[Bug-ddrescue] Re: ddrescue crashes system on kernel 2.6.18

[Arik Raffael Funke]

Hi Andrew,

thanks for taking the time to help out!

I originally had the drive on a usb interface when I was still trying my 
luck with dd for rescuing the data. After that was less a than rewarding 
experience, I assumed: the least layers between program and drive, the 
better. - Boy was I wrong...

After a first run with --no-split I am now only missing merely 3.2 Mb 
from a 20Gb harddrive.

ddrescue is an amazing tool. I especially love the logging function. 
Thanks for Antonio for his great work and to those supporting it here!

Can anybody point me to a tutorial on how to figure out which files were 
affected by the ? Given the small error, I would like to avoid having to 
be unnecessarily suspicious of every file. I can imagine in principle 
how to find files that were entirely in the error region, but partially 
affected files I am not sure about.

Many thanks,
Arik


On 10/01/2010 00:13, andrew zajac wrote:
>   Hi Arik.
>
> I don't know if the list of bad blocks is stored on the media itself or
> on memory on the drive's PCB. If the latter is the case, then you will
> run into recovery issues as the drive's OS assumes the bad blocks of the
> previous drive. Is that supposed to crash your system? I don't know.
>
> Can you try changing the interface? For example, are you able to use a
> USB interface instead of plugging in the drive directly to the
> motherboard? That would be less likely to bring down your system. Yes,
> it's much slower. But at least you can continue where you left off.
>
> Good Luck.
>
> Andrew Zajac
>
>
>
> --- On *Sat, 1/9/10, Arik Raffael Funke
> /<address@hidden>/* wrote:
>
>
>     From: Arik Raffael Funke <address@hidden>
>     Subject: [Bug-ddrescue] ddrescue crashes system on kernel 2.6.18
>     To: address@hidden
>     Received: Saturday, January 9, 2010, 3:29 PM
>
>     Hi,
>
>     Background: I have had a power supply unit failure and as a result
>     the PCBs of three hard drives were fried as well as the processor.
>
>     I have replaced the PCB from an identical drive for the data drive,
>     a WD200EB-00CSF0. The harddrive is seems to work fine for almost all
>     files: It can be mounted, data is generally accessible, etc.
>
>     When I try to image the drive with ddrescue, it proceeds without
>     errors within 20 minutes to 16.5Gb/20Gb where it encounters the
>     first errors. It then takes approximately another 5-10 minutes
>     apparently only getting errors before finally crashing the whole system.
>
>     A photograph of the screen can be seen here: (Sorry for the flash...)
>     http://img64.imageshack.us/img64/745/crasho.jpg
>
>     The result is the same with:
>     ddrescue -n
>     ddrescue --direct
>     ddrescue --raw
>
>     All attempts were made in runlevel 1 on a Centos 5.4 system with
>     kernel 2.8.18. Resuming recovery with the log file makes the system
>     crash shortly after resuming.
>
>
>     Can anybody tell me how to either:
>     - read out a complete image (obviously without the presumably bad
>     sectors) so that I can run file system recovery? The image now is
>     truncated.
>
>     - fix the hard drive problem? (I.e. why should the hard drive have
>     bad sectors in the last few GB given a mere power surge?)
>
>     Many thanks for any help!
>
>     - Arik
>
>
>     _______________________________________________
>     Bug-ddrescue mailing list
>     address@hidden
>     </mc/compose?to@>.ڝò‹À'j¸w7ŒØŽš>
>     http://lists.gnu.org/mailman/listinfo/bug-ddrescue
>
>
> Instant message from any web browser! Try the new * Yahoo! Canada
> Messenger for the Web BETA*
> <http://ca.messenger.yahoo.com/webmessengerpromo.php>
>
>
>
> _______________________________________________
> Bug-ddrescue mailing list
> address@hidden
> http://lists.gnu.org/mailman/listinfo/bug-ddrescue