 Okay, we're back, this is Dave Vellante, and we're live here at the Dell Storage Forum. This is day two of Dell Storage Forum. We're in Boston, and we're joined by Terry McClure, who's a senior analyst with the Enterprise Strategy Group. Terry, welcome to theCUBE. Thank you, David. It's a pleasure to be here. It's great to have you. We love to bring the independent perspective, you know, because we bring in all the executives and we get the messaging that's laid on thick at these conferences, which is very good. It used to be, you know, years ago, you know, you'd go to these conferences, it was sort of grassroots, it was kind of hack, not anymore. I mean, we're talking about big-time productions, we've got bands, we've got dancers. Acrobat. Acrobat's amazing visuals, right? It's really quite a change in our industry, isn't it? And you've been there for a while, so you've seen that sort of growth, and I think it's a good thing now. I think this industry's growing up in a good way. Yeah, storage isn't boring anymore. Storage has really become a core and strategic piece of the infrastructure. Remember, in the olden days, storage was a peripheral they had to have. Snorage. That's how my wife calls it, still to this day, you know, I can understand. But us, we love it. We never found it boring. But so, at ESG, you've been there for a number of years now. You're a lead analyst on file-based storage, and I presume you cover a lot of object? Yes, I do. I saw you at NAB. NAB was an interesting show in that the prevalence of object storage, I mean, it just makes so much sense for object storage to be the storage medium of choice for really, really large object stores with the flat file system that makes it so much more manageable and searchable. So what's happening in NAS these days? I mean, we've seen NAS evolve. I mean, NetApp got it all started, obviously. I guess not really. I guess Aspects. Aspects, yeah. And then... Showing our age. How it should be done. You know, and then EMC sort of came in and the EMC's way with Celera and Coach ended up buying an Icelon just to extend that. But so what's happening with NAS? It seems to have lost a little bit of its shine, but it's still a really critical component of the infrastructure, isn't it? I mean, that's an interesting question. NAS hasn't really lost its shine. I mean, we've still got to deal with this massive amount of unstructured data growth. And unstructured data is still the biggest growth component of all the storage out there, right? We've seen a real transition over these past few years in the NAS market to scale-out storage architectures, and we've seen that certainly reflected in the Dell announcements this week with the new fluid file system that they've announced. So NAS has just undergone a massive transition in the architecture to make it something that's more scalable and manageable. You'd just be pretty fainful to manage a lot of itty-bitty small file systems and a lot of itty-bitty small servers scattered about the enterprise. So we needed this transition in order to get these massively scalable solutions out there so that you can get terabytes and petabytes and even ex-sum systems go up to exabytes of data under management. Yeah, so when Dell bought Exynet, I didn't really know much about Exynet. You probably followed them pretty closely at the time, but they weren't really a well-known company, right? They weren't a household name, right? No, they were still pretty early in their evolution. In fact, ESG Lab had tested the Exynet product, and at that point in time, when you think about a couple of years ago even, these scale-it-and-ask systems were almost the lunatic fringe. They were used where they had to be used in highly throughput-intensive environments that had really big files and really needed to scale. And these guys put them in place because they had to because they couldn't survive unless they had these really scalable systems. The challenge is it's really hard to make those types of really scalable systems easy to use and do the techie things that you need to do, the cache consistency across a multi-node cluster to keep all your data clean, right? Snapshots against multiple nodes. It's really hard to make all that functional and easy to use. But Exynet was one of those generations of storage systems that was coming in to really accomplish just that, make it enterprise-ready. And we had tested it, and Brian Garrett, who runs ESG Labs, was really impressed with the potential that Exynet had to be a highly scalable enterprise-class scale-out storage system and start bringing that lunatic fringe scale-out, scalable capacity into the data center. People talk a lot about a global namespace. That comes up a lot in the NAS business. It sounds good. I'd like a global namespace. So what actually is meant by that? And there's a lot of marketing behind that. Oh, yeah, we have a global namespace. But you hear different stories about those who do have it, those who say they have it and don't. What is it? Why does it matter? And who really does have it? That's a good question. And different people, I mean global namespace, it's, gosh, it's almost as meaningless as cloud nowadays. You know, cloud can mean six different things. So many people mean different things now when they talk about global namespaces. But it's essentially one mount point for a file system that enables you to create one really large namespace that can span multiple nodes, multiple storage nodes. So it simplifies things. Instead of having a separate mount point for every storage NAS storage node you have, you've got one mount point that kind of maps to everything, right? So you get a highly scalable system. Some systems have kind of more of a namespace aggregation which virtualizes a bunch of smaller namespaces into one global namespace. And that's what they mean by a global namespace. Others have a, you know, a true global namespace, they'd say, that virtualizes the hardware. So there's different ways of implementing it. But the important thing is it's the fewer things IT has to manage, the easier it is to manage them. And that's what the global namespace does for you. It gives you one mount point, one management point of control. So you hear a lot about object stores. We've been hearing a lot about object stores for a while now. And it finally seems to be at least getting some traction in certain segments. You know, cloud archiving, for example. You're seeing some service providers like Nervonix claiming any way they're doing some significant business. And they also claim that they're going after a lot of the traditional NAS space. Is that how you see it? Is that, again, more vendor marketing? Help us squint through that. Gosh, another good question, David. And, you know, object storage, because of, because you can have really rich metadata around it, it's, and it's a flat namespace. You're not stuck with a hierarchical file system where you try to figure which directory and file folder structure that you put all your data in. So that metadata point you're making, so the metadata inside the metadata together and encapsulated in the object, which is important because it's not trapped in an application or a file system. And it makes it searchable and manageable and you can manipulate the object and you can put some context around it to be able to do further operations with it. So that's attractive. That's certainly attractive. And the scalability of having a flat namespace instead of having to navigate through folders to find things, if you can just search it based on the richer metadata or whatever, it makes it, I mean, so there's this manageability aspects of it that make object storage suitable for really, really large. I mean, we're talking, all of a sudden, we're talking about petabytes and exabytes of data. I had an exabyte scale discussion with a vendor the other day. They've literally can scale their systems into the exabyte range. And how do you manage that if you've got a hierarchical file system that you're browsing through folders? It gets really tough. So the web guys are all over this. So the web guys are all over this. The media and entertainment guys are all over this. And we saw that at NAB. There was a lot of space about this. But now that we're getting more and more long-term online archives, sure, it's a natural fit for that. So let's talk about Dell a little bit. This is, were you at Dell Storage Forum last year? Yes. Yeah, so this is second year. Now, you may have been at earlier ones as well, but I don't know. No, this was, last year's my first. It was the C drive was sort of block based, and that's not your wheelhouse. But I'm impressed with the progress that they've made. I mean, last year was sort of like a trial balloon, and I thought that was good. Michael Dell was there to give it a little charge, but these guys are doing it on their own now without the big guy. And I think it's, to me, it feels real. I mean, Dell is relevant in storage. What do you think about them, and what do you think about this event? Dell is relevant in storage. They've definitely, when you start digging under the covers and you look at the whole portfolio, they've got quite a compelling storage portfolio. They've beefed up the NAS portfolio this week with the fluid file system announcements. They've got from Power Vault all the way up into the Compellant line now with the Equalogic line. So they've got quite a NAS portfolio, the Block portfolio that they've had with Equalogic in the Compellant. So I'd say they're a serious storage player. One of the, and they certainly take it seriously. I think one of the challenges is that, when you come to an event like this, you hear a lot about what Dell does, but they haven't done a great job of painting kind of the vision about why. What's the end goal? How do they see IT look? What's the vision? Five, six years down the road that they're delivering towards. They've got it. They've got it in the bowels of the company. They're doing some really cool stuff when you look at the Project Hermes. Is it Project, is it Hermes? I think, yeah, with some of the technology that they acquired a year and a half or so ago for our network. Presumably Flash, yeah. Yeah, there's some really interesting stuff under the covers there, and they've got some really great vision, but the challenge they have here is they're a serious storage player. They've got a great storage portfolio. They haven't done a great job of telling kind of the big picture story about what the end game is that they're marching towards. So, but they're doing a great job of telling us what they have and what problems they're solving today. It's unbelievable. You walk around their Solutions Expo, and it's small. But you get a sense that Dell's got like one of everything. I mean, there's not a lot that they don't do, right? They've got block, they've got several flavors of block. They've got iSCSI, they've got Fiber Channel, they've got a NAS play, they've got Object Storage, they've got this App-A-Shore thing, which is really kind of interesting. And then the Blade Array. The Blade Array, they rolled out this week. Yeah, the Conversion Infrastructure, right. I mean, the whole Equalogic... That's a really interesting piece. And when you look at what Dell has, not just from the storage standpoint, but all of the engineering that went into that from the server and even the laptop division to shrink down that controller and make it fit into a Blade system, it really showed the power of owning the whole portfolio end to end. I mean, there's some pretty impressive engineering work that went into that. It is impressive. And Dell made a lot of acquisitions, but good ones. They've done a really good job. I mean, I'm trying to think of any duds that they've had. I can't, but I'm not that familiar with the lineup. But I mean, they seem to be leveraging, everything, Ocarina, Exynet, obviously, Compel & Equalogic were sort of out of the box successful. App Assure looks like it has a lot of promise. The RNA I don't know much about. I mean, that's sort of the new kid in the block. And that's the basis, I guess, of Hermes. So this seems to have a lot of potential. Whereas a lot of storage companies struggle with acquisitions. Why do you think Dell is so competent at acquisitions? But first of all, do you agree that they're competent? I hadn't thought about that a lot. So that's a good point. A lot of companies, they make the acquisition and you might go to the storage conference or go to the analyst events and it kind of disappears. And you wonder what's been going on under the covers. But yet here we are talking about what they've done with Equalogic and App Assure and RNA. And RNA's taken a big center stage role this year. And I think it kind of goes towards Dell's core competencies and operations. When you think about Dell, you think about operational excellence. And in a company that's got that level of operational excellence, I just think that they've got this clear vision of how things fit together that allows them to leverage them instead of kind of buy it and tuck it away and have it come out later. Yeah, or just leave it alone and let it grow and don't integrate it in any way and let the customers figure out the integration path. Which happens, right? We've seen it. And actually it works. You've seen companies, look at EMC, they basically bought Legato and took a long time to figure out what to do with that. I think they're finally now getting around to it. I think EMC's doing a better job, but historically they've kind of left a lot of their applications or acquisitions alone. Not necessarily the technology acquisitions. But I mean, they've not all worked out and EMC's really good at acquisitions. Took you so long time to get there. Well you remember some of the earlier EMC acquisitions. I don't know if you remember Epic. Yes. So that wasn't good. Yes. Those are the early days. And then the Documentum Legato, they struggle a lot with some of those early acquisitions. Even RSA has been some of the struggle. Then they hit like a super grand slam homer on the greatest acquisition in the history of the computer industry with VMware. But then subsequent to that, they seemed to get really good. But nobody knew why. At the time nobody knew why. I remember they were trying to integrate it and they were like, that didn't work and there was all this stuff going on with Diane Greene. But after they sorted that out, it's almost like Tucci brought in, okay, we screwed up a lot of acquisitions. Now we know what we did wrong and now we're going for it. NetApp is another one that's struggled with a lot of the acquisitions that it's made. And I'm actually predicting that NetApp will figure that out. I think it has to. I think it's doing a much better job with the LSI acquisition. Yeah, for sure. In Genio, right? Yeah. We're going to hear more about it. Are you going to be out there next week? I'm looking forward to it. They always do a really good job with the endless defense. They do. They spend a lot of time with the endless. They really put a high priority on it. So of course we love that. IBM's another one. They've been pretty acquisitive. I guess done pretty well. I mean, XIVs worked out for them. The NAS space, what's going on with IBM and NAS? It's so NAS. It seems to be struggling to get your huge traction. It had some bumps. Conceptually, really interesting, right? Conceptually, it's great. I mean, but we had some, we talked earlier about how hard it is to take these clustered file systems and make them easy to use and make them functional across these multi-node systems and just make them simple to manage. I mean, so NAS is based on GPFS, a really powerful file system that has a ton of functionality. So packaging that up and harnessing it and make it easy to use and make it reliable. And I mean, it's just quite an engineering challenge to even accomplish that. And it had, it certainly had some bumps. They seem to have figured it out now. They needed to figure out the right target markets and how to position that versus the N-Series that they sell. But I think they've straightened out a lot of the road bumps that they, the speed bumps that they get. Okay, so they've got their own IP for figuring out if an HP's got iBrix, which is sort of their version of Exynet, right? And they're bringing that, that's their version of fluid data architecture. They're bringing that to all their various plays, including 3-Power, right? I mean, so, right? I believe, I don't know. I've given a statement in a direction, but I don't know. Yeah, they, yeah. They've hit it that way anyway. Yeah, so, I mean, it's kind of a logical play, right? You can do that. So that's, it looks like the big guys, IBM, HP, Dell, EMC, buying Iceland, they all get their NAS store together. So where does that leave some of the NAS startups? Like, I would script a name, Avere, Avere. Avere, sure. Yeah, so they've got some interesting stuff going on. They're doing some cool stuff with tiering and optimizing for different media types. Are, are, is that, is still a good startup market? Are they able to innovate and, you know, be a good acquisition candidate? Or is everybody's NAS pretty much solid? For the big guys, you know, the high end NAS is pretty solid. There's still a lot of opportunity kind of on the low end, kind of below the radar of these guys. And that's where you see the people like scale computing and grid store and some other, there's some others coming into the market now. So there's still a lot of room in that, with probably the SMB space. But, but most of them, and that's where you see emcee with the VNXE targeting, right? Most of them have the higher, higher end NAS pretty well locked. Avere's an interesting play though. Avere's, Avere's a, a chameleon of sorts. Because you can take it and you can, the global namespace discussion we had, you can put it in front of a scale up system and multiple scale up systems and have a global namespace and virtualize the NAS storage infrastructure and kind of speed everything up with the automated tiering. But you can also take edge nodes with them and they've got a very interesting edge node use case where you can centralize all your NAS storage at the home data, at the home data center and put a node on the edge which will cache all of your edge operations yet have the primary storage back at the core. They've made some interesting technology investments and had a product release maybe a month or two ago that significantly sped up the performance of those edge nodes, increasing the use case whereas before it just used to be not very right intensive workloads, they got about a 50 to one reduction in latency for those edge operations that had to go back to the data center which is pretty impressive. With a new release they're claiming 500 to one. So now you start looking at the use cases like maybe virtual desktop support at the, at the Robo remote office. Back office, yeah. Yeah, so some interesting use cases on the edge for Avere that they can address. Serious geeks too out of Carnegie Mellon, right? I mean, some really, really smart people. Very, very smart enthusiastic group over there, yes. Yeah, yeah, like the Pittsburgh startup scene, right? Yeah. And so, let's see. So what's happened at ESG these days? You got any research you can talk about that's new? We're digging into a number of areas. We've done some interesting research on the big data, you can't walk around a corner without hearing about big data nowadays. And we just finished surveying a bunch of IT folks on their big data plans. And we wanted to understand the analytics environments. Not just, I mean, we weren't just looking at the Hadoop side of big data. We wanted to understand analytics in general and what users were facing and what the tipping points were regarding their largest data sets and their fastest growing data sets as well as find out just where Hadoop is. And so we found some really interesting things regarding how mission critical the analytics is becoming in data centers. You know, it was before people were doing their batch updates, you know, once a day or once every couple of hours. Now that so many more people are looking at doing real-time updates. About a third of the people we talked to are doing real-time updates. When you talk about how that translates into the storage infrastructure, what used to be kind of a commodity-based dash storage infrastructure may not stand up to the levels of availability you're looking for when you're doing real-time analytics operations. You need to start worrying about a little bit more about availability and the cost of downtime if you can't do your real-time analytics. So we're seeing users willing to pay a little bit more for their storage infrastructure now instead of putting all that analytics stuff just on DAS. Start looking at high-availability external network-attached storage systems to support that. That's an interesting shift in analytics. Yeah, so the big data thing is obviously exploded. In fact, this week, not me unfortunately, but we're out at the Hadoop Summit, my colleague John Furrier's out there with a bunch of folks doing the cube out there. A lot of interesting things going on. The number of companies are taking their NAS product, for example, and maybe doing a deal with Cloud Era, or doing some kind of connector into Hadoop. Is your sense that that's the right play? That's the right storage for HDFS? Or is it still too early to say? I mean today there's a lot of white box stuff, right? A lot of Seagate. Which makes sense, when you look at the business units that have been playing with Hadoop right now, it's not the data center that's been playing Hadoop, it's more the business units, and they're just putting it on a group of servers with internal storage as a prevalent use case right now in Hadoop environments. As those applications are getting brought into the data center, IT's looking at it and saying, wait a minute, we don't use internal storage, we network our storage. We've got high availability, we've got this data protection processes in place, DR processes in place that we have to map to. So as you start seeing the Hadoop stuff move from the business units into the data center, we see an uptick in the intentions to put it on sand storage, and even cloud, an uptick in cloud. But the approach is interesting, and I don't want to, when you start looking at Hadoop and some of the things it does and the way it operates, breaking up the big operation with a bunch of subtasks and farming the compute operation out to where the storage lives. If you network the storage, and you have a storage pool that you're sharing, by nature you can't put the compute operations where the storage lives, because the storage is striped across a whole bunch of nodes. So what's the right approach? Some vendors say they can mitigate the latency associated with not running the compute at where the storage data lives, and they say that you don't really notice that, their storage is fast enough, maybe they use InfiniBand back ends, but I think we'll have to wait and see what the trade-offs are regarding that locality of reference benefit you get with the Hadoop architecture versus the ability to have your high availability shared network storage. I mean, there's trade-offs to be made for both use cases. Yeah, and Hadoop is inherently batch, as you know, and everybody's trying to push to real-time. You've seen some interesting developments, and guys like Adapt are really trying to reach around the connector strategy and do stuff sort of in an integrated fashion. I know Vertica's made some statements in that direction, so it's very exciting space, you know. We love new trends like this, right? We're analysts, so we take that stuff. Now you guys also, ESG had a big presence at EMC World, you were at EMC, you guys had a booth reaching out to sort of, I don't know, I guess practitioners. Steve gave a big, Steve Duplassie, the founder of ESG gave a big keynote, Steve and his Dancing Girls, that was, he's a character. So that's, so you guys have had the labs, you got a zillion analysts, you guys are really, you know, the storage, you know, I don't even call you a boutique anymore. I mean, you guys are, you know, a large, you know, mid-sized research company, but really best in class in storage, no doubt. And seems like, you know, you guys are growing great, so congratulations on that. Thank you. And love to see the outreach to the practitioner community. And, you know, we're getting some interesting traction with the practitioner community because of some of the other research that we've recently started doing. We're looking, taking a good hard look at the online file sharing collaboration space, because, and I guess to sum it up, Enterprise Dropbox is causing some big challenges in IT today, primarily because, you know, endpoint device sprawl, I myself, this is, I'm a geek. I've got seven endpoint devices that I use. It's a little sad, it's okay, you can tell me it's sad. I bet you, maybe not that many, but getting close. iPads, laptop, I've got a desktop, I've got a smartphone, I've got an Android tablet. And the big challenge is sharing data between all my endpoint devices. I want access to one set of data from all these devices, right? So Dropbox, if you're familiar with Dropbox? Yeah, sure, I mean. Simple solution. I was going to ask you what you thought of the Syncplicity announcement that MC made. And I think Syncplicity is a fascinating announcement because the challenge here is, if I'm using Dropbox and I go out and get my account to solve that problem for solving, sharing one set of files across multiple devices, I own that Dropbox account. And if I leave VMs, yeah, there's a Freudian slip. If I leave ESG, I know I said VMC a long time. If I leave ESG and I have subscribed to Dropbox, that Dropbox account, it comes with me. It's my, the data comes with me. Nobody knows what's in there. Nobody's erasing your. Nobody's erasing my Dropbox account. And that's what's interesting about it. Or they can erase your endpoint devices, but it's your data. It's my data, it lives on my home computer, right? Because it's my account. So that's what was interesting about the Syncplicity acquisition. Syncplicity looks at giving an IT dashboard and control and management of those shares. And that's what a lot of IT practitioners are struggling with today. Dropbox is coming into the enterprise. People are using it without permission. There's a big article last week about IBM outlawing it completely from the enterprise, along with Siri, which was interesting. But it's happening. Well, that's because Siri sucks. I don't know, I don't have Siri. Although the commercials are still spooky. It's good if you're really lonely, and you want to use it, but it's useless, I can attest to that. Hopefully it's going to get better. But okay, so that's the play. So the play is to give, so IT is looking at what do I do about Dropbox? Is there an enterprise class solution? And that's what Syncplicity gave EMC. Something that they can deploy across the enterprise that allows users to share one set of the truth across all their endpoint devices, but gives IT the administration security and control that they need to continue to protect corporate assets. And that's a fascinating market. We're having end user discussions, and part of our end user practice is centered around having these discussions about decision making. We've got research coming out in this space, looking at all of the business class offerings in this space, and we're just getting a ton of traction in digging into these solutions. And you called it a collaboration you referred to before, I mean it's storage, but it's really a different use case for storage, and it's just collaboration related. And it's maybe disruptive, maybe not. I don't know, now Dell doesn't have one, right? Dell doesn't have a Dropbox for the enterprise, and I'm not sure they should. They don't have a Mosey. Well, Dell partners, you know, if you look at some of the partnerships, the DX Group partners with the Cetera actually has a Dropbox-like functionality as well as a gateway, and there's a couple other people that they partner with to offer this type of functionality. Yeah, I'm not convinced you have to own that just yet, I mean, it's an interesting move by EMC, but they put it into the IIG Group, yeah, so it's not really a pure storage play, it's more of a collaboration play, as you were saying. But what's interesting is when you start, when I talk to end users about what they're doing, when they're deploying Box or Dropbox or Syncplicity, or not Dropbox so much, because it doesn't have the administration and security and control you need, but when they're deploying Box or Syncplicity or Ignite, and I say, are you augmenting and creating new functionality in the enterprise or are you replacing? There's two things that they're replacing. They're replacing, they tell me they're Linux and Windows file servers, and they're EMC and NetApp NAS storage, they're replacing their file storage with these online file sharing collaboration platforms, and they're replacing some of their SharePoint, right? Right. Because the two use cases for these platforms are the single version of the truth that I can share across all my endpoint devices to enable me as a remote or mobile worker, and easy collaboration and sharing of files and workflow. So it is disruptive in that sense. It's going to be interesting. I mean, we're going to be digging into the potential share shift from NAS into these OFS, OFS is our internal term, sorry, but online file sharing collaboration solutions. We're going to be digging a little bit into that share shift. Interesting topic, so look for that research. Terry McClure, what's your Twitter handle? ESG analyst, TMac. ESG analyst, TMac. TMac. ESG analyst, TMac. Follow Terry McClure, very knowledgeable. Now, we couldn't be here without the great support of Dell, of course, inviting us in, and also legal seafood. Do you like legal seafood? I love legal seafood. So legal seafood, Father's Day is coming up. Follow you out there, Father's Day. I happen to love it. Shoplegalseafood.com, check it out. Shop.legalseafood.com. They ship all over the United States, so check them out, and they've given us, they've been great with some gift certificates, so I'd like to give you this gift certificate to legal, so go enjoy, and thank you very much for coming on theCUBE. Really appreciate it, it was great to see you. Thank you, it was great to see you, I appreciate it. All right, keep it right there, we'll be right back after this word.