Tidbits from my daily-life

Backing up is important. People tell us that all the time and there must be some kind of truth in this. Else why would people keep on stressing how important it is?

Well… I always believed in backup, but never put it into practice. Since my first PC computer with a hard-drive inside (aged 10) I have been loosing data on a regular basis. The only reason why this happens less and less frequently is because I run Linux and do not have to re-install my system every other day and also because I have more than one drive, thus spreading the data over several mediums.

During my recent holiday in France I was lucky enough to slightly fry one of my drives. The cause? It was an external drive that suffered from the several power cuts that happened at home. You see, this winter the power consumption was the highest, ever, due to the intense cold and many people switching to electric heating.

Enough background. How am I trying to save the data (obviously essential data. Work, exhibition material, video footage) and how will I try to remedy to this problem in the future.

How to save my drive:

Since the drive is still under warranty, I decided that it was not an option to open the case and plug the drive straight into my computer. Furthermore, the logic of the USB case was not dead so I did not see the point. From the preliminary tests, the issue seemed to be noise and possibly the reading heads hitting a platter while the power cut, thus damaging some sectors. FYI, the drive was formatted in NTFS since I need to use it on my office computer running XP.

First unsuccessful attempts were the following:

• Mounting the drive under Linux and using the cp command to put the files on another drive. The copy would stall at some point and would stay stuck.
• Same as above using rsync
• Same as above using dd_rescue for each of the files
• Then I tried mounting the drive under XP and using Roadkills unstoppablecopier, some of the files would copy, but it would take forever and leave a lot of errors.

After all these trials, I decided that the NTFS partition must have been corrupted and that there is some physical damage to the disc. If I eliminate one of the problems, the recovery should be easier. I took the step and finally did the expense of buying a hard-drive that would have enough capacity for me to image the broken drive, fix the image if need be and then mount the NTFS partition and go from there.
For this, after some reading, I used ddrescue (and not dd_rescue). GNU ddrescue is praised by many and seems to be the way to go. It combines all the good aspects of dd_rescue and dd_rhelp and where all the others have failed so far, it seems to be doing quite good.

The current ETA is: 227663MB copied at an average of 3438kB/s and still counting.
The partition that is being copied is about 300gig.
More update on this later.

How to prevent the same thing from happening

I have a total of 5 systems currently:

• “home” computer / multi-purpose machine / home server (ubuntu studio)

• SATA II 250GB – system disk
• SATA II 320GB – home folder
• SATA II 250GB – misc documents
• SATA II 250+250+250GB raid 5 array – backup of important documents

• IDE 80GB – system and data

• IDE 30GB – system and data (on different partitions)

• SATA II 160GB – 80gb for data and 80gb for system
• FW External drive 250GB – for work files, video files, office-related data (the data which i’m trying to recover)

• SATA II 160GB

The new drive I bought is a E-SATA/USB2.0 1TB drive.

My new plan for backing up is to use the 1TB drive to make backups of the most important files on each of the systems. The Raid array, since it can be easily reconstructed, will also be holding some of the most important files. The really important files will be burned to DVD and spread over several physical machines so as to make sure that there is also a working copy in at least one place.
I also have some online server space. I will make sure that this will also be put to use for some of the important files.

Further plans:
Frequent check of the SMART status of the drives.
Frequent backup
Try and add more space to RAID array so that more data can be kept safe. (I know that RAID is not a replacement for a good backup, but it sure does help keeping the data safer than having just a single point of failure. Also, the RAID is software so it can be reconstructed easily if the motherboard came to die.)