Over the past few years, I’ve been working to rid myself of as much paper as possible.
As part of this, I’ve been scanning/shredding lots of sensitive financial documents.
To protect against data loss, I’ve been using an online backup service. And for file security, I’ve been relying on TrueCrypt, which is a piece of open source software that lets you create encrypted volumes.
The big concern with storing everything in an encrypted archive is that there is now a single point of failure. What if the archive gets corrupted?
According to the TrueCrypt FAQ, one corrupted bit typically causes loss of an entire block of encrypted data. The good news is that TrueCrypt stores your data in 16 byte (128 bit) blocks and corruption of one block doesn’t affect others. Thus, if data corruption occurs, the damage is limited.
Of course, if the file header gets corrupted, it may be impossible to open your archive. Here again, there’s good news… TrueCrypt allows you to backup the volume header (under “Tools > Backup Volume Header”) so, with a bit of forethought, a corrupted header is little more than a bump in the road.
But still, this all makes me a bit nervous. Thus, I pinged my pal Mike Piper, who likewise uses TrueCrypt to store sensitive financial documents.
Mike’s excellent suggestion was to simply keep multiple versions of your data archive. One way you might do this would be to duplicate your archive before you add data. As long as you don’t have huge archives, this is a straightforward solution.
Here’s how it might work:
Let’s say you have an archive called “important_data”. Whenever you open it to add data, duplicate it and append the date. So today, for example, I would name the new file “important_data_2012_12_03″. And when I add more data next week, I would create a new file called “important_data_2012_12_10″.
In other words, you’re keeping versioned backups. If one fails and you’re not able to restore from your online backup for whatever reason, you’ll only lose a small increment of data — whatever you added since the previous update.
For added safety, I suggest closing and re-opening the current and next most recent versions of your archive whenever you add data. That way you can be sure that everything is working as intended before you walk away.
All that’s left is to decide on a reasonable number of prior versions to keep, and then get in the habit of deleting the oldest whenever you create a new one. Easy peasy.
Relevant queries: archiving files data loss.