raid6 problem.
Robert Heller
heller at deepsoft.com
Sat Aug 9 11:24:21 UTC 2025
At Fri, 8 Aug 2025 23:30:36 -0400 "Ubuntu user technical support,? not for general discussions" <ubuntu-users at lists.ubuntu.com> wrote:
>
> On 8/8/25 20:11, Robert Heller wrote:
> > Ok, if the volume group is vg0 and the volume name is myraid, then there
> > should be a device
> >
> > /dev/vg0/myraid,
> > ^ ^--- logical volume
> > |----- Volume group
> This turned out to be a symlink to /dev/dm-10.
Yes that is normal. *Don't* hardwire /dev/dm-<mumblw> as this can change in
unpredicable ways (eg boot order changes). Always use /dev/<vgroup>/<lvolume>
OR /deb/mapper/<vgroup>-<lvolume> to refer to a LVM volume device.
> >
> > OR
> >
> > /dev/mapper/vg0-myraid
> >
> > is another path to the same place.
> Except: ls /dev/mapper (uts a big raid, 5 ea 4TiB drives)
> control   vg0-myraid_rimage_0 vg0-myraid_rimage_2
> vg0-myraid_rimage_4Â vg0-myraid_rmeta_1Â vg0-myraid_rmeta_3
> vg0-myraid vg0-myraid_rimage_1 vg0-myraid_rimage_3 vg0-myraid_rmeta_0Â
> Â vg0-myraid_rmeta_2Â vg0-myraid_rmeta_4
>
> >
> > (There is another option under /dev/disk/by-id as well.)
> >
> > so assuming you did something like (creating a Ext4 file system)
> >
> > sudo /sbin/mkfs.ext4 /dev/vg0/myraid
> I switched to mkfs.gfs2 -p lock_nolock before that would proceed where
> it made over 43xxx entries, culminating in:
> root at amanda:~# mkfs.gfs2 -p lock_nolock /dev/vg0/myraid
> /dev/vg0/myraid is a symbolic link to /dev/dm-10
> This will destroy any data on /dev/dm-10
> Are you sure you want to proceed? [y/n] y
> Discarding device contents (may take a while on large devices): Done
> Adding journals: Done
> Building resource groups: Done
> Creating quota file: Done
> Writing superblock and syncing: Done
> Device:Â Â Â Â Â Â Â Â Â Â /dev/vg0/myraid
> Block size:Â Â Â Â Â Â Â Â 4096
> Device size:Â Â Â Â Â Â Â Â 10752.00 GB (2818572288 blocks)
> Filesystem size:Â Â Â Â Â Â 10752.00 GB (2818572253 blocks)
> Journals:Â Â Â Â Â Â Â Â Â 1
> Journal size:Â Â Â Â Â Â Â 1024MB
> Resource groups:Â Â Â Â Â Â 42974
> Locking protocol:Â Â Â Â Â "lock_nolock"
> Lock table:Â Â Â Â Â Â Â Â ""
> UUID:Â Â Â Â Â Â Â Â Â Â Â 39a76260-c304-4c90-a6a8-ec89407fd2b3
> > sudo mkdir /myraid
> ok.
> > /etc/fstab could then have
> >
> > /dev/vg0/myraid /myraid ext4 defaults 1 2
> /dev/vg0/myraid gfs2 defaults 1 2
> >
> > Even better (hardwiring /dev/... in /etc/fstab is to be avoided whenever
> > possible -- using LABEL=r UUID= is prefered to avoid boot time "crashes" due
> > to hardware changes):
> >
> > sudo /sbin/mkfs.ext4 -L MyRaid /dev/vg0/myraid
> > sudo mkdir /myraid
> >
> > then in /etc/fstab:
> >
> > LABEL=Raid /myraid ext4 defaults 1 2
> I put LABEL=Raid /myraid gfs2 defaults 1 2
>
> trying to do this but there's no LABEL. I forgot to add it to mkfs.gfs2,
> in fact it doesn't have the -L name option. the doc says tune2fs can do
> that, but how isn't clear. Hints plz?
*I've* only every used Ext file systems, so I don't know what options gfs2
has,,, You should check the man pages...
>
> Thank you Robert Heller
> > At Fri, 8 Aug 2025 17:29:03 -0400 "Ubuntu user technical support,? not for general discussions" <ubuntu-users at lists.ubuntu.com> wrote:
> >
> >> Got Q for admin types, I trying to run amanda on a bpi-m5.
> >>
> >> I have created this raid6 on arm64 but can't seem to concoct a mount
> >> recipe for /etc/fstab.
> >>
> >> I have this:
> >> root at amanda:~# lvs
> >> ÃâàLVÃâàÃâàÃâàVGÃâàAttrÃâàÃâàÃâàÃâàLSizeÃâàPool Origin Data%ÃâàMeta%ÃâàMove Log
> >> Cpy%Sync Convert
> >> Ãâàmyraid vg0 rwi-a-r--- 10.50t 2.16
> >>
> >> I have this seemingly done, lvm and mdadm are all installed.
> >>
> >> Can anyone help?
> >>
> >> Thank you.
> >>
> >> Cheers, Gene Heskett, CET.
> >
>
> Cheers, Gene Heskett, CET.
--
Robert Heller -- Cell: 413-658-7953 GV: 978-633-5364
Deepwoods Software -- Custom Software Services
http://www.deepsoft.com/ -- Linux Administration Services
heller at deepsoft.com -- Webhosting Services
More information about the ubuntu-users
mailing list