Using Knoppix for data retrieval

Yesterday I was given the dubious task of data retrieval. A client’s laptop had a faulty hard drive that contained much needed data. How did I know the hard drive was faulty? Well it went “brooomm guu conk” when I tiled the notebook on its side.

Windows wouldn’t load so on went Knoppix, a bootable live Linux distro.

The first task was to mount the offending hard drive, hoping that it was still in a working condition.

mount /dev/hda1 /mnt/hda1

Once mounted, I could snoop around the hda1 device and suss out what needed to be copied. The only question left was what’s the best transfer method? I tried FTP but found it to be slow, and with a corrupted hard drive, it often timed out with “input output errors”. I needed something that would auto retry and also resume without complaining about errors or incomplete fragments.

Jonesy suggested I start up apache, serve the hda1 folder as its Document root and wget -r the bastard. This worked fine at first, until it started timing out and reporting more “input output errors”. It had been stuck on a 200kb file for the last 20mins.

New tactic, the fridge.

I disconnected the laptops hard drive and popped it in the fridge for half an hour. It managed to stay there untouched and unnoticed by co-workers who were continuously opening the fridge for milk and food.

Popped the hard drive back in and THIS time I gave SCP a try with the -r option.
Bing batta bing, a flawless copy of all the critical files I needed.

Fridge + SCP + Knoppix = all you will ever need for simple data retrieval.

No related posts.

2 Responses to “Using Knoppix for data retrieval”

1. Jonesy says:

   May 11, 2006 at 11:06 am

   Who’s idea was it to whack the drive in the freezer?
2. Jamsi says:

   May 11, 2006 at 1:50 pm

   Mine !