Freezing hard drives for data recovery [COLUG]

[Jim]

On Wed, 26 Jul 2006 12:44:32 +1000 (EST) Duane <duane at cacert.org> wrote:

> I'm trying the freezer trick first...

Good. 

There are things you can do to improve the chances of it working, 
which have been discussed  previously on the list more than once. 

Here's part of one such discussion: 



Begin forwarded message:

Date: Tue, 01 Feb 2005 22:28:20 -0500
From: Jim <jep200404 at columbus.rr.com>
To: Central OH Linux User Group <colug432 at colug.net>
Subject: Re: [COLUG] hard disk recovery?


Oh boy, 

Here's the omnibus response.  

On Tue, 01 Feb 2005 12:37:08 -0500 Mike Lietz <codger at gmail.com> wrote:

> I've got a 20GB Maxtor drive that decided last night to give up the
> ghost. 

> Does anybody have experience with hard disk recovery? 

I've had a number of drives die and have also heard of the freezer 
technique.  I thought it was nonsense until I tried it.  

  It worked for me and I got the data off.  

When the drive warmed up it quit working.  
After a good cold soak in the freezer, it worked yet again.  
Of course, I can not trust it.  I have put it back in the freezer.  
Hmmm.  Maybe this could be a COLUG meeting demo.  
Could I get yet another "last time read" out of it?  
Of course, I'd have to do the last time write to replace 
the proprietary info with innocuous fluff.  

Instead of just a couple hours, let it soak overnight 
to let that asymptote get close the freezer's temp.  

Also, when you pull it out, it will collect condensation which 
can confuse the electronics, so I put it in an antistatic bag 
_before_ I put it in the freezer.  Also, I put it in the freezer 
with the IDE cable attached to minimize handling of the drive 
when I remove it from the freezer.  

When you remove it from the freezer, you don't really know how 
much time you'll have, so I prepare as much as I can before I 
remove it from the freezer.  

The PC is open, ready for the IDE cable to be plugged in.  
The main (OK) drive already has pretested scripts for 
sucking the data from the frozen drive.  The drive in the 
freezer already has an antistatic bag taped shut with the 
IDE cable sticking out.  Of course, master/slave jumpers 
are already installed.  I also put some insulation 
(styrofoam or my wool cap) in the freezer.  

Then when it's time for action after the cold soak, 
1. Open the freezer.  
2. Using gloves, put the frozen drive in the insulation.  
      (The insulation keeps air from circulating over the drive, 
       so this also puts another time limit on using the drive if it 
       works.  
3. Plug the frozen drive into the PC.  
4. Boot the PC.  
5. Run a few commands (like fdisk -l /dev/hdx) to verify 
   very basic info.  
6. Run the scripts (with rsync/dd/tar/nc/whatever).  

On Tue, 01 Feb 2005 14:24:58 -0500 Mike Lietz <codger at gmail.com> wrote:

> To the people who froze a drive: was
> it recognized in the BIOS beforehand? 

I don't remember.  Certainly it was after freezing, 
or else I couldn't have recovered the data.  

On Tue, 01 Feb 2005 13:07:28 -0500 Pat Collins <pat at linuxcolumbus.com> wrote:

> On Tue, Feb 01, 2005 at 12:50:05PM -0500, Brett Stauffer wrote:
>  
> > Sounds crazy, but put it in the freezer for a couple hours, then try it
> > again.  This will sometimes buy you enough time to get the data off of
> > the drive.
> >
> 
> I knew somebody would say this.  This is not good.  

It was a dead drive, so there was little to lose.  
It worked for me.  

> It is better to buy the electronic freeze spray

But how do you know which thing to freeze?  
If it is inside, how can you freeze it?  

On Tue, 01 Feb 2005 12:50:05 -0500 Brett Stauffer <brett at bnbstauffer.net> wrote:

> In any case, buy 2 replacement drives, not one.  
> use rsync

Amen.  

Hard drives are cheaper than backup tapes these days.  
Old PCs to put the hard drives in are free.  

Other folks talked about swapping parts between different drives.  
The insane density of today's drives preclude opening them up.  
Yup, I've opened up old drives in the past and they worked OK 
afterward, but the bits were so big then and so small now.  
Also, the controller board can have calibration info that is 
matched to the platters and heads, so swapping controllers might 
be crazy.  Of course, if freezing doesn't work, you could try it.  
After all, it's a dead drive with little left to lose.