I have a single boot SUSE (11.4, I believe) system with a single hard drive. I have made no changes to the system in a couple of years. When I rebooted the system a few days ago I got a grub error 17. After searching some forums, I was able to get the system started with a SUSE LiveCD. I ran an fdisk -l and findgrub:
fdisk -l
Disk /dev/sda: 122.9 GB, 122942324736 bytes
255 heads, 63 sectors/track, 14946 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x0006a426
Device Boot Start End Blocks Id System
/dev/sda1 * 1 14945 120045681 f W95 Ext’d (LBA)
/dev/sda5 1 131 1052194+ 82 Linux swap / Solaris
/dev/sda6 132 2090 15735636 83 Linux
/dev/sda7 2091 14945 103257756 83 Linux
findgrub
Find Grub Version 3.5.1 - Written for openSUSE Forums
unknown query type
reading MBR on disk /dev/sda … → Legacy GRUB found in sda MBR => sda6 0x83 (openSUSE)
reading bootsector /dev/sda1 * (Extended) …
skipping partition /dev/sda5 (swap)
reading bootsector /dev/sda6 (LINUX) … → Legacy GRUB found in /dev/sda6 => sda6 0x83 (INVALID)
reading bootsector /dev/sda7 (LINUX) …
Press <enter> to Exit findgrub…
I’m concerned that I’m getting the 0x83 (INVALID) when findgrub looks at /dev/sda6. I haven’t been able to mount /dev/sda6, I receive an error that it is the incorrect fs type or has a bad superblock. fsck on /dev/sda6 aborts because of a suspected harware problem.
Any help on being able to correct the MBR or mount this drive to copy the data to another drive would be very much appreciated.
The liveCD is version 11 also. I didn’t do the original install, and this is running software from a company from which I no longer have support, but the application is tied to the version of the OS, so there isn’t much I can do there in terms of updates.
The system is not booting by itself at all (without the LiveCD). Things were working ok until a few days ago when the grub error 17 started. I really need to either repair this hard drive so the system can find the MBR, or at least recover the files to copy to a new hard drive.
From this description it is most likely that this is a hardware problem.
So if you can run the system from the LiveCD then it should be possible to you to backup at least a good part of your data which you should do as soon as possible.
If your hard disk is in the state of inreasingly showing failures then odds are that you can not ‘repair’ it.
So first backup your data.
Thereafter you may use ‘smartctl’ as root from the command line to get to know more about the health of your hard disk.
On 2013-07-25 00:16, rdnet wrote:
>
> The liveCD is version 11 also.
That’s saying nothing. Notice that version 11.3 is very different than
version 11.4, there are no “minor” versions in openSUSE, despite what it
may appear looking at the names/versions.
Told otherwise: 11.2 is a major release, 11.3 is another major release,
11.4 is another major release. And the next major release is 12.1. Or
were, all of them are obsolete.
> I didn’t do the original install, and
> this is running software from a company from which I no longer have
> support, but the application is tied to the version of the OS, so there
> isn’t much I can do there in terms of updates.
Ok… that partition is probably bad. Do what ratzi says, get a backup
done, then “smartctl -a /dev/whatever”, analize, the short and long tests.
On success, you need to run fsck on that partition. On no success,
replace disk and clone from backup.
If you can not mount the partition for backup, then do an image backup
with dd. If this aborts because of i/o errors, then use dd_rescue, or
dd_rhelp.
–
Cheers / Saludos,
Carlos E. R.
(from 12.3 x86_64 “Dartmouth” at Telcontar)
Thanks for all the pointers. dd did not work due to drive errors, dd_rescue has been running for about 18 hours and has identified over 10000 bad blocks on the disk. I will be doing a fresh install on a new hard drive. I have backups of a good amount of what was on the drive, but anything that can be recovered directly from this drive will be a bonus.
On 2013-07-26 18:56, rdnet wrote:
>
> Thanks for all the pointers. dd did not work due to drive errors,
> dd_rescue has been running for about 18 hours and has identified over
> 10000 bad blocks on the disk.
I’m sorry to hear that.
I would suggest you restart with “dd_rhelp” instead. The reason is that
this script locates bad regions and restarts dd_rescue as many times as
needed to skip bad regions. In a relatively short time it has copied all
the good regions of the disk or partitions, and then it tries to copy
the bad region, starting at the easy and ending on the difficult. In
some hours the basics are copied, but you may leave it for days to try
to copy everything.
The documentation explains it better than me, but it’s worth it. I
should have explained this on my previous post, but I hoped it would not
be necessary.
–
Cheers / Saludos,
Carlos E. R.
(from 12.3 x86_64 “Dartmouth” at Telcontar)