 Okay, we're back. This is Dave Vellante at Wikibon.org. We are live VMworld 2011 here in Vegas, Sin City with Two of my squeaky clean colleagues Brian Reagan Steve Kenes. You say that after saying sin city coming on. You know, we don't have time to enjoy Vegas I know running around like crazy. Absolutely right. I'll work welcome back I think this is my like hundredth time in Vegas this year. I don't know why but everybody's having shows in Vegas So we'll take it VMworld 19,000 people Quite an event. We had Rick Jackson on we're talking off camera next year. It's moving back to San Francisco, which is Definitely a change of pace, but Brian back at IBM great to see you as an executive in a cloud role Absolutely. How's that going? Great to be back here at Wikibon and and silicon angle and and the cube. Yes, it's it's 30 days in learning a lot Really trying to understand And the reason we're here at VMware is really trying to understand what this inflection point means is more Mission critical applications become virtualized. What is that? What are the implications in terms of disaster recovery business continuity? I was encouraged to hear Paul Moritz talk about how DR is one of the most natural use cases for the cloud And we couldn't agree more so that's that's the business that I'm focused on in running right now And Steve you are a storage efficiency evangelist at IBM. So tell me what that means so Essentially what I go out and do is I talk about how storage efficiency both for IBM and for the end user is really affecting them What are the key technologies? What are the trends? What should they expect to see from IBM? Where's it coming? What are the different products? It'll be either embedded in or how can they acquire it? Where is it going? You know so Brian and I do some work together about talking about as we make storage efficient in the cloud How important is that? How do you sell it? How do you promote it? That sort of thing So Brian, I actually think you're in a great spot. So your scope is a disaster recovery archiving as well archiving as well That's correct. Okay. You're in really good shape. Here's why Everybody is talking about how your applications for the cloud are cloud 1.0 That means it's becoming real. There's money to be made here, right? We actually we had Sikhi Jinta on and She's with a competitor at CSC and she's saying look, you know, it's a lot of hype out there But the real stuff is happening, you know in You know these day-to-day applications absolutely are in archiving are right there So what are you guys seeing in the customer base in terms of adoption cloud adoption specifically in those areas? well, it's I would break it down into two main areas one is as In terms of traditional DR the the traditional model of I have applications sitting on infrastructure at my primary location I need a secondary site so in case Irene or other natural events occur or if it's just a natural downtime event which can occur from you know Data integrity problems or human error and the like we all know the the reasons why systems and applications go offline People are naturally starting to evaluate how the cloud can Change the dynamics of that disaster recovery model for the better cost will come down naturally And maybe they'll get more agility out of their infrastructure Maybe they're they're 24-hour RTO and RPO can come down to six hours if the data is persisted in the cloud as opposed to just at a Secondary physical site, so that's sort of use case number one Use case number two, which is related is very often in the traditional DR model Tape was the the media of choice We would back up our systems images back up our data put them on trucks ship them to the Recovery site and hope we never had to use them And very often those tapes would be kept for a very long period of time Fast forward to the disk-based world of storage and all of a sudden what we're talking about is An archive use case that first really serves a storage optimization purpose But ultimately as we start layering and more information management Governance compliance search and discovery capabilities on top of that That obviously expands and extends the value that we can offer to our clients as they preserve that data for long periods of time So I want to drill into the cost a little bit and then I come back to the optimization piece cloud is rental Right and rental is almost always. Maybe maybe always more expensive than than owning. Why is cloud less expensive? Well, I think it comes down to very similar to a traditional managed service model where yes, the physical the capital the capital cost is more expensive to rent But the operating expense for rental I'm essentially nearly eliminating the operating expense and as data grows and as the Information environment becomes more complex those opera op-ex costs go up So what we're really doing is we're neutralizing or minimizing or almost eliminating the op-ex Well, perhaps seeing a slight bump in the cab-ex. So if I could translate that a cloud service provider Which you are is going to do be more efficient at managing infrastructure be able to pass those savings on to the customer The point I've been making a lot today Is every customer we talked to? Almost everyone anyway Budgets are flat We're asked to do more with last headcount is down is does that describe your business that? If it's going to the roof yeah Absolutely if our CFO we're here you'd be talking about how we need to continue to take costs out of our environment Well, you know why efficiency is absolutely vital to our business, but you're growing right at the point Oh, yes, absolutely. So absolutely and so the innovation is the cloud service providers are innovating They're pouring money into that innovation Whereas the traditional IT shops are like, okay, keep it flat cut it. It's not our business, right? And so now that brings me to optimization. What's Steve? What's the intersection between what we're talking about here in cloud services and optimization? Is it fundamental? Is it a nice to have is it a one-time hit give us give us some insight there I think Brian nailed it on the head in you two talking about the economics of it All right if I'm going to continue to get hammered on on the economics of running my complete infrastructure And I'm going to sell to the CFO that I'm going to outsource this And I've already sold the fact of the matter that I can actually take the operational costs and extract it from from my My environment right now. What am I doing? Okay the next question from any CFO? I know is what's next right? So if I can compress it locally or if I can shrink it locally or deduple locally or provision it better Locally and become more efficient and then whatever I do in the cloud I can get a better dollar per terabyte rate because I'm using less physical space right all of a sudden I'm really now having a much deeper impact on that financial budget at the end of the day right it's really going to come down to finances and You know we've all witnessed that curve that shows you know Disdoubling every 18 months and when they double they're the most expensive ones But I need the capacity using optimization technology You don't have to ride that wave anymore right you can actually stay behind the curve Wait for the expensive diffs prices to fall be able to put more data and have more capacity available to you without shelling Out the money every time you turn around because they need more capacity So how about pricing? You know everything's cloud services. You think Amazon swipe the credit card. Okay get a monthly bill for s3 or EC2 DR is a little bit more complicated than that isn't it so what's the pricing model? How's it shaping up? Well, so we do have a few services in our portfolio where you literally can go online to IBM comm and swipe a credit card and download and that Typically would serve for example a PC or laptop backup type of need So pricing on that is going to be you know as you would expect it to be somewhere in the five to ten dollars a month You know per user proceed DR on the other hand really takes the shape of How much of your infrastructure we really talking about it's verily one server right is it a physical or a virtual server? So that's going to change the pricing is it you know, what is our your objectives in terms of recovery? Would you like a you know very short window of exposure? Can you tolerate 24 48 hours of exposure all of those variables kind of play into? Both a capacity-based cost as well as there are service costs on top of that So in the case of a declaration in the case of a test those tend to be more one-time charges versus ongoing rental charges Is that a belly-to-belly discussion that that an IBM sales rep or an IBM channel partner has yes with the customer? It's so we're not at the point now where we can have that discussion in a way That's standardized across the customer base do you do you see that ever happening where we can just sort of dial an RTO RPO because people don't know what RTO and RPO is correct explain it to how much data you willing to lose right nun Right, okay. You want to pay five hundred million dollars? No, right? Okay. That starts the conversation. I don't know how it ends I think Dave that what I'm seeing here at VM world is We are getting closer to that non belly-to-belly conversation, but it's going to be very workload specific It's going to be you know first in the wind-tell and lintel environments where there are fewer variables in terms of the application sitting on top of them So we can let the users sort of self-select into some You know online a little more click transaction type of DR services, but when it comes to the I need to protect my enterprise my enterprises, you know Revlon or Southwest Airlines or any of these other enterprises that we've heard from They're not doing that by just swiping your credit card They're they're designing and really building resiliency into their business Dave I think what you're seeing right is is this whole notion of RPO and RTO really just changing to a quality of service when you get to The cloud right so it's now no longer. You know, what is it? It's here's what I want How do I get it and if I can actually get get a better quality of service because I can move more data more efficiently by Optimizing it in some form or fashion now all of a sudden I've increased my my QoS and if I can do that and we can be the the leading cost effective Evangelizers of a better QoS for the user it you know this stuff just kind of works now a lot of its commodity I've been walking the show floor, you know in the old days used to have to internet in to go see a demo through now This stuff they show it up. They wheel it on the floor just works. So we've gotten past that hurdle now It's I want it to work better. I want it to work smarter I want it to work more efficiently and this is what the cloud is enabling right now I think that point about quality of service is a really good one because that allows you as the cloud service provided a Package things guarantee a level of quality of service and maybe it's not just cost-plus. Maybe it's a value There's a buyer. I'm happy to have you to pay that so that's um, that's good My last question Steve is is Ken's things like real-time compression actually speed recovery or is Is that not the case no absolutely right if you think about if I can actually Compress and let's just hypothetically take a 50% compression ratio If I can move a file back to my storage over the over the land That's 50% smaller in theory. I get 50% better performance So instead of moving a terabyte of data. I'm moving 500 gigabytes if I'm moving 500 gigabytes I can move 500 gigabytes faster than a terabyte that becomes hugely valuable to me and it helps me increase my quality of service And what did I do? I added a little layer of compression. All right guys. We're out of time. Thanks very much for coming to the Cube Brian Regan Steve Keniston from IBM appreciate you sharing your knowledge and your insights great to have you