07-09-04 08:53 PM
Mike Lascuola wrote:
> I am new to linux (just installed Fedora last week), and here is my
> task. I have a 30GB hard drive from an Alesis HD-24 hard disk
> recorder, on which I had recorded over an hour of multitrack data when
> the power failed. In my hard luck of hard luck stories, the new UPS I
> purchased just for this gig was DOA, so I was running without a UPS.
>
> That is water under the bridge...
>
> I've heard tell that if I overwrite the header on the hard drive with
> what the header would have looked like, I can get all of the data.
> So, I took an identical Maxtor hard drive, and recorded over an hour
> of blank space using the Alesis HD-24. I've heard that linux users
> can view the data on the drive no matter what the file system (it is
> Alesis proprietary)
You can read/write raw bytes, but in this case you're unlikely to be
able to make sense out of the data unless you know a *lot* about the
internals of the Alesis system.
Linux will not be able to read the proprietary Alesis filesystem as a
filesystem. Linux will just see a stream of raw data. It is up to
*you* to interpret that data by yourself. IMHO, for that reason, you'd
be better off talking to other Alesis users (or Alesis directly) about
data recovery. I can't stress enough the need for *you* to understand
the Alesis disk format/filesystem/whatever, and that's not likely
information you'll be able to get here in a Linux newsgroup.
> to see where the header and data start and stop,
> and write the good header to the original hard drive.
You either have to know that the header occupies certain parts of the
disk, or know how to tell the difference between the header and your own
data. Do you know if the header is always a fixed size at the start of
the disk? Alternatively, do you know the on-disk format of the data?
(In that case anything that didn't match that format _might_ be header.)
> It looks like I'll need the dd command and maybe SFdisk, but --
> how?!!!
>
> 1) What command do I use to read the data on the devices to make sure
> I know what is what?
dd if=/dev/<whatever> of=/tmp/disk.img
Substiture /dev/<whatever> with the device node for the disk you're
trying to recover. (E.g., /dev/hdc or /dev/sdb)
Now /tmp/disk.img is a complete copy of the drive. You can edit it with
a hex editor or do whatever you like with it from there without altering
the original.
> 2) (STUPID) The only way I have found to get a command prompt in
> Fedora is to use Terminal, and then I have to close the window and
> open another one to enter the next command! I know there is an easier
> way (with command history, etc., like doskey in Windows)
Fire up an xterm or equivalent. Or press Ctrl-Alt-F1 to leave graphical
mode entirely and work at the command prompt only.
> 3) Do I have to copy the data to the OS boot drive, or can I copy
> directly from the one 30GB device to the other?
You certainly don't need to copy to the boot drive or anywhere else
unless you want to. However, I'd strongly recommend making a copy of
the failed device, working with that, and putting the original in a safe
place. That way, if you make a really big mistake, you can always go
back to the beginning and start over if you have to.
You also might want to consider contacting a company that specializes in
data recovery. They're not cheap, but if your data is valuable, it
could well be worth it.
[ Post a follow-up to this message ]
| 
