Norton Ghost 9.0: Reliability of Hot-Imaging
The following letter is from Dan Goodell (who has a great article on understanding multi-booting) and the Dell Utility Partition). I asked his opinion of Ghost 9 and hot-imaging: the (controversial) process of imaging a live operating system *from* a 'live' operating system.
He's been using imaging/cloning products before I knew they existed, and uses them frequently (professionally). He also has much experience with Drive Image, upon which Ghost 9.0 is based.
I think you'll find his comments interesting. He stipulated they represent mostly opinion (though an informed one). Posted with permission:
________________________________________
Hi Rad.
I'm not ready to accept hot-imaging as reliable, but I also think it's too early to judge Ghost 9 until it builds more of a track record. Yet, since it's based on the clearly inadequate Drive Image 7, I'm not holding my breath.
I do take issue, though, with readers who argue Ghost version 9 is a mature product, just because the name has a good reputation. Ghost 9 is a rebranding of a PowerQuest product and bears no resemblance to Ghost 2003 other than the name on the box. It would be more appropriately called "Windows Ghost version 1".
I have no experience with Corporate Edition of Ghost. The bulk of my experience is with PowerQuest products, and PQ never had a product in that in-between tier. (DeployCenter came close, but not very.) As I see it, with the purchase of PowerQuest, Symantec now provides three product tiers:
• Ghost 9 (PowerQuest DriveImage 7): for home users
• Ghost 8 Corporate Edition (Original Ghost, with all the bells & whistles): for Corporate users
• LiveState Recovery (PowerQuest v2i Protector): for Enterprise users
Historically, PowerQuest was the original developer of imaging/cloning technology. DriveImage was DOS-based and Ghost was subsequently developed as a competing product. For much of their lives, Ghost & DriveImage competed directly, with each offering similar features.
Product life-cycles ended with DI 2002 and Ghost 2003. DI provided a cleaner GUI interface, and was marginally easier to use. Ghost's GUI was not as clean--in large part, I believe, because of a less-optimal DOS-video mode chosen by its designers.
DI's ease-of-use advantage was, in part, a byproduct of its limited features. Ghost had more bells and whistles, but versatility begets more choices, which the user may need to contend with.
Both were reliable as DOS products and established admirable track records. Along the way, both products acquired tacked-on Windows shells--probably just to counter the public's negativity toward DOS. And, again very similarly, neither Windows interface was particularly robust. Each had a Rube-Goldberg feel.
________________________________________
By the time DI2002 hit the market, PowerQuest was already well underway with development of Windows-based imaging, which resulted in DriveImage 7 (for home users) and v2i Protector (for enterprise users).
DI7 only worked on XP, so PQ tossed DI2002 into the box to support legacy versions of Windows. (I'm not aware if Norton ever developed an enterprise-level companion to Ghost.) Banking their future on the Windows-based versions, PQ allowed DI2002 to languish and abdicated "King-of-the-Hill" status to Ghost 2003. Then, when DI7 stumbled in the marketplace, PQ became buyout fodder.
Despite the similarity of DI2002 and Ghost 2003, it was DI7 that PowerQuest was marketing against Ghost 2003. In that matchup, Ghost was regarded as the more reliable product, obscuring the fact that DI2002 was as reliable as Ghost 2003 when used from DOS. However, all that's moot now, as DI2002 is no longer available and left no successor.
When Symantec bought PowerQuest, they gained DI2002, DI7, and v2i Protector. DI2002 was tossed because Symantec already had Ghost 2003. DI7 was reworked--how extensively, I' not sure--and remarketed as Ghost 9. (FTR, I see Ghost 9 purportedly supports Win2K Pro, whereas I don't believe DI7 did.)
To support legacy Windows versions, Ghost 2003 was tossed into the box rather than DI2002. PQ's v2i Protector was probably also reworked to some degree, and it is now being remarketed by Symantec as LiveState Recovery.
Note that Ghost Corporate Edition is not targeted at big corporations with big-shot IT professionals. Rather, they're marketing LiveState Recovery (~$800 per license) to those users. It's possible the Corporate Edition may have originally been intended for enterprise targets, but with the purchase of v2i Protector, Symantec may have repositioned Ghost's "Corporate Edition" as a middle tier. So calling it "Corporate Edition" seems misleading.
Based on info online, Ghost 8 Corporate appears to be based on v2003. But I doubt that's because Symantec feels v9 is too unreliable for corporate users. Rather, I suspect it's due to the fact that Ghost never developed a Windows-based version, and the Corporate Edition was already under development with the only product they had: v2003, as its starting point.
It appears Corporate Edition is essentially Ghost 2003 with volume pricing and tweaks targeted at network users, like better network and remote management features. I suspect these are largely enhancements to the Windows shell, so I'm skeptical whether underneath it's really more
reliable than the Windows shell of v2003.
As for this middle-tier, I'm inclined to think that IT staffers using Ghost Corporate may experience somewhat better reliability than the accumulated body of home users. Businesses rarely test the range of Ghost's capabilities. They often settle on one tried-n-true backup/restore method company-wide. Once they find & proof a method that fits their needs, they tend to stick with it and avoid changes.
They're unlikely to use multiple partitions, resizing partitions on-the-fly, and other complications home users are famous for. It's plausible both v2003 and v9 work equally well here. I think it's just a matter of Symantec getting around to it before there's a "Ghost 9 Corporate Edition".
I doubt Symantec feels Windows-based hot-imaging is less reliable than v2003's DOS-based method. If Symantec surveyed their support personnel, I expect the comparison would be: Ghost 9 support issues vs. Ghost 2003 *Windows* support issues, and in that light v9 probably comes off no less reliable. Plus, the added benefit to Symantec of product activation would overshadow minor reliability differences anyway. In other words, they feel it's not really a matter of whether hot-imaging is more/less reliable than v2003's DOS-based imaging, but simply: "Is it reliable enough?".
Now, what about LiveState Recovery? I've gone back and reread several PowerQuest press releases about v2i Protector, and I'm struck that there were clearly three marketing points:
1. Overstretched/downsized IT staffs--so ease-of-use (punch a button and go) and time-savings (not having to boot to DOS or find image CDs) were paramount goals;
2. Remote management--to relieve the IT person from having to physically visit each workstation;
3. Minimal downtime--in a big corporation you don't want your server offline at all, so hot-imaging is desirable and booting to DOS to make an image is out of the question.
Reliability was never raised as a marketing point. With both v2i and DI, the attitude was simply, "Of course it's reliable", with never an intimation that either method could be more or less reliable than the other.
In other words, v2i stayed in Windows not because they believed that was more reliable, but because of other factors that were considered worth the monumental increase in complexity. (And $800 per seat goes a long way to making the complexity worth it.) I doubt any of that has changed under Symantec's stewardship.
________________________________________
Back when they were still under the PowerQuest banner, DI7 and the Windows shell of DI2002 refused to work on server OS's (XP Server & 2000 Server). It would generate a message that said a server OS required v2i Protector.
I've always suspected that wasn't true. I suspect it was a Windows limitation, not a limitation of the imaging technology. My feeling is that sectors are sectors, and a server OS does not use a different file system than non-server versions of Windows.
From DOS, a server version should look no different to an imaging/coning program than any other version. I don't have a server OS to test. Most techs who manage servers accepted PQ's word and bought v2i instead.
Over the years, I came across two people who used DI2002 from DOS on their server OS. Both claimed success. A sample of two prove nothing (who knows how complex their configurations were?). But it suggests I may be correct.
I suspect the real reason PQ insisted on v2i for servers was because servers are not intended to be offline (as DI2002 would require) and are too busy for DI7--handling things like CALs and domain controllers while live were beyond the ability of DI7.
I don't know Ghost 9's policy on server OS's, but I see server OS's are not listed as supported. I expect Symantec to claim LiveState Recovery is required. I have a hunch however, Ghost 2003 from DOS would be successful imaging a server OS.
Anyway ... my two cents worth.
Dan
The END.
