
> remove the loop back devices
>   losetup -d ...
> rebuild the loop back devices
>   mknod -m 660 ...
> build the overlay files
>   truncate -s 300G overlay-...
> reassociate the loop-devices and the overlays
>   losetup... dmsetup..
>
> and try again.
>
> (Yeah, I recognize that there is code to do this (I think) in the
> article, but my script-fu is not up to fully understanding those
> examples.)
>
> On Mon, Aug 29, 2022 at 6:58 PM Eyal Lebedinsky <fedora@xxxxxxxxxxxxxx> wrote:
> >
> >
> > On 30/08/2022 07.45, John Stoffel wrote:
> > >>>>>> "Peter" == Peter Sanders <plsander@xxxxxxxxx> writes:
> > >
> > > Peter> Phil,
> > > Peter> fstab from the working config -
> > >
> > > Peter> # <file system> <mount point>   <type>  <options>       <dump>  <pass>
> > > Peter> # / was on /dev/sda1 during installation
> > > Peter> UUID=50976432-b750-4809-80ac-3bbdd2773163 /               ext4
> > > Peter> errors=remount-ro 0       1
> > > Peter> # /home was on /dev/sda6 during installation
> > > Peter> UUID=eb93a2c4-0190-41fa-a41d-7a5966c6bc47 /home           ext4
> > > Peter> defaults        0       2
> > > Peter> # /var was on /dev/sda5 during installation
> > > Peter> UUID=d1aa6d1f-3ee9-48a8-9350-b15149f738c4 /var            ext4
> > > Peter> defaults        0       2
> > > Peter> /dev/sr0        /media/cdrom0   udf,iso9660 user,noauto     0       0
> > > Peter> /dev/sr1        /media/cdrom1   udf,iso9660 user,noauto     0       0
> > > Peter> # raid array
> > > Peter> /dev/md0    /mnt/raid6    ext4    defaults    0    2
> > >
> > > Peter> No LVM, one large EXT4 partition
> > >
> > > Peter> I have several large files ( NEF and various mpg files) I can identify
> > > Peter> and have backup copies available.
> > >
> > > Peter> I have the overlays created. 300G for each of the six drives.
> > >
> > > So that's good.  Now you have to try and figure out which order they
> > > were created in.  As the docs show, you setup the overlayfs on top of
> > > each of the six drives.
> > >
> > > Keep track by noting the drive serial numbers, since Linux can move
> > > them around and change drive letters on reboots.
> > >
> > >
> > > Then using the overlays, do an:
> > >
> > >       mdadm --create /dev/md0 --level=raid6 -n 6 /dev/sd[bcdefg]
> > >       fsck -n /dev/md0
> > >
> > > and see what you get.  If it doesn't look like a real filesystem, then
> > > you can break it down, and then modify the order you give the drive
> > > letters, like:
> > >
> > >        /dev/sd[cdefge]
> > >
> > > and rinse and repeat as it goes.  Not fun... but should hopefully fix
> > > things for you.
> > >
> > > John
> >
> > An aside, I would think the way to specify a list in a nominated order is something like
> >
> > $ echo /dev/sd{c,d,a,b}
> > /dev/sdc /dev/sdd /dev/sda /dev/sdb
> >
> > rather than
> >
> > $ echo /dev/sd[cdab]
> > /dev/sda /dev/sdb /dev/sdc /dev/sdd
> >
> > which will be in sorting order, regardless of the order of the letter.
> >
> > --
> > Eyal Lebedinsky (fedora@xxxxxxxxxxxxxx)



[Index of Archives]     [Linux RAID Wiki]     [ATA RAID]     [Linux SCSI Target Infrastructure]     [Linux Block]     [Linux IDE]     [Linux SCSI]     [Linux Hams]     [Device Mapper]     [Device Mapper Cryptographics]     [Kernel]     [Linux Admin]     [Linux Net]     [GFS]     [RPM]     [git]     [Yosemite Forum]


  Powered by Linux