Personalizing The NAS Space.

Network Attached Storage from one vendor's perspective

NAS has evolved from a niche storage alternative for low-end networks to a serious storage solution for a much larger market. NAS is now forecasted to grow at over 60% per year, and approach $7 billion in annual revenue by year end 2003. NAS sales are expected to double in this year, and may grow even faster as solutions scale further and add functionality.

To give our readers an insider's point of view on NAS in the Enterprise, we present excerpts from a recent interview with R. Daniel Smith, president and Chief Executive Officer (shown) and Steven Paulhus, vice president, business development at StorLogic (www.storlogic.com) a NAS pioneer headquartered in Lake Mary, Florida. The complete interview is available on the CTR website: www.wwpi.com.

CTR: How did StorLogic come to the NAS space?

DAN: It's actually kind of an interesting story, the company is nine years old. It started really to develop network management utilities for optical media. It was very early in the NAS markets with their own small network attached brick running their embedded software, again handling mostly optical media, CD caching, etc. So the company really has its roots in the NAS environment and really got started because of the early requirements to share CDs and to make it easy for users to be able to manage and share those devices over large integrated networks. So the company really was an earlier pioneer in NAS development. Unfortunately for a number of reasons, it never really developed marketing and sales wise until the last two years when the entire management structure changed. So it's an interesting story, it's almost as if this were a turnaround company in itself because not only did we turn the internals of the company around but we've also completely turned, we think, the paradigm for being able to manage not only optical media now, but also storage and storage applications and applications themselves over large distributed networks.

CTR: StorLogic is the first NAS solution to utilize the Windows NT Embedded 4.0 platform. How did it come about?

DAN: About two years ago, we leveraged the relationship we have with Microsoft to become a beta site for their NT embedded products. So basically, in a nutshell, what we did was we took the NT operating system in a pure headless environment--there's no mouse, there's no monitor, and there's no keyboard--stripped away all the Microsoft IP that you would normally see in the product and wrote our own wrapper around the core operating system. So as a user, you never see NT. You only see SerView, which is our management product. So we have taken what we would consider to be an entirely different path in terms of product development for being able to share, now, not only CDs, but DVD-RAM and DVD-ROM, tapes AlT and DDS, any type of disk drive, any type of manageable storage subsystem, i.e. RAID, up to the point we can actually manage now via a remote agent, your normal NT 4.0 server system.

CTR: How did you bring NT to heel? As you know, it's kind of pushy in terms of mapping storage for its own usage.

DAN: Actually we have a fairly well trained and educated group of programmers who were able to really go through the NT space, separate out the pieces we required, the pieces we didn't require, and around that we wrote an application using Java virtual machine that sort of acts as a virtual interface between all of NT and NTFS and our own UDF and NFS file systems. So it was very easy for us at that point then to manage down a very, very low level all of the disk space, all of the share points, and both logical and physical conditions you see in an NT environment.

CTR: Are you more a block level management or a file level management device?

DAN: Right now we are a file level management utility. We do have some opportunities coming up to change that, but right now we specifically concentrate in a NAS environment. We have a pretty robust set of products. that are soon hitting the marketplace or has hit the marketplace that really leverage a company's ability to distribute storage and the application down to the user level, yet maintain centralized management and centralized control over the operating environment itself.

CTR: Does that mean you're going to be expanding to the SAN space?

DAN: We feel like we obviously have an innate bridge into the SAN space, whatever that space eventually becomes. So we think we got some clear technology paths, there's a whole set of issues that I believe the industry as a whole has to overcome, but I think as a player in the NAS market, we provide a well rounded, highly functional device that does both single function and multi-function bias, yet we have a very clear ability to migrate to SAN as those standards develop.

CTR: How do you see your building on your Microsoft relationship?

DAN: It's NT versus Linux in the embedded space and there's a war. So they're very excited about this space. We've got really two things that we're building on with Microsoft. One is from an engineering technology standpoint. We are very much in tune and have great communication going on from the engineering side. We've just participated in one of their design reviews for the next generation of their embedded NT product. There's a confident dialogue between the engineering groups of both StorLogic and Microsoft. And then we've got the business side, which is the side I'm managing right now where, again, they're excited that we've selected embedded NT as our operating system base for our NAS products, we're the first guys to do it, so they're excited about helping us and making sure that we succeed. In terms of business opportunities we're doing joint road shows, trade shows, sales calls, they're putting StorLogic products in their executive briefing center in Redmond, some of their major district sales office s around the country, they funded a case study, produced a case study on StorLogic that was sent out to their user base and sales people, so there's lots going on there.

CTR: How do you differentiate between SAN & NAS?

DAN: Since we play primarily, now, in the NAS environment, we almost need to differentiate a little bit between those two segments of the marketplace. SAN, we think, is a good concept. But the concept isn't there yet. We think some of the interoperability problems have obviously given SAN sort of a black eye in the industry. It is very difficult for your average company, without hiring massive amounts of external consulting help, to really structure and deliver a SAN that fulfills the promise of the concept. In NAS, we don't have those obstacles. We're still typically at the SCSI level it's very easy to move files, it's very easy to share data. It really moves to, in our case, where do we take the flexibility that you've conceivably obtained by implementing the SAN and move that down to the NAS level. And that's the structure that we took in developing our SerView software, which really gives the end user and the IT manager the opportunity to co-locate the device, the appliance, down at the user level, yet m aintain control over the physical hardware, the application space, and the storage that underlies the whole NAS product.

CTR: Have you identified storage service providers and application service providers as strong potential customers?

DAN: We have, but again we sort of tried to break the mold on the concept of what those guys can offer the end user to the customer. Our product really is designed to not to be a data center product. As an ASP, I think the biggest issue you have to face is that ability to get a large number of users and a large number of requests for data or application space out to your center and back down to that end user. We took a different approach. With our approach you put the device, the appliance down at the end user level. The ASP then dials in to manage the device, the applications for and the storage support at a local level. It's a much cleaner model, we think, and hopefully will allow ASP type of business to grow a little more rapidly.

CTR: We've discussed your involvement with NT embedded, you're experimenting some in the Linux space, I believe. Can you tell me something of what you're doing?

DAN: Well we actually support Linux in terms of our management client, the client site. We don't run Linux internally, nor do we plan on running Linux internally, as an operating system. We also support an FS internal to our SerView software, so as a client you can access as a IT manager if you happen to be running at a Linux workstation, you can manage our appliance remotely through that.

CTR: File systems are becoming a more and more serious issue as people are trying to bring value to the marketplace. You're running an FS in your software suite. Are you working, or looking, at some of the other software suites, such as the one that Hitachi's pushing: the DAFS?

DAN: DAFS is an interesting concept. I think it's really a design to supplant the NFS system we all have grown to know and love. I think actually NetApps is pushing DAFS probably harder than anybody. They've put together an interesting collaborative group, there's a lot of good companies. That and what we call the VIA, virtual interface architecture, brings another level of interoperability, if we want to use that word, to being able to, again, share files. Now the system, as I understand it, was really designed to be used in a cluster file sharing environment. It does not do what a NAS appliance, in our opinion, should do. It is not good in terms of security, it is not good in terms of general file sharing operations, and it is not good in terms of wide area networks. So while the concept itself is very interesting, I think the premise that it is going to be pervasive as NFS has become is probably not correct. I could be wrong, but I think as we see this roll out, there seems to be a lot of limitations. In terms of what StorLogic did, if you take a look at our software and some of the concepts that we took, we'd like to think we have created our own virtual architecture. And we do that at a couple of levels. One, we manage both the physical and logical interfaces between NT and the end user. We also provide a virtual interface between NT, UDF, and NFS. We provide a routing mechanism, very similar to VIA, without being bound by hardware requirements that actually runs between our server appliances that gives us a centralized one to many view which is unique in the industry. And we provide a virtual interface between the appliance and the application space that links that back to the logical units. On top of that, we offer two more layers of software, and you've probably heard us talk about storage domain and virtual domain, where we've basically virtualized the devices based on what their logical use is in your company infrastructure. Again, we want to put the devices down at the user level. By being able to co- locate these down in the accounting department or down in the engineering department, we then can logically group those devices together, single view, see everything that's going on inside that logical unit, not only in terms of storage and storagemanagement, but in terms of device management, who the users are, what files they've been using, and what application and application space is available to them. So we took a similar path as to the VIA standards, yet we really made it kind of real world. We went out and we looked to see what people were really doing with appliances and with NAS devices, what we thought we could bring, because we do run NT, there is a wide variety of applications that we can run on the devices, so we obviously encompassed the ability to manage those inside of the software. So while both of those concepts, DAFS and VIA, are good, I think that what we did was very technically elegant, yet very easy to manage. As we see the industry progress I think we'll see another switch in terms of how we're going to move files and application space between servers and end user. I don't think we're there yet.

CTR: Who do you think your competitors are, and will be?

DAN: Let me clarify what we consider a competitor. We have what we consider competitors to be on basically two levels. Those who took existing RAID hardware storage products and the associated technologies and re-packaged them into what are now called new NAS appliances. We really don't consider them to be competitors. We think as the market begins to differentiate us from those levels of products, they will quickly fall by the wayside. It's a lot like RAID was four, five, six years ago where there was a general education process going on. There were a lot of integrators, "manufacturers," that were able to roll their own products simply because the technology was such that you could take existing controllers and a box and a series of cables and make a RAID device. And there are a lot of companies that have been very successful in doing that. As NAS becomes a little more mainstream and as people become a little more educated, not just to what it does functionally, but also the opportunities that it gives the c orporate manager and the IT manager responsible for really the value of the data and the value in the data, we will see, I think, those companies falter as they did and as they continue to do in the RAID storage subsystems business. We look at Sun obviously, EMC, and Network Appliance simply because of their size and their ability to impact the market. And they also drive a lot of the perception as to what SAN and NAS are. So we look to those guys in terms of where we think they're driving the space and we simply try to get there first. Which is something that I think we've just done with this release of the product.

CTR: Where do you see the integrator grow in the successful implementation of NAS or SAN? Is it a time thing, that he's the early adopter par excellence and then he passes along his expertise to people who need his ability less as time goes by or is the complexity curve such that there's always going to be a need for somebody who knows more than the other guy?

DAN: I think it's the same old story, there's obviously a requirement to add value to any successful product. The added value could be, in the old school, hey, I'm the guy that knows how to put it all together. It could really be that simple. I think as the application space and the client space changes, you're really going to have to differentiate. It is going to be less technical in terms of how the product is assembled, in fact with our product, you do not need an IT manager to put the product on the network. It is very intelligent and it is very easy to add one of our appliances to an existing network or to really structure the product to be your single file server in your small home office. So we think the model is going to change more or less over the next year or two. Right now as an integrator or a VAR, you have to have certain levels of network or enterprise software application experience to really be successful as an integrator or VAR. We think that model is going to change a little bit. I think as we see some of the flexibility move into our space, some of the power of mainframe applications become easier to use at the appliance level, it will become more focused on customer support and customer service and less of that focus on the technical support of a router, or a hub, etc. We think it will open up the market, somewhat, to that integrator or that VAR, that is really more tuned to customer resources or the management of a customer as a resource versus the management of the technology as a resource. I think it would be good for a certain level of our integrator that recognizes a change in the landscape. You have to find someone, and I also think it depends on how vertical some of these appliances become, but I think you're going to find VARs and integrators that really know the environment that customer works in. That's probably the key to making any integration products successful: they know what the customer's expectations are, they know what the business model is, and they know what products are going to fit into that. That really becomes, I think, will become more important than the fact that I can take an NT network and Novell network and a couple of Sun workstations and get them all talking to each other. I think it's going to be more or less a factor in terms of how well you know your customer, how well you know his vision, and how well you know his business model versus how well you know technology.

CTR: The next question is, do you find resellers ready, willing, and able to aggressively tell them solutions, at this point in time?

DAN: Yes. They are willing able and ready. Most of these businesses, and they really run the range from small, single locations to some of the large, multinational, thousands of lines of integrators, and we really have seen a rapid acceptance of our product in those. First of all, I guess you've seen the IDC report that says we're particularly well suited for the reseller channel?

CTR: Yes.

DAN: So that right there validates my argument. We actually see an upside potential in the VAR reseller channel. And we see the ability for that VAR reseller to garner some of that upside profit potential himself. We offer the product and the ability to install the product in a manner which allows him to target new markets, ones that he hasn't been in, in the past. To target new application integration opportunities, in other words sell into a vertical or an application space that he has a level of expertise in that again meets that customer requirement. To add peripheral support opportunities that he didn't have in the past. We can deploy as either a multi-function or a single function device. There is no change in the hardware, no change in the software; we have the ability to do that. So it really gives the VAR and integrator that wants to run with the product a tremendous amount of opportunity to make money.

STEVE: We're recruiting channel partners and it's amazing to me, whether I'm talking to a reseller, integrator or a large commercial distributor in the country, that the smaller resellers and integrators are looking for a NAS supplier, but the large distributors have actually formed NAS divisions, which gives credibility to the momentum that's going on. And they're looking now for vendors to put into this NAS division with NAS specific programs and things like that. We're getting an increase in international distributors now that want exclusives for NAS devices in their country. So it's really becoming very visible out there in the channel. There's an absolute distinction between NAS and whatever else they've been selling up to this point. Which is entry-level servers.

CTR: How do you walk resellers through the technology maze?

DAN: We have an answer, it's an easy one because we have a quick install guide We have a quick install guide and it's a program we call BizNAS that really is a solution for the reseller that wants to walk through that technology maze and wants to be able to resell our product. It includes education. It includes a certification program. It includes promotions: a way for them to make extra money. It includes vertical support and includes solution selling. So we're looking, again, for that reseller that's willing to make that commitment and grow with the marketplace and make a profit.

CTR: How are you getting new resellers? How are you keeping them? How are you different from anybody else?

DAN: We are sort of the new guys on the block, but obviously we bring a lot of history and we bring a lot of expertise to the NAS marketplace. We also bring a very clear product roadmap with substantial technology built into it and a very clear operating path to the next level of technology. We think that will clearly segregate our reseller base from the guys that sell repackaged, price sensitive, low margin, let-me-add-a-disk-drive-to-the-network kind of technology. We support AIT, EDS 3 and 4, CD changer technology, NFS, UDF 2.2, RAID and RAID storage technologies, DVD-RAM, DVD-ROM, CD-R and CD-RW, and disk drives, and all of the technologies that that encompasses. We interface through the same piece of software, through the same piece of network support software, and through the same operating system. It really gives the reseller a broad approach to selling into his niche opportunities, and that's what will keep them and that's what's going to attract them to the company.

CTR: Where do you see your horizon?

DAN: Obviously if you annotate EMC as the 800 pound gorilla and NetApp as the 400 pound gorilla, we'd probably like to be a 600 pound gorilla. There are definitely some tough business models out there that we would like to emulate, but we have our own substantial opportunities going for us. We got a very solid management team. Most of us have been in technology or the storage/storage management business for most of our lives. We have a very consistent product development philosophy inside the company. We have a financially sound business model. We're not out to educate the public and we're not out to lose money. And we're going to continue to advance our core technologies which we think are pretty applicable to the industry and to the state of the business right now.

CTR: StorLogic supports tapes as well as disks, NAS is historically disk intensive; where do you see the role of tape in NAS?

DAN: I don't think tape has yet been completely defined, especially in this marketplace. You do need to have backup, you do need to have snapshot, you do need to have, in some cases, roll forward and roll back. You do need the technology to preserve and to archive and to manage the data. Personally, I see HSM and hierarchical storage management and tape coming a little closer together. Right now data protection--that's a key corporate responsibility for the IT manager. It's his job to make sure that what we say, what we store is both protected and available. So we really kind of see a convergence in that and the ability to manage what's underneath the data structure coining together quite quickly, probably in the next year or two.

CTR: What's your take on Linux in the Microsoft space?

DAN: We support Linux from our management application and from our NFS capabilities. We just don't see Linux malting major inroads into Microsoft's dominance of this space. We don't see that for a number of reasons. NAS and eventually SAN need to be successful and for them to be successful you have to have acceptable profit generating applications that can be ma over top those operating systems. You just don't get that with Linux. You do get that with Microsoft. And Microsoft has become a dominant player and obviously Windows has driven that, but you just will not see that anytime soon in the Linux environment. You have to have a lot of interoperability between various clients and platforms and you have to be able to make a profit with that. I just don't see a model Out there with Linux where you can do that. The developer or the manufacturer, whoever puts the product out on the street, has got to be able to add value and be able to sustain his IP, his intellectual property, over a given period of time in ord er to recoup his investment. With Linux, you just don't have that clear an opportunity to be able to do that and it's just not been proven that you can sustain that IP over a given period of time. Just look at the losses that the Linux companies have suffered.

CTR: Okay. Your NAS appliances scale from 18 to 730GB and from SOHO to enterprise, Where do you see your greatest opportunity?

DAN: We actually see our greatest advantage in what we call internally the middle market. And that is above a Procom or an HP SureStore product and above the competitors that we consider to be emulators and not innovators, yet it's below the full support that's required to put a net app appliance out on your network. But it has all the software architectural and scalability that's typically inherent in an EMC sale. So we provide the ease of use, the management, the cost of ownership that we're hearing now in this environment, yet we do scale in terms of massive storage and massive application space, up to say an EMC level. When we designed the product, we had really five requirements that we outlined internally before we ever put code to paper. First we had to be scalable in terms of storage, RAID, and other storage topologies. Secondly we had to offer a software management application itself that was scalable in terms of not only providing the file system and the file system management routines that were req uired, but also the ability to manage the application space that typically resides on our style of NAS appliance. We had to provide a tuned NAS appliance that seemly integrated the hardware and the software. Our box is a closed box. The end user doesn't open it up and add cards to it. It is a very small, tightly bundled package. Our motherboard is only 3.5 inches by 7 inches long. So it is tightly coupled with not only the software that we manage that by, but the RAID engine, any safety or SES functions you have in the enclosure, the network topology, and the 10 requirements. And those are all balanced to the function that that appliance is designed to accomplish. We also provide our SAN ready architecture that is an architecture that was designed to support our plug-in technology, accommodate tomorrow's performance and functionality requirements today without you losing value in that product, without having to replace that product. We expect to meet tomorrow's market demand with substantially the same produc t only with new input from fresh technology. Finally, we design the product to sell primarily in the VAR and the reseller channels and distribution. So we made it a requirement in not only our internal marketing documents, but also in our product design requirements that we would make it for a reseller very easy to do four things: penetrate new markets and spaces and strategies; to provide those application integration opportunities, that we spoke of earlier; to provide peripheral support opportunities in peripheral areas that he wasn't supporting at the present time, and to allow him to integrate the product in such a way into vertical software markets that he is familiar with.

CTR: The next one has to do with multifunction versus dedicated function. Are your customers using all StorCom.e's capabilities?

DAN: The product is being located in everything from hospitals to RAID enclosures, manufacturers to investment banks and small, local, standalone offices. It really lends itself well to both scenarios: same software again, same management utility. It's just a matter of where the VAR or the integrator puts his focus. With our product you have a full NT plug-and-play functional file server, without the management overhead that comes with NT 4.0. You have an application device that comes with IIS and web hosting capabilities and you have a departmental or workgroup NAS appliance that comes with all the robustness, all of the fault tolerance, and all of the scalability that is required for those markets.

CTR: How do you communicate those things to your VAR, like so-and-so's using it for this and somebody else's doing this, so you increase the ability to understand?

DAN: Because of our background in the business, we understand that education is a continual process that goes on with any type of product development or solution sales process. We are set up to really help share data with our BizNAS partners and to educate them and to spread those stories. Internally, we have a system set up for them on the web. We're in the process of giving them as much support as they're willing to handle.

CTR: Finally, what was it like--a small company working with the giant, Microsoft?

DAN: I think the relationship, from my point of view, with Microsoft has been very good. Obviously, we talked about EMC being the 800-pound gorilla, I wouldn't want to characterize how big I think Microsoft is. I think the untapped potential of that company is just tremendous. It's been interesting to work with them. It's been very good for us as a small company to be able to leverage their name in association with our product. It gave us--obviously we were, as we mentioned before, part of the early beta group with NT embedded--some opportunity. It really gave us time to market. If you go back and look at where we came from and where we started with this product, we laid out in February or March of '99 an upgrade path from our existing software to NT embedded. We really didn't start writing software code until about August of '99. So here we are just a little over a year later with really a pretty functional product with a lot of depth and a lot of breadth and we think we're able to do that because: one, we w ere working in a Microsoft environment that gave us the ability to develop in a very stable, very well accepted file system. It helped us really maintain an absolute focus on our intellectual property. It helped us leverage their expertise in file systems and network topologies and bring that down to our level. It gave us instant application

support. And it gave us the ability to develop this SAN roadmap and our SAN-ready architecture via the use of their file system. So it really was good for us. It has expanded our technological opportunities significantly and it has expanded our business opportunities significantly. So, in my view, it's been a great relationship and it can only get better.