Error attempt to read or write outside of partition entering rescue mode

I am about to cry…..

I left the computer as a 3rd night running ddrescue doing a cloning disk to disk.

When i got up this morning just the terminal was running but anything else had dissapeared…. the desktop and bars had dissapeared so i could work or do anything (i dont really now how to you the terminal much i use xwindows for linux mint even to install programs) . So I closed the terminal and switched off the laptop manually since i knew i could continue ddrescue with the log file!!

I disconnected the 2 disks i had attached and I restarted the computer….. this is what I got

error: attempt to read or write outside of partition
Entering rescue mode…
grub rescue>

what do I do now? take into the account i do not have another computer and i have a lot information in there since i emptied my new disk to clone the old disk in my laptop….. info that i dont want to loose!!

I have tried this

grub rescue > ls (hd0,msdos5)/ # this is root and bootloader of linux
grub rescue > set root=(hd0,msdos5)
grub rescue > set prefix=(hd0,msdos5

but when I do

Grub rescue> ls /boot says error: attempt to read or write outside of partition as well if I do
Grub rescue> ls /home gives the same error but

I can do fine:
Grub rescue> ls /media or ls /root or ls /lib

I have tried

grub rescue > insmod normal but it says the sames
error: attempt to read or write outside of partition

what can i do??

Thanks in advance

Make sure you didn’t just disconnect the wrong disk, because that’s what it sounds like. Unless you were trying to ddrescue off the disk you are now trying to boot from, and since using ddrescue implies that the disk was already damaged it might simply be too far gone now.

Either way I’d boot from the Linux Mint installation medium (create one at another computer if you don’t have one, ideally the 19.2 version) and check the status of the disk from there before writing anything to it.

The other 2 disks where external hard disks… what is wrong is the internal HD.

is there something I can do from the Grub rescue prompt before trying to create a booting usb?

I was hoping not to have to do an live usb to boot… i don’t really know how… the other computer have is a old tower (bought in 2000) on windows xp full of dust…. i will try to connect it to a monitor and see what I can do.

Well, that depends on how accurate your claims in the OP were. If (hd0,msdos5) really is the one that holds /boot and thus the kernel, then I don’t see a way around booting from any live environment to repair the drive and likely reinstall GRUB and the kernel and whatever else may have been lost.

But it’s also possible that you were not actually looking at the correct partition. Since this is an old legacy BIOS/MBR boot situation it would not surprise me if you had a separate partition for /boot (and /home). In that case you can load the kernel from that partition.

After a night leaving ddrescue running to clone a external disk to another external disk and creating the log file in my internal hard disk, I got up and the laptop was stuck so i restarted it.

When it restarted:
error: attempt to read or write outside of partition
Entering rescue mode
Grub rescue>

I tried few things on internet but didn’t work…
So i have done a live usb…
I can see 2 devices in /media… one Acer whic is my windows partition and i can access to all the files… Great!!!

The second partition is the linux one 76GB having a large name b7f22c6d-85c1-445a-83f6 and same more numbers and later…
In the terminal when I get into this folder and I do a > ls it shows:

ls: cannot access ‘home’: Structure needs cleaning
ls: cannot access ‘boot’: Structure needs cleaning

And the same for the folder opt, etc, lib64, mnt, usr,cdrom

I can access media, dev, lib

I have tried to run fsck and fsck.ext4 many times… it starts great and then aborts …e2fsck aborted!
I ran it many times but it does fix some errors and aborts again.
it seems like the access to /home /boot is locked
What else I can do… to be honest I want to access to /home folder to save all the info i had there… if i cant recover the system … well i will install linux mint again (i hope this is possible) but i dont want to loose the info i had there!

How can i recover the info in /home?
How can i repair the boot?
How can I repair the system?

Oh well… I decided to run fsck again

fsck -y /dev/sda5

It aborted and I ran it again… like that 6 times and finally it finished it without aborting!!
I thought it was great but oh my god!!

fsck deleted all the folders that were locked and could access to it, therefore /home /boot /opt /etc /lib64 /mnt /usr /cdrom folders have been deleted… I had 34GB free in that partition … now I have 61GB… what do i do to recover my /home folder?
I am guessing with those folder missing my linux system is gone!!

At this point, I think a clean reinstall is in order. I hope you had your /home backed up. Nothing like a situation like this to emphasize the importance of back-ups.

Larry

tonilib wrote: ↑

Sun Dec 01, 2019 12:43 pm

… to be honest I want to access to /home folder to save all the info i had there…

Sounds like you don’t have backups. That’s unfortunate. You might be looking at data recovery, which is complicated and a lot of work. Never had to do it myself, but TestDisk is the utility I see recommended. More info at Ubuntu Forum and CG Security.

Don’t do this right away. Wait to see whether anyone has a less drastic remedy. TestDisk is your last resort.

I am getting the following grub error: attempt to read or write outside of disk 'hd0'.

Is there any way to fix this error? I can’t boot into Debian because of this.

Partitions:

SCSI1 (0,0,0) (sda) - 250.1 GB ATA WDC WD2500BEVE-0
    #1 Primary 249.0 GB B ext4
    #5 logical 1.1 GB F swap swap
SCSI3 (0,0,0) (sdb) - 7.8 GB USB Flash Memory
    #1 primary 7.8 BG B fat32

SCSI3 is a USB device to load in some firmware needed for the install of Debian.

Because you’re usign debian, it will be very easy to recover grub, using the installation CD.

Boot the install Debian CD and select Advanced options -> Rescue Mode
There you will have options to reinstall grub or to go to a chroot shell, and reinstall grub using:

grub-mkdevicemap  
grub-install /dev/sda  
update-grub

answered May 25, 2016 at 17:25

I pulled my 17 yr. old Intel D850GB based build out of the closet, and fired it up. It had XP Home on it that (for some reason) wanted re-activation. I didn’t care about that, I wanted to put Fedora on it. However, during the boot process of the install disc, got the «display out of range» signal on the monitor. Not much I can do with that, so I tried other discs. I tried Debian 9.5 netinst, Ubuntu 16.04 sever…anything that would fit on a CD, as this machine does not have a DVD drive, and is too old to boot from USB. I would get the distros installed but upon reboot would get:

error: attempt to read or write outside of disk 'hd0'

entering rescue mode
grub rescue>

Just for the record- the HDD is a Western Digital IDE 7200 RPM.

O.K., so I tried to fix this with:

grub rescue > ls (hd0)
grub rescue > set root=(hd0)
grub rescue > set prefix=(hd0)/boot/grub
grub rescue > insmod normal
grub rescue > normal

But….that didn’t work. Upon reboot would still get «error: attempt to read or write outside of disk ‘hd0′».

I was successful getting an old copy of Debian 6.06(badly EOL) installed and working without any «hd0» errors.

Question:What does «error: attempt to read or write outside of disk ‘hd0′» actually indicate? Something wrong with the hard drive?
Something not working during partitioning?[/QUOTE]

bw123 wrote:Did you try any of the solutions you found when looking up the error msg?

Yes, the solutions I came across from numerous sources were:

grub rescue > ls (hd0)
grub rescue > set root=(hd0)
grub rescue > set prefix=(hd0)/boot/grub
grub rescue > insmod normal
grub rescue > normal

Which did not work.
However, one solution mentioned here was:

sashkello wrote:OK, SOLVED!

I figured hard drives can’t be the problem, and the motherboard doesn’t seem to have any obvious issues either, otherwise there’d be problems with working from USB boot, which works perfectly.

Then, the only place the problem might reside is somewhere in between these two. So, I cracked it open once again, removed the cable connecting the drive to the motherboard and noticed the tiniest of dust particles sitting on one of the contacts. It must have got there when I was replacing the drive for the very first time, and since I only disconnect the part of the cable which goes into the hard drive, it was sitting there unmoved all this time.

Phew! Such a stupid thing and so much trouble from it. Like my dad taught me: «If it doesn’t work, disassemble it and assemble it back, that will solve it in 90% of situations»

Worth investigating, I wasn’t looking for hardware problems up to this point, although as I mentioned- Debian 6.06 did install nicely without a hitch.

I have ‘new’ ribbon cables and IDE drives, I’ll tinker further, as I really want an up-to-date OS. I can pull packages from old-stable archives but no security updates.

It could be that the old BIOS is unable to read/write disk sectors beyond some limit, but GRUB needs to read data beyond that limit.
What is the capacity of the disk ? How did you partition it ?
You can try to install with a small partition (100 MB-1 GB) located at the beginning of the disk and mounted on /boot.

PS : the «solution» you mention cannot work on a partitioned disk. Anyway I do not see how it would help with your issue.

p.H wrote:It could be that the old BIOS is unable to read/write disk sectors beyond some limit, but GRUB needs to read data beyond that limit.
What is the capacity of the disk ? How did you partition it ?
You can try to install with a small partition (100 MB-1 GB) located at the beginning of the disk and mounted on /boot.

PS : the «solution» you mention cannot work on a partitioned disk. Anyway I do not see how it would help with your issue.

The hdd is a WD 250GB 7200rpm drive, The BIOS recognises the drive correctly. I checked(actually replaced) the cables and blew out the PATA connector on the drive, like was suggested in the quote in my post above. That made no difference. I tried a small partition(/boot) at the beginning of the disk, made no difference when trying to install the newer OS’s, still the «attempt to read or write outside of disk hd)» error.
However, I once again slipped in my Debian 6 disc, told it to use the whole disk, automatic partitioning…gave me a 249GB ext3 partition, and a 1GB swap. Boots great!
I’ve been wondering «what the difference is» between the Debian 6 and all the other newere install discs…maybe ext3 vs. ext4 file system writing…?
https://www.golinuxhub.com/2014/03/what … 3-and.html
^^makes note to self^^ try another Debian 9.5 install, using ext3 file system.
Anyway, I bought a refurb’d WD 80GB hdd for testing, maybe something wrong with the 250GB drive. :shrug: I’m pretty sure that an OS using ext4 can install on an IDE PATA drive. Ext4 was incorporated into the 2.6.28 kernel back in December 2008, but I only had one SATA drive at that time, so Lenny would have been installed on one of the older drives.

**In other news**

As part of my annual Fall update on all my old machines, I did a Debian 9.5 netinst on a Athlon xp1800

machine with a Maxtor 250GB IDE hard drive, that had no read/write outside of disk issues whatsoever.

Only real issue is both Firefox esr and Chromium refuse to run, has to do with the processor not having SSE2 instruction set.

https://support.mozilla.org/en-US/kb/yo … -supported

Epiphany is crippled, presumably for similar reasons.

Midori does work, but when visiting gmail displays in [very basic] basic html mode.