Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.os.linux.hardware > #898 > unrolled thread

Recover zonked USB sticks

Back to article view | Back to comp.os.linux.hardware

Contents

  Recover zonked USB sticks Haines Brown <haines@HistoricalMaterialism.info> - 2011-12-05 11:37 -0500
    Re: Recover zonked USB sticks Lew Pitcher <lpitcher@teksavvy.com> - 2011-12-05 12:24 -0500
      Re: Recover zonked USB sticks Haines Brown <haines@HistoricalMaterialism.info> - 2011-12-06 13:03 -0500
        Re: Recover zonked USB sticks "David W. Hodgins" <dwhodgins@nomail.afraid.org> - 2011-12-06 16:01 -0500
        Re: Recover zonked USB sticks Henrik Carlqvist <Henrik.Carlqvist@deadspam.com> - 2011-12-06 23:13 +0100
          Re: Recover zonked USB sticks Haines Brown <haines@HistoricalMaterialism.info> - 2011-12-07 10:42 -0500
    Re: Recover zonked USB sticks Richard Kettlewell <rjk@greenend.org.uk> - 2011-12-06 19:29 +0000
      Re: Recover zonked USB sticks Haines Brown <haines@HistoricalMaterialism.info> - 2011-12-07 10:39 -0500
      Re: Recover zonked USB sticks Haines Brown <haines@HistoricalMaterialism.info> - 2011-12-07 12:13 -0500
        Re: Recover zonked USB sticks Richard Kettlewell <rjk@greenend.org.uk> - 2011-12-07 17:25 +0000

#898 — Recover zonked USB sticks

I am in the habit of putting an ISO on a USB stick to netinst install
the Debian Squeeze system. However, when I try to do it with a custom
ISO and installer combo, the # zcat command destroys the stick. After
doing it, I now plug the stick in (with automount) I get:

  $ dmesg
  ...
  [250357.735511] sd 28:0:0:0: [sdb] Assuming drive cache: write through 
  [250357.735514]  sdb:
  
  $ df -k
  ...
  /dev/sdb        489208     25408    463800   6% /media/usb0

  $ mount 
  ...
  /dev/sdb on /media/usb0 type vfat
    (rw,noexec,nodev,sync,noatime,nodiratime)

With the custom boot image I do: 

  # zcat boot-i386-0808.img.gz > /dev/sdb

  $ ls /media/usb0
  adgtk.cfg    f2.txt f8.txt     initrd.gz   rqtxt.cfg     vesamenu.c32
  adtxt.cfg    f3.txt f9.txt     ldlinux.sys setup.exe  win32-loader.ini 
  disk.lbl     f4.txt g2ldr      linux       splash.png
  exithelp.cfg f5.txt g2ldr.mbr  menu.cfg    stdmenu.cfg
  f10.txt      f6.txt gtk.cfg    prompt.cfg  syslinux.cfg
  f1.txt       f7.txt initrdg.gz rqgtk.cfg   txt.cfg
  
However these files are probably remnants of a prior run of zcat on
a stock ISO, but they are not in /dev/sdb1 for some reason.

  # cfdisk /dev/sdb
  FATAL ERROR: Bad primary partition 0: Partition begins after
         end-of-disk 

I've done this to two sticks. Although they are old, small and cheap, it
would be nice to recover them. Any suggestions?

Haines Brown
 

[toc] | [next] | [standalone]

#899

On December 5, 2011 11:37, in comp.os.linux.hardware,
haines@HistoricalMaterialism.info wrote:

> I am in the habit of putting an ISO on a USB stick to netinst install
> the Debian Squeeze system. However, when I try to do it with a custom
> ISO and installer combo, the # zcat command destroys the stick. After
> doing it, I now plug the stick in (with automount) I get:
> 
>   $ dmesg
>   ...
>   [250357.735511] sd 28:0:0:0: [sdb] Assuming drive cache: write through
>   [250357.735514]  sdb:
>   
>   $ df -k
>   ...
>   /dev/sdb        489208     25408    463800   6% /media/usb0
> 
>   $ mount
>   ...
>   /dev/sdb on /media/usb0 type vfat
>     (rw,noexec,nodev,sync,noatime,nodiratime)
> 
> With the custom boot image I do:
> 
>   # zcat boot-i386-0808.img.gz > /dev/sdb
> 
>   $ ls /media/usb0
>   adgtk.cfg    f2.txt f8.txt     initrd.gz   rqtxt.cfg     vesamenu.c32
>   adtxt.cfg    f3.txt f9.txt     ldlinux.sys setup.exe  win32-loader.ini
>   disk.lbl     f4.txt g2ldr      linux       splash.png
>   exithelp.cfg f5.txt g2ldr.mbr  menu.cfg    stdmenu.cfg
>   f10.txt      f6.txt gtk.cfg    prompt.cfg  syslinux.cfg
>   f1.txt       f7.txt initrdg.gz rqgtk.cfg   txt.cfg
>   
> However these files are probably remnants of a prior run of zcat on
> a stock ISO, but they are not in /dev/sdb1 for some reason.

Given that you zcat the image to /dev/sdb, you no longer /have/ a /dev/sdb1. 

>   # cfdisk /dev/sdb
>   FATAL ERROR: Bad primary partition 0: Partition begins after
>          end-of-disk

Yup. No valid partitions on /this/ device.

Remember, when you write to /dev/sdb, you are writing to the complete
device, without regards for it's partition table. In fact, since the
partition table sits within the first 512 bytes of /dev/sdb, your zcat
overwrites that table with data, and you no longer /have/ a partition table
(as such).

> I've done this to two sticks. Although they are old, small and cheap, it
> would be nice to recover them. Any suggestions?

How do you mean "recover them"? Do you mean
a) regain access to the data that you placed on /dev/sdb ? or
b) regain access to the data that originally sat on /dev/sdb1? or
c) clean them up so that no data remains, and that they can be reused as
   standard devices again? or
d) something else?

If you mean a), then simply mount /dev/sdb to a mountpoint.
If you mean b), then unless you took backups before you started, you are out
of luck. You clobbered the partition table, and any data on that partition
with the zcat command.
If you mean c), then you can simply fdisk /dev/sdb, and create a new
partition table. Then mkfs /dev/sdb1 with the proper filesystem.
If you mean d), then please elaborate.

HTH
-- 
Lew Pitcher
Master Codewright & JOAT-in-training   | Registered Linux User #112576
Me: http://pitcher.digitalfreehold.ca/ | Just Linux: http://justlinux.ca/
----------      Slackware - Because I know what I'm doing.         ------

[toc] | [prev] | [next] | [standalone]

#900

Lew Pitcher <lpitcher@teksavvy.com> writes:

> How do you mean "recover them"? Do you mean


> c) clean them up so that no data remains, and that they can be reused
> as standard devices again? or


> If you mean c), then you can simply fdisk /dev/sdb, and create a new
> partition table. Then mkfs /dev/sdb1 with the proper filesystem.

Lew, sorry I wasn't clearer. It is your option (c) that I want: make the
stick usable again. The problem is that the FATAL ERROR message is when
I try to run cfdisk on the device. Cffisk can't see it.

Perhaps I should add that while a # df -k command tells me the device is
/dev/sdb, the kernel can't see it ($ dmsg does not report it).

The reason for my wish to recovery is that I want to try and zcat a
custom boot.img to /dev/sdb again and have my supply of UBS sticks to
dwindle.

Haines

[toc] | [prev] | [next] | [standalone]

#905

On Tue, 06 Dec 2011 13:03:29 -0500, Haines Brown <haines@historicalmaterialism.info> wrote:

> I try to run cfdisk on the device. Cffisk can't see it.

cfdisk -z /dev/???

Regards, Dave Hodgins

-- 
Change nomail.afraid.org to ody.ca to reply by email.
(nomail.afraid.org has been set up specifically for
use in usenet. Feel free to use it yourself.)

[toc] | [prev] | [next] | [standalone]

#906

Haines Brown <haines@HistoricalMaterialism.info> wrote:
> Perhaps I should add that while a # df -k command tells me the device is
> /dev/sdb, the kernel can't see it ($ dmsg does not report it).

If "df -k" tells you that the device is /dev/sdb it means that you have
mounted the device, you would not have been able to mount it without the
kernel to see it.

When the device is mounted you should not try to alter the partition table
or the file system.
 
> The reason for my wish to recovery is that I want to try and zcat a
> custom boot.img to /dev/sdb again and have my supply of UBS sticks to
> dwindle.

If you once again want to put something to sdb like that there is no
reason to put a partition table on /dev/sdb first. However, you whould
first unmount the device before doing any such operation.

regards Henrik
-- 
The address in the header is only to prevent spam. My real address is:
hc123(at)poolhem.se Examples of addresses which go to spammers:
root@localhost postmaster@localhost

[toc] | [prev] | [next] | [standalone]

#908

Henrik Carlqvist <Henrik.Carlqvist@deadspam.com> writes:

>> Perhaps I should add that while a # df -k command tells me the device is
>> /dev/sdb, the kernel can't see it ($ dmsg does not report it).
>
> If "df -k" tells you that the device is /dev/sdb it means that you have
> mounted the device, you would not have been able to mount it without the
> kernel to see it.
>
> When the device is mounted you should not try to alter the partition table
> or the file system.

Sorry, should have made clearer that I unmounted the stick before
issuing the cfdisk command. 

It turns our that zeroing the first partition as suggested did the
trick.

Haines

[toc] | [prev] | [next] | [standalone]

#902

Haines Brown <haines@HistoricalMaterialism.info> writes:
> However these files are probably remnants of a prior run of zcat on
> a stock ISO, but they are not in /dev/sdb1 for some reason.
>
>   # cfdisk /dev/sdb
>   FATAL ERROR: Bad primary partition 0: Partition begins after
>          end-of-disk 
>
> I've done this to two sticks. Although they are old, small and cheap, it
> would be nice to recover them. Any suggestions?

Zero out the first sector, e.g.:

  dd if=/dev/zero of=/dev/sdb count=1 bs=512

Make sure you get the device name right.

cfdisk will probably tolerate it thereafter.

-- 
http://www.greenend.org.uk/rjk/

[toc] | [prev] | [next] | [standalone]

#907

Richard Kettlewell <rjk@greenend.org.uk> writes:

> Haines Brown <haines@HistoricalMaterialism.info> writes:

>>   FATAL ERROR: Bad primary partition 0: Partition begins after
>>          end-of-disk 

> Zero out the first sector, e.g.:
>
>   dd if=/dev/zero of=/dev/sdb count=1 bs=512

Yes, thanks. That seems to be what was needed.

Haines

[toc] | [prev] | [next] | [standalone]

#909

Richard Kettlewell <rjk@greenend.org.uk> writes:

> Zero out the first sector, e.g.:
>
>   dd if=/dev/zero of=/dev/sdb count=1 bs=512

Well, I spoke too soon. Still can't partition the USB stick. 

When the stick processed as above is inserted, dmsg tells me that the
stick is "sdc: sdc1" (since initially writing I added another HD and
thus the stick no longer sdb). But only sometimes (see below).

The automount did not mount it, and the kernel fails to see it (df -k
returns nothing for it). 

When I try to mount it:

  $ sudo mount -t vfat /dev/sdc /media/usb0
  mount: wrong fs type, bad option, bad superblock on /dev/sdc,
       missing codepage or helper program, or other error
       In some cases useful info is found in syslog - try
       dmesg | tail  or so

When I run cfdsk on the unmounted stick I now get:

   $ sudo cfdisk /dev/sdc
   sdc1 Boot Primary FAT16 <32M>  ---- 1008.43 MG

Isn't this what I should expect?

However, it may not be the stick's fault, for I try to run cfdsk on a
fresh mint unmounted stick, and get this error:

   # cfdisk /dev/sdd  (dmsg tells me the device is sd0: sdd1)   
   FATAL ERROR: Bad primary partition 0: Partition begins after
                          end-of-disk 

So obviously I'm doing something stupid along the line. 

I should mention that I have erratic USB optical mounse movements
(pointer hangs a bit, but this behavior comes and goes), and
often when I expect dmesg to tell me about a key I just inserted,
it instead tells me again about the mouse (I do pause):

  usb 4-1: Product: USB-PS/2 Optical Mouse
  usb 4-1: Manufacturer: B16_b_02
  usb 4-1: configuration #1 chosen from 1 choice
  input: B16_b_02 USB-PS/2 Optical Mouse as
    /devices/pci0000:00/0000:00:1a.1/usb4/4-1/4-1:1.0/input/input90 
  generic-usb 0003:046D:C024.0056: input,hidraw0: USB HID v1.10 Mouse
    [B16_b_02 USB-PS/2 Optical Mouse] on usb-0000:00:1a.1-1/input0

Sounds like some kind of driver confusion, and the mouse driver is
being loaded for the USB stick.

Haines

[toc] | [prev] | [next] | [standalone]

#910

Haines Brown <haines@HistoricalMaterialism.info> writes:
> Richard Kettlewell <rjk@greenend.org.uk> writes:

>> Zero out the first sector, e.g.:
>>
>>   dd if=/dev/zero of=/dev/sdb count=1 bs=512
>
> Well, I spoke too soon. Still can't partition the USB stick. 
>
> When the stick processed as above is inserted, dmsg tells me that the
> stick is "sdc: sdc1" (since initially writing I added another HD and
> thus the stick no longer sdb). But only sometimes (see below).
>
> The automount did not mount it, and the kernel fails to see it (df -k
> returns nothing for it). 
>
> When I try to mount it:
>
>   $ sudo mount -t vfat /dev/sdc /media/usb0
>   mount: wrong fs type, bad option, bad superblock on /dev/sdc,
>        missing codepage or helper program, or other error
>        In some cases useful info is found in syslog - try
>        dmesg | tail  or so

mount's error reporting is atrocious  l-(

> When I run cfdsk on the unmounted stick I now get:
>
>    $ sudo cfdisk /dev/sdc
>    sdc1 Boot Primary FAT16 <32M>  ---- 1008.43 MG
>
> Isn't this what I should expect?

It sounds like you've created a partition but not initialized any kind
of filesystem in it.

-- 
http://www.greenend.org.uk/rjk/

[toc] | [prev] | [standalone]

Back to top | Article view | comp.os.linux.hardware

csiph-web