 Hi everybody, we're back, this is Dave Vellante with Stu Miniman, this is theCUBE, Silicon Angles production of Edge 2013, we're here live at the Mandalay Bay. Jeff Guntner is here, he's the director of storage practice, lead at CMI, integrator, reseller, flash systems expert. Jeff, welcome to theCUBE. Thank you. You know I did some research before I came on and you guys maybe know this, but out on the web they call you the ESPN of tech. I love that. Thank you. I've been with Dan Patrick and Vakini May, this is like incredible. We tried to dress the part as you can tell. We should do a top ten afterwards or we're done. That's actually a great idea. Well anyway, welcome and thank you for the compliment, we appreciate that. So Edge is a big show for you obviously as IBM partner, we're going to get into that and talk about it, but just give us your impressions of Edge. It's fantastic, so year two, I know you guys were here last year, IBM's really needed a conference like this for storage specifically because IBM is a player in storage, but you always hear EMC in that app and IBM is somewhat of an afterthought, but as you heard Ambush say we're becoming a forethought and we've been selling IBM storage for 36 years, we've been in business. So IBM storage is very much a part of our DNA and we have a very solid portfolio and it's great to share it with the industry, clients, prospective clients and I love it. I think Edge is a great event, it's only going to get bigger and better and just like our storage portfolio. So it's exciting, it's going to be here. So talk about CMI a little bit, talk about your specialty and how you're helping clients solve their problem. Yep, well we're located in San Francisco Bay Area, so we are regional, Northern and Central California and we specialize in a very specific segment of the market where its customers mid-market as IBM's defined space and GB, so less than 100 employees up to maybe 10,000 employees and we really have three practice areas that we specialize in. So the first is business continuance, the second is storage management and the third is data center optimization and everything that we do is backed by two tenants, strategic principles, strategic consulting, assessments, CIO as a service, implementation of strategy and then tactically we go through and we can help clients not only develop the plan through our process but also do the implementation and be very tactical and we use IBM technology to do it. So what are some of the big themes you're hearing from customers? You go on an initial call or even a client that you're seeing for the nth time, what are the patterns that you're hearing in the client base? There's two big ones right now, it seems like it's the back, the summer of backup and recovery. Backups are fundamentally broken, customers know that they're broken and they're really scared about doing restores so the two discussions we get into, backup being one, the second one is flash systems, you mentioned at the top and we see a lot of growth in this space, we've actually done a lot of flash implementations, end of 2012 into 2013 and it's a hot area for us solution-wise. Well so let's unpack each of those, backup, I love what you're saying here, ESP on attack, backup is broken, you sound like you've worked for Wikibon and so backup is broken and we've often said data protection as a service is the fix but getting from point A to point B ain't easy but why do you say that backup is broken? It's what our customers tell us and I think it's true for any vendor, it's backup is, I like the way you position that, data protection as a service, it's a good framework, we can unpack that but I always tell people backup is an application, you might say but it really is an application in the data center that gets neglected and people take for granted that data is just getting backed up, nobody does a restore, so... That's a very high performance demanding... It is, right and we see multiple methodologies in the market, you've got the incremental forever, you've got a flat file backup, you've got continuous data protection, so you've got some of the things that we do with our N-Series, NetApp, SnapManagers, disk to disk and a lot of discussion for us is around cloud too, backing up to the cloud, getting it off-site but still providing enough data protection for the customer and moving away from tape, people hate tape but fundamentally economically for your storage environment, it plays a critical part, so I don't think you're ever going to get away fundamentally all the way from tape but people are really trying to do it. Well, I mean we agree that you probably don't want to use tape as your primary backup but you still, you probably should have some tape. People do, they're actually here at IBM, there's a bunch of sessions on LTFS, so there is use. I'm curious when you talk about kind of that low cost, does Amazon Glacier come into discussions of your customers? It does. How can it not at, what is it, penny per gigabyte per month? It's penny per gigabyte but it's, you know those iceberg charts? You got to pay attention to what's under the iceberg. It's an attention grabber though. It is. You saw your penny and then you really belly up to the bar and you have to figure out, okay well how much data am I going to take out? What's my cost to get it out? What's my cost for the pipes? Will they ever see it again? Yeah, exactly. What if the Amazon site goes down? So it does come up. It comes up in an archive, a deep archive, where they know they're not going to do the restore within 90 days, maybe six months. So definitely in the context, we're also being an IBM partner with the IBM Object Store Cloud, which is an Erbonics offering. We're also an Erbonics partner. So that comes into play as well. Yeah, so we often say that one of the problems with backup and one of the reasons why it's broken is it's a lot of customers do one-size-fits-all. In fact, we just did a study on this, saw the preliminary results. The predominant response was yeah, backup is for us, is we got one service level. That's it, take it or leave it. So by definition, you're never going to get it right for the workload and the application. You're going to either be spending too much or too little for your backup. And that's a problem. I totally agree with you and that's why we start when we enter all those discussions, the customer says, hey, my backup is broken. So well, when was the last time you really did a business impact analysis? Because the next question is why backup parlays into DR, which is big for us as well. And folks don't want to do DR because it costs too much. But they don't spend the time to really sit down and figure out what those critical applications are. And really, well, maybe I need to back up these three with this LA, SLA and then the rest of them can have a lower one. So when you really sit down and go through that exercise and spend the time to do it, you get that homo with a client, they realize, hey, I can do it. So are you able to point to some examples where you're actually implementing backup as a service or data protection as a service? Not yet. But we've had a lot of conversations with our clients about optimizing their backups, doing business impact analysis, figuring out what truly does need to be backed up. And then that waterfalls into what their DR plan is and getting them close to the dream. Yeah, we live in the dream here. Alright, now let's get into the flash. Sure. So IBM put a billion dollars, you know, up on the table. So we're going to invest this in flash and on April 11. Big move. That was made you happy. It was very happy. Very happy. What is the cut? Did the customer see that? Did they notice that? What's what's the reaction been in the client base? Customers did notice I was in the actually the solution center last night. And one of the guys from IBM was holding the flash system and there was literally like 20 people crowded around him. So it's it's hot, right? People want to know. So there's no shortage of that. And our clients love it because for two reasons. Number one, it's accessible now. It's real in terms of price performance. And nine times out of 10, they're willing to spend the money to solve the problem. What's database acceleration, it's just improving an application to increase revenue because they need to do more with less. So those are really the two odd areas for us around flash database being one of them. We've got a lot of case studies. We actually have an TCO ROI chart where we can go in and we can tell you works well with the Oracle clients, do an Oracle AWR report, get the report, we sit down, we analyze it, we can tell them with 99% certainty whether flash is going to improve the environment and by how much that water falls into our TCO where we go through and we tell the client, well, how many Oracle cores are you running today? Okay, yeah. This is where the CIO starts to care, right? Flash, okay. But when I start to save you software costs, licensing costs, maintenance costs, that's where it becomes really interesting. And we've done that and done it successfully. And the model works. So you got examples of that? We do. You know, handful, dozens? Yep. We have clients here at Edge. They're walking around and we did, I did a panel yesterday. So we uncovered that. I don't know, Stu, when did we do that study? But it was several months ago, maybe six, nine months ago. Yeah, earlier this year. And it was, wow, the light bulbs went off and we talked to a number of customers. It was basically where they were spending maybe $100,000 more on storage. Right. And they were saving literally millions on the back end on cores, but more importantly, Oracle licenses and maintenance, massive savings. Yeah, we did a case study with a client last year where it was a non IBM vendor, let's say, a very large array. They were running on. And we went in, we did the proof of concept with them, showed them Flash, showed them how much they could reduce their core count by and they saved a few million dollars by moving to Flash. So I understand Oracle is going to change its licensing policy. It's now not going to price just on cores. It's cores times Flash capacity times two is actually how the new pricing is going to be. The America's Cup isn't September. Larry does have to, yeah, it's kind of, boat problems. Well, but I mean, I'm joking, but Oracle is actually quite good at morphing its model, you know, whether it's virtualization or whatever to make sure that that net contract value increases. Jeff, I wonder if I can ask you a question about just kind of IBM and Flash in the field. So many in the industry have kind of criticized IBM because they're not first and maybe not even second, but, you know, what's your customer adoption bin of Flash? Where do they, do they see IBM as a thought leader in this space? Is it hitting the market at the right time? You know, where does it fit for your customer environment? It's a good question. So for us, being in San Francisco, we're in the hotbed of Silicon Valley. So we see pure storage. We see nimble. We see Nimbus. You name it, we see it. And those guys do a good job of innovating and get driving, helping drive customer demand, validating the market. And then what you see happen is EMC buys Extreme IO, right? IBM buys Texas memory. So everybody starts gobbling up the small players. So in terms of IBM, they may be slow to do something, but I think they always generally make the right decision. And they take their time to assess the market, which sometimes is frustrating for us. However, we know that when they make an acquisition, they're going to improve it and make it better. I think XIV is a great example of that. The store-wise acquisition with real-time compression for V7000. And now this Texas memory acquisition has been phenomenal. We're doing at least one client engagement a week sold, but we're having two, three, four customer dialogues about improving their environment with Flash. And it generally centers around the database, or it's a C-level led initiative where they need their reports at 6 a.m. And right now, they're getting at 10 a.m. So it's about better business insights, and Flash is a very quick, simple, easy way to do it. So I love this discussion. So you lived through the tier 1.5 bubble, right? Which is great. I mean, good acquisition with XIV. I mean, 3-par going for 2.5 billion. Compellent went out. Icelon, you know. That was a phenomenal wealth creation event for the storage industry. Well done to many of our friends out there. So that's fantastic. And several have posited that you're going to see the same. Sort of, I guess, deduplication went through a mini version of that. It did. And many have said, OK, Flash is going to go through a similar type of phase. But then you've seen IBM made a move with TMS. Granted, TMS has been around a long time. EMC, instead of paying 2.5 billion down the road, spent whatever, let's say a few hundred million for extreme my own. Do you expect the same kind of, you know, but who's left, right? Oracle, I guess, could buy somebody HP, really, is paying down debt before it buys companies. You know, what do you think about that? I think there'll be some more acquisitions. The next kind of two targets in my mind are violent memory, which SAP has put a lot of money behind, and pure storage. So we've seen in Northern California a lot of pure storage. They have 2,000 customers now. I'm sure you guys have talked to them on the weekend. Yeah, we know Scott. He's from theCUBE. Great product, and they're growing very well. So they're either going to IPO, or they'll get acquired, or maybe, and or both. You never know. So I think those are the two that I have in my mind. Nimble seems to be catching on a little bit, but not as at scale as violent or pure storage. Right. And now, do you resell other Flash, or just pure Flash systems? We do. We resell violin. You do? Mm-hmm. So where do you, let's have that conversation. Where do you use each? Where do you recommend each? Presumably it's client-driven, but help us squint through that. It really gets back fundamentally for us of having the conversation with a client, and having a business discussion, and figuring out what the total problem is, what the solution needs to be, and then going from there. Even though we are a violin partner, I do fundamentally think that, excuse me, TMS is a better technology. And in the things you can do with it, coupled with SVC, it's the right direction. You guys have probably seen it, there's kind of two things going on in the market. There's either an appliance model, or you take a pure storage model where you've got compression, deduplication, Flash, and it's truly a Flash array. It's innovative in that it's using Flash, but it's not really innovative for the storage industry. It's still the same paradigm, it's still an array. So I think we'll see who wins out. So you said TMS fundamentally a better product, architecture, whatever, with SVC. Did I hear that right? Because of the stack, because of the storage stack that you get? You can do both, we do both. So we do TMS straight up attached to a server, and then we do TMS, or IBM Flash now, with SVC. And the former, who's doing the storage management, the application presumably. Right. Okay, so you look for situations, that's the TMS legacy, that's really where it's sold. Correct, or you manage it as an array in the environment. It's very easy to manage. Okay, and then compare that to the violin approach. Violin, you know, OEM's a semantic stack, right? That's relatively new. Okay, so what makes TMS better? I think for us, it's the simplicity. It's the data density, so in 1U, 20 terabytes, 500 watts, compared to violin 3U, same densities. So it's packaging. Packaging is a huge, and our clients have small data center footprints. Even if you have a large data center footprint, so density, power, and cooling performance, they all matter. And that's not to say that violin doesn't perform, or somebody else doesn't perform, they all perform well. It's just, I think density matters. And the innovation in engineering is what we like. Awesome. All right, Jeff, well listen, thanks very much for sharing your deep insights. Definitely a tech athlete there, Jeff. I appreciate you sharing with us. Yeah, you are welcome on theCUBE anytime. Boom goes for dynamite. Yeah, to boom. And we actually have, you know, John Furrier, my partner, we have an office in Palo Alto, we've got a studio out there, so love to do more with you. Yeah, absolutely. All right, everybody, keep it right there. We're right back with our next guest. This is theCUBE, we're live from Edge.