There’s been a lot of discussion lately about what exactly is a “backup” of data, and how does a “copy” differ from a “backup”. Channel Register posted a recap of the discussion in, “Storage world asks: Is a copy a backup?”
There’s a lot of nitpicking going on, but some key points are:
- A backup resides on a different device than the original data
- Backup management is done by an application (as the manual process of sorting thousands of files is unwieldy)
- At some point, the backup must be moved off site
- Backups are to be used for the restoration of data, and nothing else
- Backups must allow for recovery at a specific point in time
P. Lee commented on Channel Register’s story by saying, A backup is defined by intent, not format or medium.
Well said, P. Lee. For instance, taking your child’s favorite DVD and making a backup of it (while still illegal) has a much different intent than making a copies of it and handing them out.
Following P. Lee’s argument, Joshua Goodall chimes in: what matters isn’t that a copy is made. What matters is what you want to do with it afterwards.
This also follows the rationale that once a copy of the data is made, it should be organized and saved to another device, preferably off site.
Part of the discussion about backups revolves around the format of the backup. Some argue it should be in a different format, while others say the same format. Why does this matter? Well long ago, when storage wasn’t that affordable, Unix systems packaged up data and compressed it before backing it up for efficiency and storage considerations. Carter George recognizes, Great for backup vendors, not so great for users.
Scott Waterhouse, an employee of EMC (Mozy’s parent company) concludes,
You can have a backup only. You can have a backup and a copy. But you can’t have a copy only.
So are you backing up your data, or just making a copy?