Continuing on About Backing
Up"> One colleague makes a suggestion youll also find in the PCMag story: separate your code and data. I sort of do this already in the sense that I use the Windows My Documents folder as a root for all my data that I consciously save. I could go further and allocate a separate disk partition for data and back up that partition. This is, by the way, basically the Linux/UNIX way to do things. But in Windows there is other data, and you cant easily stick it on another partition. For instance, a lot of configuration information and data, like your Mozilla bookmarks, are stored in the users profile folder off the Documents and Settings folder. Its possible to change the location for this folder but its easy not to. And its still possible (although a bad idea) for programs to store data in other fixed locations, like the Windows directory. Yes, doing all this makes it easier to do a complete data backup, but its still not a complete system backup since it lacks applications, not to mention Windows itself.Other readers recommended what are essentially mirroring systems: two hard disks, keeping mirror images of each other, so if one fails the other can take right over. But this is only useful for hard disk failure; if you accidentally delete a file, it gets deleted on both disks. You still need a real backup system. I have also learned that NAS systems are even cheaper than I thought, and there are NAS enclosures for well under $100. This seems like the future of the low-end to me. The price of these enclosures will certainly drop even further, and then no home network should be without them. Unlike the mixed online-offline and separating code and data, I like the idea of backup being something that happens automatically and mindlessly, and I only have to think about it when a disaster strikes, and then its there for me. Thats when I know Ill do it right. For me the really big open question is testing: You can never really be confident of a disaster recovery plan unless you test it, and that means testing a restore of your backup. Its one thing to test a restore of your data, and you can do that to a test location, but testing a full system restore is a risky and time-consuming thing to do. Heres one way you can do it: Buy yourself that new, larger hard disk you know you need anyway. Test your backup by restoring to it, as if your current hard disk crashed. When youre done you can put your old hard disk in an enclosure and back up to it. Its still a bit of a shell game; where had you been storing that backup you just restored? But the security from having a disaster recovery plan in which you have confidence is profound, the kind that really makes you feel better. Its worth going to some trouble and cost. Security Center Editor Larry Seltzer has worked in and written about the computer industry since 1983. He can be reached at firstname.lastname@example.org. Check out eWEEK.coms for the latest security news, reviews and analysis. And for insights on security coverage around the Web, take a look at eWEEK.com Security Center Editor Larry Seltzers Weblog.