- Bruce
On Fri, Apr 1, 2022 at 12:45 PM Wol <antlists@xxxxxxxxxxxxxxx> wrote:
>
> Hmmm... what drives are the damaged array on? There's an intact raid1
> there ...
In my script, I focused on /dev/sd[ce]1 because those partitions have
(or had) the data.
I'm guessing they're toast, but I never intentionally had a new RAID
formatted, let alone a new FS installed.
Given that the installer did stuff I did not intend for it to do, I
cannot really guarantee anything anymore.
> On 01/04/2022 19:21, Bruce Korb wrote:
>
> So sda5 has a raid1 on it ...
new disk, no data -- same with sdb5.
> and mounted as /home.
New disks, new /home hierarchy.
> > ├scsi 2:0:0:0 ATA HGST HMS5C4040AL {PL1331LAHEZZ5H}
> > │└sdc 3.64t [8:32] Partitioned (gpt)
> > │ ├sdc1 3.20t [8:33] MD raid0 (0/2) (w/ sde1) in_sync 'any:1'
> > {f624aab2-afc1-8758-5c20-d34955b9b36f}
> > │ │└md1 6.40t [9:1] MD v1.0 raid0 (2) clean, 64k Chunk, None (None)
> > None {f624aab2:-afc1-87:58-5c20-:d34955b9b36f}
> > │ │ xfs 'User' {fe716da2-b515-4fd6-8ea6-f44f48038b78}
>
> This looks promising ... dunno what on earth it thought it was doing,
> but it's telling me that on sdc1 we have a raid 0, version 1.0, with an
> xfs on it. Is there any chance your install formatted the new raid?
Chance? Sure, because it didn't do what I was expecting. It was never,
ever mounted.
During the install, I made certain that no mount point was associated with it.
Once the install was done, I did do a manual "mount /dev/md1 /mnt",
but said I needed
to run the "xfs_recover" program. I started it, but then I realized
that it was looking at
7TB of striped data, whereas it was actually 3.5TB of redundant data.
That wasn't going to work.
> Because if it did your data is probably toast, but if it didn't we might
> be home and dry.
If I can figure out how to mount it (read only), then I can see if a
new FS was installed.
If it wasn't, then I've got my data.
> Can you mount the raid? This just looks funny to me though, so make sure
> it's read only.
>
> Seeing as it made it v1.0, that means the raid superblock is at the end
> of the device and will not have done much if any damage ...
+1 !!
> It's probably a good idea to create a loopback device and mount it via
> that because it will protect the filesystem.
>
> Does any of this feel like it's right?
I don't remember back all those years ago about which file system
openSUSE decided to use as default.
I'm pretty sure it was either XFS or EXT4. I did file systems many
years ago, but that was back in the mid-80s.
IOW, I don't know how to look for the file system layout data to
figure out how to mount this beast.
I thought I could rely on the install detecting the RAID and doing the
right thing. Obviously not. :(
