This is why I prefer Symantec (Norton) Ghost 8 for my full system backups. Ghost 8 runs in DOS. That means that Windows isn’t running when you use Ghost 8 to back up, which in turn vastly increases the reliability of the backup. I’ve restored machines from Ghost backups many times with complete success. The only hitch was that it was designed to be run from floppy disks, and modern laptops rarely have floppy drives. Oh, and it didn’t always want to recognize USB or FireWire external drives, though I had no problems with my FireWire XHD on my previous laptop. (My current laptop doesn’t have FireWire.)
When Ghost 9 came out after Symantec bought DriveImage, I didn’t upgrade. DriveImage had a good reputation, and I recommended it to clients without floppy drives (prior to the new Bart-PE CD which runs Ghost 8 but is, shall we say, somewhat ethically complicated), but I didn’t want to use a product which tried to back Windows up from within Windows, and that’s what Ghost 9 is. (The Ur-Guru wrote me paragraphs of outrage when he tested it.)
Why am I bringing up all this history? Because IntroAnalytic just released a new study involving DriveImage (the pre-Symantec product), Norton Ghost 9 (whether Ghost is billed as “Norton” or “Symantec” seems to depend on whether it’s the consumer or the corporate version), Acronis TrueImage 7, and Microsoft System Restore. System Restore (which has actually been useful to me in the past) rated the lowest: IntroAnalytic gave it 0%. But of the third-party products, Ghost 9 came out at the bottom, with only a 75% chance of successful recovery from a major computer error (non-functioning OS). Drive Image came out on top, at 90%.
They didn’t test Ghost 8, presumably because their aim was to compare products which operate from within Windows and work with external hard drives. They also didn’t test Dantz Retrospect, but I found Retrospect decidedly underwhelming anyway.
Where does this leave someone just embarking on a backup plan? In an uncomfortable position, given that Drive Image and Ghost 8 are no longer sold by the manufacturers. In order to get the most reliable software for making a full system backup, you have to go hunting around on eBay and other places where used software is sold. If you can’t use diskettes or just need to be sure Ghost 8 will recognize your external drive, you then have to create a Bart-PE CD (or get someone like the Ur-Guru to create it for you).
And remember—whatever backup software you use, you need to test your backups. If you can’t bear to take the plunge and do a test-restore of your whole machine, try restoring one or two files. This will at least assure you that the image is not corrupt. (One drawback of Ghost is that if you span a Ghost image over CDs or DVDs, a single damaged disk renders the entire image unrecoverable. I learned this the very hard way.)
Read IntroAnalytic’s press release at http://www.introanalytic.com/BackupResearchMediaRelease.pdf
The DM Review editorial staff reprises the press release: http://www.dmreview.com/article_sub.cfm?articleId=1049420
Comments on the study from Chris Mellor at TechWorld: http://www.techworld.com/storage/blogs/index.cfm?blogid=3&entryID=152
This is a very valuable info. I have been using Ghost 2003 and found it to be bullet proof. When you buy later verisons of Ghost, 2003 version is included. The issue of reliability of later versions of Ghost is extensively discussed in the User Guide available on the Internet. Thanks.
You might be interested to know that the Infoanalystics “review” tested a two-version-old copy of Acronis True Image (ver. 7) vs. the just-replaced Ghost 9. However, I’ve not been able to find out much about this “research firm.” Makes me question the validity of the review and the folks who did it.Seems to me that if you plan to test products, use current versions and test on equal footing — don’t test products that are so old they’ve been replaced —- TWICE.
Sometimes the older products are the better ones. I would agree that it would make more sense to compare versions of the same vintage, even if the final conclusion is that one of the older products is better than any of the newest ones. However, InfoAnalytic clearly disclosed which products they were testing.When the Ur-Guru tested Ghost 9 extensively (he has enough machines and VMs to test pretty well everything) and pronounced that it sucked and I should stick to Ghost 8, I believed him. Apart from Ghost 8, I can’t think of a single Norton/Symantec product that I’d recommend, so I was fairly well horrified when they bought PowerQuest. Not unexpectedly, they managed to ruin two good products by combining them. But if you’ve read my posts about TrueImage and Retrospect, you’ll know it’s not just Symantec’s products I have a problem with. If it’s trying to copy Windows from within Windows, there’s going to be trouble. That’s the fault of Windows, not the software manufacturers.