CDP: the next generation of backup.

Those of us who have been following the PC software industry since the days before 'NASDAQ' entered the public lexicon may remember the utilities software category. We remember fondly indispensable utilities such as Quarterdeck QEMM386 and Central Point PC Tools. But alas, as technology marched on and Bill Gates' empire grew ever larger, these software tools became absorbed into the basic operating environments PC users work in. The purpose they had served ceased to become a separate function; memory and disk management became an extension of computing itself.

Sentimentality aside, we can generally agree that this has been a good thing. PC users don't want or need to think about disk or memory management; they just want to surf the net in peace. In fact the shifting demands of PC users have accelerated the pace of this absorption. It's easy, in fact, to identify modern and more socially-relevant examples of this ongoing enthalpy. Anti-virus software, today a large and profitable software category, is clearly destined to join Hyper Terminal in the software graveyard that is the "accessories" group.

Can an exploration of the evolution of the anti-virus software category provide us with insight into "the next anti-virus"? The path of technical evolution is quite revealing. Anti-virus applications began as standalone applications that actively scanned your computer for these little nasties, on demand. The process was slow and cumbersome, and out of pure necessity the convenience of "scheduling" became the must-have of the anti-virus category.

But technology did not stop there. The strategy of anti-virus software shifted from one of passivity to one of activity. Rather than focusing on identification and removal of viruses after infection, anti-virus software providers sought to prevent virus attacks in the first place. This led to the concept of anti-virus "blockers" which originally were terminate-and-stay-resident (TSR) applications. (Who remembers good old "Sidekick"?) These lent themselves well to adaptation in the Windows world, where they became those little icons in the system tray that we are now accustomed to. The end-user was moved out of the equation, and the function of anti-virus "scanning" had now become an extension of computing.

Of course, even today there is no shortage of utility applications that rely on scheduling. Backup software is the most obvious example. Indeed, the concept of the periodic backup predates personal computing itself; Iron Mountain arguably delivered the first backup solution in the mid-1950s when it began burying carbon copies of documents in an abandoned iron mine in upstate New York.

Backup software is similar in many respects to anti-virus software. The purpose of both applications is to protect data and system configurations; users routinely avoided using both applications until they felt the pain by losing data or being attacked by a virus; at the outset, both were relatively difficult applications to use.

However, anti-virus evolved steadily, eventually taking the form that we are familiar with today (and are no doubt running right now), whereas backup has remained a relatively un-evolved concept. And while anti-virus has taken firm root in both corporate and consumer computing, backup software, perhaps owing to its complexity, has largely remained the province of IT administrators. The home user still doesn't back up their data.

Why is this so? Successes with anti-virus establish that even unsophisticated home PC users understand the basic concept of preventative maintenance. Perhaps the inability of backup to break out of IT and into the home has to do with the aforementioned lack of evolution. Today backup software is still a scheduled process, one that essentially runs in the foreground when it is doing its thing. It is still sufficiently invasive and annoying that even backup vendors themselves liken backup to flossing--everyone agrees it's a good thing to do, everyone pretends to do it, but if we fail to do it or do it incorrectly, the consequences can very unpleasant.

The emerging category of continuous data protection (CDP) software may be the first sign that backup software is ready to evolve. There are a host of small vendors marketing CDP software, such as Storactive, Mendocino, Mimosa, Asigra, Revivio, XOSoft, TimeSpring and FilesX. The activity has caught the attention of industry heavyweights such as IBM, Microsoft, Symantec and EMC, each of whom has either announced or indicated future plans for CDP products. As is typically the case, most of these solutions are geared toward servers and aimed IT administrators. However, a few products, most notably Storactive's and IBM's, are designed for laptops and desktops.

CDP itself is a relatively simple concept in theory, though as a practical matter the members of the SNIA CDP Special Interest Group engaged in spirited debate before settling on a definition. To paraphrase their technical definition, SNIA defines CDP as continuous capture of data and arbitrary recovery points.

As applied to desktops and laptops, CDP software resembles anti-virus in terms of how it functions; there is an "auto-protect" mechanism which monitors data for changes, just as there is such a component in anti-virus products which monitors for suspicious activity. Also strikingly similar is the extent to which end-users interact with CDP and anti-virus software. In CDP software, end users do not work with a backup application at all--they instead work with recovery applications. Essentially end users will not interact with CDP backup software unless they actually need to recover something. This is almost identical to the existing antivirus paradigm--users are only aware of the anti-virus agents when they catch a rogue program in the act of doing something nasty.

By removing the end user from the backup process, CDP software overcomes a traditionally formidable obstacle to routine backup: the labor. End users have been the most unreliable component of any backup system. By removing this component from the system, CDP software dramatically improves reliability. As an added benefit, the continuous nature of CDP backup lends itself well to the rapidly changing data created on today's PCs. Shorter data lifecycles demand that the most recent backup be very recent; CDP ensures that there is always a recent copy available.

If the evolution of backup software plays out in a manner similar to how anti-virus played out, the emergence of CDP over the last one-and-a-half years is an indication of things to come. We can expect to see the CDP category make inroads not just in the server backup space, but into areas where backup has failed to resonate with the user--in the home, in the remote office, and on the laptop. We can expect data protection products to become a natural part of modern computing, just as anti-virus had done over the past decade.

Steve Sussman is senior product manager at StorActive Inc., Marina del Rey, CA

www.storactive.com