I had the unprecedented luxury of almost completely shutting down over the holiday season. OK, I peeked at my CrackBerry a few times, but -- mentally -- it was a total shutdown. That's a good thing to do every so often, I think.
One of things I did think a lot about was this post -- what's going to happen in 2008?
Hopefully, I can keep my accuracy at something close to last year's levels ...
Organizing the Discussion
There's a lot to talk about that's coming in 2008, more than you might think.
To keep things from wandering, I've decided to organize my thoughts into three categories: technologies (a perennial favorite), usage models (how people actually use the technology) and industry (how the landscape will change).
Not that anyone needs any encouragement, but please -- feel free to add your own thoughts to mine.
Technology Trends in 2008
First, we'll see an even wider range of service level options for storing information. There will be even higher highs and lower lows when it comes to tradeoffs between service levels and media costs.
1TB drives are in the market today (albeit in limited fashion), they'll become widely used in many enterprises. Data deduplication in all of its sundry forms (actually another service level option) will find its way into all sorts of storage use cases.
At the other end of the spectrum, we'll see flash memory and larger storage caches become popular again. Yes, these approaches will be relatively expensive, but -- despite the costs -- they'll join the spectrum of service level options for storing information.
Unlike many in the popular media, I don't think it's productive to predict technology "winners" and "losers" -- nothing is going away (even tape!), we'll just have new options to consider for different parts of the storage portfolio.
Second, I think the storage networking landscape will be largely unchanged in 2008. FC protocols will continue to be popular in environments that started that way; iSCSI will be popular in newer SANs but won't cannibalize FC SANs, and FCoE will be in a tire-kicking period during 2008.
But I think there's going to be a new market: as we see more interest in Web 2.0-type information (active content), and cloud-oriented information infrastructures, the high-end scalable NAS discussion will re-ignite.
This time it won't be about supercomputing or render farms, it'll be about distributing information at internet scale.This type of storage enviroment is so different than what we've seen in the past, we'll probably need new words and concepts to describe it ... it doesn't do it justice to just call it "SAN" or "NAS" anymore.
Encryption of storage media will start to become more of a non-issue for many customers: not only has the paranoia thankfully peaked, but -- more importantly -- the encryption technology will now be simply part of the fabric (HBA, switch, array, IO driver, etc.) and there will be decent storage-oriented key management tools available.
In a continuation of what we saw begin in 2007, I think there will continue to be strong interest in putting storage functionality in the network, as opposed to either server or array. This isn't just storage virtualization, this list includes encryption, archiving, replication, etc. -- more and more people will want to separate their functionality decision from their storage decision.
Usage Model Trends in 2008
The headline story here has to be SaaS (software as a service).
Now, I think this particular industry moniker does a poor job in describing what I think will be interesting in this space -- it'll be more about information management service providers (aka IMSPs) who will offer to backup your data, manage security event information, records retention -- all the nasty information management housekeeping challenges that have to be done, but no one really wants to do.
This means that customers will start to seriously considering the tradeoffs between buying the technology and running it themselves, or contracting for a service. And I'm sure much will be written in the popular press on do's and don'ts, e.g. "Ten Things You Should Know About SaaS For Storage".
Behind this, we'll probably see a new raft of entrants into this business. And they'll all be faced with the same two challenges: one business-oriented, one technology-focused.
On the business front, you're convincing a customer to trust a service provider with their most precious asset -- their information. I think larger, more established companies will be generally be more successful in winning this argument rather than the smaller and more aggressive outfits.
On the technology front, there's the architectural challenge of multi-tenancy: building environments that can be easily shared as a service.
Building environments for multi-tenancy is harder than you might think. And it's not something that can just be added to an existing product with traditional usage models. All forms of resource and information management are affected by this (not just storage), and -- of course -- the providers of multi-tenancy services will want all the pieces to work together with a minimum of fuss.
There's no good, broad solutions out there today for this, but I think we'll see some activity begin during 2008.
Tiering of information (aka ILM) will take on a renewed interest during 2008, largely driven by all the new options for storing information at different service levels. Yes, you'll be interested in terabyte disk drives and aggressive data deduplication, but you'll have to figure out what goes where when, won't you?
Already, some people have re-learned some old lessons with large stores of video files -- remember, they're already compressed, and thus aren't amenable to any form of additional compression. Or that data deduplication extracts a performance penalty -- there's no free lunch.
If all the customer interest in 2007 was around new features (e.g. thin provisioning, data dedupe, et. al.) I think most of the interest in 2008 will be around integration of these features so that they just work as expected, and aren't a "special case".
Thin provisioning, as an example, is of marginal use unless you can use it with everything else you've got, e.g. replication, storage management -- and it all just works as expected. Same for data deduplication, or QoS (quality of service) management, encryption, replication, etc.
As an example, I have a pretty integrated home theater system. It works the way I want it to. A nice Harmony remote with big buttons that say "watch TV" or "play movie". I don't know about you, but I'm not taking the plunge on HD-DVD or Blu-Ray simply because I'll have to upgrade a bunch of components, plus spend a lot of time figuring it all out. And 1080p movies just aren't worth it to me -- yet.
But there's one new bit of technology that everyone will want integration with, and that's VMware.
For many customers, Vmware isn't just another add-on to their existing portfolio of operating environments, it's the target of their next-gen IT architecture -- as it should be.
During 2007, I saw many customers get more interested and sophisticated around the whole discussion of "optimized for VMware" rather than simply "works with VMware". I think it's a safe prediction that we'll see much, much more of that in 2008.
And, finally, I'm betting that we'll see the nascent "information governance" discussion move front-and-center for many customers. They'll realize they're sitting on a pile of information that can be used to make money, save money, or get them in a whole lot of trouble.
And they'll want a formal governance process to wrestle with the tradeoffs (or at least get more fingerprints on the answers!)
The Industry Trends in 2008
First, I think that server vendors have re-awakened to the fact that storage isn't just a peripheral anymore.
Maybe it's the server that's become the peripheral to the information?
We can speculate on how successful they'll be, but consider (1) Sun's continued noise around storage (including many re-orgs), (2) Dell's expensive acquisition of EqualLogic, and (3) IBM's recent (and expensive!) acquisition of XiV, which is an interesting story in itself.
To me, it's pretty clear what's happening here. Processing has become commoditized, information isn't. And, if you're a server vendor, you better have a storage play that isn't purely reselling other people's stuff. The only other major vendor that hasn't checked in is HP -- anyone want to speculate here?
The high prices being paid will probably encourage a handful of new storage startups, probably focused on Web-2.0-ish active content, probably in a multi-tenancy model. The existing gaggle of startups (3PAR, Isilon, Pillar, et. al.) will probably either have to demonstrate success, get bought, or fold during 2008.
I've got my bets on what will happen to each -- do you?
I think technology vendors will have to start to learn the new "triple play" in 2008: sell your products traditionally, offer them as a service, or sell them to people who offer them as a service. Trust me, this won't be easy -- each is a distinctly different business -- and no one has it all figured out yet.
Tech consultancies will probably start talking to customers about their information, rather than just their technology, or their applications. The smart ones will probably position themselves as information-based analogs of their "Big 6" finance counterparts.
No discussion of the industry would be complete without a few thoughts about the "industry watchers": the press, the analysts, the freestyling bloggers, the curmudgeons, etc. A few years back, anyone could write about just about anything and seem to find an audience. Most of what I've seen isn't worth the (virtual) paper it's printed on, but the occasional gems do stand out.
Now, I think we live in a much more discerning world, at least as far as this industry is concerned.
Yes, we'll still have our fair share of info-smog in our industry, but I think the bright, insightful voices will have even more impact in 2008, and the remainder, hopefully less impact. Quality matters in this world, and the gap between the best and the rest will continue to widen -- and, for those of you doing quality work, my personal thanks go out to you!
The Bottom Line
Like every new year, there's a lot of macroeconomic turmoil in the air: the sub-prime credit crunch, the fact that it's an election year in the US, the rising price for oil and other commodities, and so on.
But for me, that's only surface chop on a very deep ocean. There are powerful currents at work down below that don't fluctuate much day to day, or year to year.
We've become an information society: information now powers our economies and our personal lives -- and there's no going back. And each of us is coming to terms what this new world means to us -- in our jobs and careers, at home and in our personal lives.
There's going to be even more information to store, protect, optimize and leverage in 2008. About 60% more than last year, if history continues.
And that's a safe prediction for 2008.
Recent Comments