In my backup program should I click compress contents to save disk space?
And encrypt contents to secure data? Why wouldn't I always want to do both?
Compression is in general an easy choice these days - turn it on unless you
run into issues I'll talk about in a moment.
Encryption requires a little more thought. It boils down to a risk/risk
tradeoff - the risk of your backup being compromised, versus the risk of not
being able to get back into it yourself.
]]>
Compression, as you probably already know, is a mathematical operation that takes data and makes it smaller by removing redundancy and repeated patterns. Compressed data can be uncompressed to restore it to its original form.
Backups can be big. Compression makes things smaller. Seems like an obvious choice, right?
And most of the time it is. There are two things to consider that might occasionally lead you to not use compression:
-
Speed: compression can be a CPU-intensive task. Processors are typically much more powerful than needed to perform compression faster than the hard disks can read or write. Usually. If you have an older, slower machine, or if you find that your machine bogs down excessively as you try to use it during a backup, you might turn of compression to reduce the impact on your machine.
-
Corruption: this varies a great deal depending on the specific algorithms used by your backup software, but if, for example, you have a bad sector in the middle of your backup, an uncompressed backup is typically slightly more likely to be recoverable than a compressed on. Perhaps if you're backing up onto unreliable media you might want to turn off compression (but then ... why are you doing something as important as backing up to unreliable media?)
As I said, it's often a pretty simple decision. I even turn my backup program's compression up from it's default level compression to one that will take a little longer, but make the results a little smaller.
Encryption is the process of scrambling your data using a secret password, pass phrase or other encryption key such that the encrypted data is effectively meaningless. Only with the appropriate decryption key can you recover the original information.
Encrypting a backup makes a lot of sense if what you're backing up contains sensitive information, and the backups themselves may end up being accessible to untrusted parties.
The risk of encryption is such a simple one, and yet I see it every day over and over and over again, ad nauseam.
People forget their passwords.
And with any good encryption or backup software, if you don't have the password, you're not getting at the data.
Backups are a tad riskier, in my opinion, than your average free email account when it comes to losing your password. Think about it: you make an encrypted backup, put it on a shelf somewhere (so to speak), and then some period of time later you decide you want to access something from that backup.
Will you remember the password?
You're betting your backup on it.
Now, of course people have strategies and systems and ways of remembering passwords, and that's fantastic. If you're one of those people who has a good approach to not only maintaining secure passwords but remembering them in a month, a year or even much longer, then you'll likely have no problem encrypting a backup.
On the other hand, if you seem to need that "I forgot my password" link on websites a little too often, then you might want to reconsider. There's no password reminder for a backup.
Hi Leo, are you saying that the compression scheme used by Windows is a lossy type of compression?
That would mean data loss wouldn’t it?
I know MP3’s and JPEG’s are lossy compression schemes but I wouldn’t think data would be compressed using any kind of a lossy method.
I would instead expect data compression to be lossless.
I use a Linux based program to create backup images of my OS installations and the program offers varying levels of compression.
The medium setting reduces image size by more than a third while retaining all the data.
If it didn’t, restores would be impossible.
23-Dec-2009
@Frank,
Lossy compression is used to compress image and video files, as much of the data is not needed at all. For example, if you have a video at 60 frames per second, then it can be compressed down to 24 frames per second as the human brain can’t tell the difference between 60 and 24 fps. This is lossy compression as data is deleted.
The compression used on backups is lossless. It checks blocks of data and if it sees a recuring pattern it replaces it with a single instance of that pattern. For example, in a text document, the word “the” may appear 20 times. The backup program takes the word “the” and in place of the other 19 “the”s it will leave a symbol or reference for it so that 60 characters (3×20) can be compressed to 6 for example. When the document is uncompressed, it converts the symbol’s into the word “the”.