Hard disk has died (trashed partition table???)

Hard disk has died (trashed partition table???)

Post by Oliver B. Tupma » Mon, 20 Sep 1999 04:00:00



Hi there,

I turned my PC on this morning, it booted up as normal and delivered me
to my log-in prompt.

I entered my username and pressed enter, and nothing happened apart
from the hard drive light came on.

After a little while I got the following messages:

EXF-fs error (device ide0(3.74)): ext2_check_inodes_bitmap_wrong free
inodes count in grup=6, stored = 1, counted = 0

read_intr:      status=0x59 { DriveReady SeekComplete DataRequest Error}
read_intr:      error=0x00 {}
recal_intr:     status=0x31

There were more errors but they flew past at the speed of light and I
was
unable to write 'em down. I then rebooted but the same errors come up as
Linux tried to fsck the partitions.

I therefore downloaded a copy of tomsrtbt ( a micro-distribution) and
booted
up my PC with that.

When I run the command 'fdisk -l' I get the following information
about the hard drive.

Disk /dev/hdb: 255 heads, 63 sectors, 778 cylinders
Units = cylinders of 16065 * 512

Device          Boot    Start   End     Blocks          System
/dev/hdb1       *       1       192     1542208+        Linux native
/dev/hdb2               193     779     4709880         Extended
Partition 2 does not end of cylinder boundary:
        phys(778, 89, 63) should be (778, 254, 63)
/dev/hdb5               193     25      514045+         Linux native
/dev/hdb6               257     283     211645+         Linux native
/dev/hdb7               283     303     158728+         Linux native
/dev/hdb8               303     316     105808+         Linux native
/dev/hdb9               316     324     68008+          Linux swap
/dev/hdb10              324     344     158728+         Linux native
/dev/hdb11              344     472     1028128+        Win95 FAT32
/dev/hdb12              472     603     1050808+        Linux native

I'm wondering whether the problem is that error given for 'Partition
2'?

Also, if I try to attempt to install a second copy of Linux to the
1 GB of space left on the drive, the partition creation dies!

So how do I go about fixing my hard drive!?!

All the data I _really_ want is on /dev/hdb12, so if I can get that
off (onto a Zip drive) then that'd be all right. If I need to kill
all the other partitions on hdb (apart from hdb12) so I can get
the data on hdb12 off, then I suppose I'd have to do it!

Another observation is that whenever a program attempts to access
the damaged part of the hard drive, it says that hdb isn't ready
for any commands and that the ide0 reset times out!

Thanks in advance to all those who respond.

Cheers,

 Oliver B. Tupman

 
 
 

Hard disk has died (trashed partition table???)

Post by vilmoss.. » Mon, 20 Sep 1999 04:00:00



> I turned my PC on this morning, it booted up as normal and delivered me
> to my log-in prompt.

> EXF-fs error (device ide0(3.74)): ext2_check_inodes_bitmap_wrong free
> inodes count in grup=6, stored = 1, counted = 0

> read_intr:      status=0x59 { DriveReady SeekComplete DataRequest Error}
> read_intr:      error=0x00 {}
> recal_intr:     status=0x31

Hello Oliver,

It can mean many things. For example, a disk in the process of dying,
loose cable, or loose power connector. Let's hope it is the latter two.
BTW what kind of disk is this?

Quote:> When I run the command 'fdisk -l' I get the following information
> about the hard drive.

> Disk /dev/hdb: 255 heads, 63 sectors, 778 cylinders
> Units = cylinders of 16065 * 512

> Device          Boot    Start   End     Blocks          System
> /dev/hdb1       *       1       192     1542208+        Linux native
> /dev/hdb2               193     779     4709880         Extended
> Partition 2 does not end of cylinder boundary:
>         phys(778, 89, 63) should be (778, 254, 63)
> /dev/hdb5               193     25      514045+         Linux native
> /dev/hdb6               257     283     211645+         Linux native
> /dev/hdb7               283     303     158728+         Linux native
> /dev/hdb8               303     316     105808+         Linux native
> /dev/hdb9               316     324     68008+          Linux swap
> /dev/hdb10              324     344     158728+         Linux native
> /dev/hdb11              344     472     1028128+        Win95 FAT32
> /dev/hdb12              472     603     1050808+        Linux native

> I'm wondering whether the problem is that error given for 'Partition
> 2'?

No, that means your fdisk and the kernel have different ideas about the
disk' geometry. I had a machine running with such setup for a long time.

Quote:> All the data I _really_ want is on /dev/hdb12, so if I can get that
> off (onto a Zip drive) then that'd be all right. If I need to kill
> all the other partitions on hdb (apart from hdb12) so I can get
> the data on hdb12 off, then I suppose I'd have to do it!

I would try to boot from a rescue disk, mount /dev/hdb12, and try to
copy off everything from the partition. You might also try to boot from
a rescue floppy and copy the whole partition to another media.

dd if=/dev/hdb12 of=/where/you/want/it/to/show/up

In this case you *might* get an image of the partition and you can try
to salvage some data from that.

The most important thing I would do is to check the data and power
cables.

Vilmos

 
 
 

Hard disk has died (trashed partition table???)

Post by MBr » Tue, 21 Sep 1999 04:00:00




> > I turned my PC on this morning, it booted up as normal and delivered me
> > to my log-in prompt.

> > EXF-fs error (device ide0(3.74)): ext2_check_inodes_bitmap_wrong free
> > inodes count in grup=6, stored = 1, counted = 0

> > read_intr:      status=0x59 { DriveReady SeekComplete DataRequest Error}
> > read_intr:      error=0x00 {}
> > recal_intr:     status=0x31

> Hello Oliver,

> It can mean many things. For example, a disk in the process of dying,
> loose cable, or loose power connector. Let's hope it is the latter two.
> BTW what kind of disk is this?

I'm not sure, but the first thing you got to do is backup up your mbr
and partition table using a rescue disk or so
( dd if=/dev/hda of=/floppy bs=512 count=1).
This so you can safely toy around with some tools...
Second thing, get your self fsck 1.15, it has really improved...
If your drive supports SMART and your BIOS also, enabled it in
the BIOS so you can see if your drive is dying...
What brand & type of drive do you got? (So it's more easilly to judge
whether
the drive is dying, if it is eg. a Quantum Bigfoot ~~~FEAR~~~)

Quote:

> > When I run the command 'fdisk -l' I get the following information
> > about the hard drive.

> > Disk /dev/hdb: 255 heads, 63 sectors, 778 cylinders
> > Units = cylinders of 16065 * 512

> > Device          Boot    Start   End     Blocks          System
> > /dev/hdb1       *       1       192     1542208+        Linux native
> > /dev/hdb2               193     779     4709880         Extended
> > Partition 2 does not end of cylinder boundary:
> >         phys(778, 89, 63) should be (778, 254, 63)

This kinda thing can also happen when using partition managers like
Partition Magic, or when you specify the size in blocks under fdisk

- Show quoted text -

Quote:> > /dev/hdb5               193     25      514045+         Linux native
> > /dev/hdb6               257     283     211645+         Linux native
> > /dev/hdb7               283     303     158728+         Linux native
> > /dev/hdb8               303     316     105808+         Linux native
> > /dev/hdb9               316     324     68008+          Linux swap
> > /dev/hdb10              324     344     158728+         Linux native
> > /dev/hdb11              344     472     1028128+        Win95 FAT32
> > /dev/hdb12              472     603     1050808+        Linux native

> > I'm wondering whether the problem is that error given for 'Partition
> > 2'?

> No, that means your fdisk and the kernel have different ideas about the
> disk' geometry. I had a machine running with such setup for a long time.

> > All the data I _really_ want is on /dev/hdb12, so if I can get that
> > off (onto a Zip drive) then that'd be all right. If I need to kill
> > all the other partitions on hdb (apart from hdb12) so I can get
> > the data on hdb12 off, then I suppose I'd have to do it!

> I would try to boot from a rescue disk, mount /dev/hdb12, and try to
> copy off everything from the partition. You might also try to boot from
> a rescue floppy and copy the whole partition to another media.

> dd if=/dev/hdb12 of=/where/you/want/it/to/show/up

> In this case you *might* get an image of the partition and you can try
> to salvage some data from that.

> The most important thing I would do is to check the data and power
> cables.

I Agree

Quote:

> Vilmos

--
                "It may hurt my pride to be wrong once in a while, but I
                rather be flamed with better information than to be left
                blissfully ignorant." /-- Manuel Beunder, maintainer of:    
                http://www.euronet.nl/~mailme - The SB Live!-Linux page
 
 
 

Hard disk has died (trashed partition table???)

Post by Jayan » Tue, 21 Sep 1999 04:00:00


did you make any changes to the bios before the
booting when you got the error?

If you change the disk access mode - some bioses
let you select between DOS / Other mode of access.
And some of them let you select Windows / Other
for the operating system. There might be an option
of LBA turned off or on for the harddisk, which made
the geomentry information wrong. This makes a lot of
difference to the way the OS sees the disk, and
would attribute to errors similar to the one that
fdisk showed when you used tomsroot.

Therefore, check the above mentioned settinngs in
your bios, if you changed them, or try giving right
C/H/S parameters of your harddisk to lilo when booting up
man lilo.conf for help on this - look for something like
hd="number,number,number"

HTH

Jayan


> Hi there,

> I turned my PC on this morning, it booted up as normal and delivered me
> to my log-in prompt.

> I entered my username and pressed enter, and nothing happened apart
> from the hard drive light came on.

> After a little while I got the following messages:

> EXF-fs error (device ide0(3.74)): ext2_check_inodes_bitmap_wrong free
> inodes count in grup=6, stored = 1, counted = 0

> read_intr:      status=0x59 { DriveReady SeekComplete DataRequest Error}
> read_intr:      error=0x00 {}
> recal_intr:     status=0x31

> There were more errors but they flew past at the speed of light and I
> was
> unable to write 'em down. I then rebooted but the same errors come up as
> Linux tried to fsck the partitions.

> I therefore downloaded a copy of tomsrtbt ( a micro-distribution) and
> booted
> up my PC with that.

> When I run the command 'fdisk -l' I get the following information
> about the hard drive.

> Disk /dev/hdb: 255 heads, 63 sectors, 778 cylinders
> Units = cylinders of 16065 * 512

> Device          Boot    Start   End     Blocks          System
> /dev/hdb1       *       1       192     1542208+        Linux native
> /dev/hdb2               193     779     4709880         Extended
> Partition 2 does not end of cylinder boundary:
>         phys(778, 89, 63) should be (778, 254, 63)
> /dev/hdb5               193     25      514045+         Linux native
> /dev/hdb6               257     283     211645+         Linux native
> /dev/hdb7               283     303     158728+         Linux native
> /dev/hdb8               303     316     105808+         Linux native
> /dev/hdb9               316     324     68008+          Linux swap
> /dev/hdb10              324     344     158728+         Linux native
> /dev/hdb11              344     472     1028128+        Win95 FAT32
> /dev/hdb12              472     603     1050808+        Linux native

> I'm wondering whether the problem is that error given for 'Partition
> 2'?

> Also, if I try to attempt to install a second copy of Linux to the
> 1 GB of space left on the drive, the partition creation dies!

> So how do I go about fixing my hard drive!?!

> All the data I _really_ want is on /dev/hdb12, so if I can get that
> off (onto a Zip drive) then that'd be all right. If I need to kill
> all the other partitions on hdb (apart from hdb12) so I can get
> the data on hdb12 off, then I suppose I'd have to do it!

> Another observation is that whenever a program attempts to access
> the damaged part of the hard drive, it says that hdb isn't ready
> for any commands and that the ide0 reset times out!

> Thanks in advance to all those who respond.

> Cheers,

>  Oliver B. Tupman

 
 
 

1. Trashed Partition table / slice table HELP!

Partition Magic trashed my FreeBSD partition table.  It set it
to sysid 181 and reset the partition entrys to funky points. -I didn't
actually tell it to do anything with the drive FreeBSD is (was...) on.

Is there ANY way to restore the partition table and the slice table?

Please help.  Thanks.

--

     http://www.best.com/~zenin/treehouse/       Just another Perl hacker
     Spelling mistakes? Their couldn't be. -My modem is error correcting.
     The oppions stated here are mine and you can't have them.  So there.

2. Unix - Frequently Asked Questions (Contents) [Frequent posting]

3. HELP! Disk Druid trashed my partition table!

4. PPP logoff problem

5. Hard Disk Partition Table Corrupted

6. Linux on the Desktop...TODAY!

7. Lost My Partition Table / Hard Disk Recovery

8. help: snmpd no response

9. Sun Netra X1 IDE hard disk and partition or slice table configuration.

10. HELP: partition table trashed

11. DOS FDISK trashed my partition table, can I save my Linux on /dev/hda3

12. Help!!trashed partition table Help!!

13. Yggdrasil Fall '94 TRASHES a HD partition table!?