GCJ: Could you tell us about RevStor, who you are, what you do, something about your products, and what makes you unique in the market space?

Felker: Sure, RevStor is a software development company and we've developed a product called iSAN which is a storage software product that allows companies to use existing hardware resources strung together in a grid storage architecture to store their data with increased availability and dramatically increased security.

GCJ: VC-funded? Do you have a number of employees or customers you could tell us about?

Felker: Well, right now, we've been funded through friends and family, so we're at a fairly early stage. We're actually just finishing up our first version of our product and starting to get out there and sell. We have beta customers. Most of our customer targets are in the outsourced IT and infrastructure support arena. So we actually sell to them, they in turn sell to their customers.

GCJ: How many employees?

Felker: We have right now four employees.

GCJ: Storage virtualization. Is that how you're encapsulating what you do?

Felker: Well, it's a buzzword, and like any buzzword, it fits kind of. The problem with what we're trying to do I think is that they haven't come up with the all-encompassing buzzword yet, so ours is a combination of storage virtualization, data virtualization, security, and then of course you have the peer-to-peer grid architecture, which is not necessarily a part of storage virtualization as it stands today. What we've tried to do is take the concept of storage virtualization and make it truly heterogeneous so that you're seeing all of your systems in one unified architecture and one unified management tool.

GCJ: You think the enterprise world is ready for this kind of product, and it is differentiated from the big storage vendors in what way?

Felker: Well, the biggest differentiator between our product and a typical storage product that might use heterogeneous equipment - for example a SAN, a storage area network, or a cluster of servers - is that we store data in a pieced or chunked form, whereas in traditional storage architectures data is stored in more of a silo form. So your data's always in one place. Now, that might be replicated to multiple servers, but the files are always together, they're always there, and they're always consolidated. What we've done is broken it out and taken the files to a truly distributed nature so that no one file ever exists in a whole state on any one computer until it's requested. So that's one of our biggest differentiators.

If you ask me if the enterprise market is ready for something like this, I don't really know about that. The enterprise market to a certain extent - definitely when you start messing with their data, it's a slower moving giant than the market we're going after which is really the small and medium-sized businesses.

One of the things we did when we designed the software was design it in such a way that we could have a pricing structure that could actually be afforded by the small and medium-sized businesses. So that our software implemented at their location is inexpensive, yet provides great benefit, and in many ways some of the same benefits that a product 10 or 20 times the cost provides at the enterprise level, that can now be provided for small and medium-sized business level. The second thing we've done is we've designed it so that it can be implemented in a managed service provider fashion, so that outsourced IT groups can actually implement it and then use it to effect an MSP model, and has to generate ongoing revenues in itself.

GCJ: So focus is SMB, is the solution today - is it a packaged product or is there a services component, how does it work?

Felker: Well, it can be either actually. The package can be installed locally and it is - all you do, you just double-click an install, run through a little wizard like we're all used to, and the product's installed. Very straightforward, once it's installed and started up it becomes a self-healing, self-managing storage network.

The other way it can be provided is as a service. So if this were to be implemented at an outsourced IT group, a value-added integrators group for example, the storage space that is garnered from installing this software can then be rented out to provide for online, automated, off-site backup to that outsourced IT provider's customers. And because of the cost basis for this product, which to a value-added reseller is only about $675 per terabyte of managed storage, they can get a return on investment extremely quickly from the managed service provider model.

GCJ: And you mentioned hosting before, would the large hosting markets be a target in the future?

Felker: Very possibly. In fact, we're trying to get in with some of those groups now because of course they have a lot of customers who already trust them with some of their data, so it becomes a natural extension of their business.

GCJ: And let me just be clear, this is very much a software package. You're not providing the storage hardware as well, right?

Felker: No, you can use whatever you want. Most of our beta testers actually used systems that were locked up in cabinets in their basements because they had outlived their usefulness on the front lines of computing as it were. Our software actually runs on anything back to a Pentium 2 with about 64 megabytes of RAM.

GCJ: Interesting. So the whole notion of virtualization and grid and using all available resources is very much a key component of your messaging?

Felker: Absolutely, that's one of our keys, one of our absolute key components. One of the other big differentiators that we designed for was shared computing. So you're a user, you have your computer, this software can also be running on it and use whatever you're not using. In a similar fashion to how CPU cycles are donated in some cases, are utilized when the CPU's not being utilized by other processes, we do the same thing but at a storage level.

GCJ: I take for SMB focus, it's not a large product that needs a big server to run on and it must be fairly easy to use for an SMB administrator?

Felker: It's extremely easy to use. In fact we designed it such that you don't even have to be a network administrator to use this. You could just be a very simple power user. Once you've installed the software and the nodes start talking to each other - node being any computer on which the agent is installed - you bring up an administrative tool, and everything's laid out right there for you. You can add files, add folders, remove files and folders, and move things back and forth. And the key to it is that it's a truly distributed architecture, so there's no one central point of management. Any node can be the administrative console at any given time, as long as you have the right logon credentials, of course.

GCJ: Right, and that also means there's no single point of failure.

Felker: Absolutely.

GCJ: What about vertical markets within SMB, are there some that you are targeting or have had early success with?

Felker: Yes, some of the markets that we've been targeting are insurance, finance, financial companies, healthcare organizations. It's really organizations that need to retain data for long periods of time and need that data to be secured during its retention process. And those two groups - the three companies I just mentioned, those two are very key groups. One other one is engineering firms and architectural firms, which have enormous amounts of data that are created on an almost daily basis, and need to be retained almost indefinitely.

GCJ: Is it specific to a specific network? So you mentioned these engineering groups, if they've got an office in New York and an office in California, are they able to do that across the network?

Felker: Absolutely, it works both on a VLAN basis, and on a LAN basis.

GCJ: You talked briefly about pricing, do you want to give your base pricing for our readers?

Felker: Sure, for a value-added reseller the price is $675 per managed terabyte, so that's - you put it on 10 systems, they each have one terabyte of free space that's given over to the software, so your managed terabytes are ten terabytes. The product is then sold at an MSRP of twice that, so $1350 per terabyte.

GCJ: Maybe it just strikes me, but that seems like an awful lot of space for an SMB, is anybody in your user community even remotely getting near that?

Felker: Most of them end up at about three terabytes of space that they're able to reclaim. Now, what you have to remember is in order to ensure data availability and data integrity, we have to store everything a couple of times. So while you might reclaim a total of 10 terabytes of space, what we recommend generally is a fault-tolerance setting of five, which means that each file is out there five different times. So really, you've only reclaimed two terabytes of space. So you have to always take your total amount of managed, divide it by your fault tolerance, and get your actual amount available.

GCJ: That makes sense. That also brings up a question of speed, is there any - is there any degradation of speed to the end user, or is this all happening, and they don't even know it's going on?

Felker: Actually they'll notice a speed increase when accessing files, especially large files over this system versus over a standard Windows network share. As an example, one of the tests that we've done is a 700 megabyte, it's a movie file, an .avi file copied directly from one Windows system to another Windows system, so what essentially amounts to a client-server relationship, it takes about five minutes to move the file across the network, 100 megabit network - fairly standard in the industry today. Using our software, it takes about a minute to a minute and a half.

And well, the reason is because there's multiple copies, each client that's requesting a file is able to request the file from multiple sources. So you're not tethered to any one direct network connection or a series of network hops. You're using multiple, all at the same time. In the same flavor as a BitTorrent or, going back a little farther, a Napster would use to increase transfer speed.

GCJ: OK. Now, from a platform perspective, supports Windows, Linux, others?

Felker: Windows, Linux, Solaris, HP-UX, AIX, and we'll have Mac compatibility here in about two, three months.

GCJ: Let's jump into some of the higher-level market things that are going on, there's been a lot of reporting and a lot of fear, uncertainty, and doubt, or FUD, this year surrounding virtualization, grid, SOA, utility computing, all of those - throw in whatever other one you want to throw in there. I'd like to get your take on that, and then what you think about the market and specifically where your technology fits in all of those scenarios.

Felker: Sure, the market as a whole, the problem with the market right now, when it comes to all of those things that you mentioned is that there's all of these terms that mean, when you get right down to a base level, almost the same thing - use the resources that you have in the manner that most benefits the situation. That's really what it boils down to. You talk about utility computing, you talk about SOA, you talk about virtualization of hardware, or data virtualization, it all boils down to that. The problem is that there's so many different methodologies of achieving that that the market is extremely confused right now and extremely crowded with buzzwords and not enough actual, here, look, here's a nice, easy packaged solution that you can just implement and it'll do what it needs to do, and you don't have to know what all the buzzwords mean.

The buzzwords are more important of course at an enterprise level than they are at the small- medium-sized business, where effectiveness and ease of use tend to be the driving factors. And so utility computing, SOA, all the things that you had mentioned are right now being discussed a great deal at larger companies and even large mid-sized companies. But the people in a 10-person shop with maybe one server - or maybe two servers if they happen to be very IT conscious - don't really care about utility computing. They care about how much they're going to have to pay to keep their data available, backed-up, and secure, and that that process is automatic and easy so that it doesn't take away from the focus that they need to put on moving their business forward.

And that's where the real problem - that's where the real discrepancy and white space is in the market right now. The buzzwords that you listed there and I've kind of thrown back are all at the enterprise level. Nobody's brought a product out that targets very specifically what small or medium-sized business need. And what they need is, again, ease of use, cost effectiveness, and it just does what you need it to do without you having to know how it does it.

GCJ: Good point. So based on that, the next logical question to me is in that market space you just defined, who's your competition and how do you stack up against them, or who do you see out there?

Felker: I think as competition for our first product, which is iSAN Pro, are other backup methodologies. This product is designed very specifically for archiving of data and record retention. Very straightforward. The product uses all the technologies that we've talked about to ensure availability and security of that data, but that's its real main focus. So we're competing directly against tape. We're competing against networked cache storage type backup, whether that be a USB hard drive that gets plugged in, or it be permanent NAS box on the network. And we're competing against other online backup technologies that are out there and that are available right now.

What we've tried to do is differentiate ourselves both in the level of security, the ease of use, and in the price that we're able to provide, such that the outsourced IT groups and value-added integrators can actually get into the online backup business for a relatively low cost, and hence pass that low cost onto their customers.

GCJ: What about your customers or case studies?

Felker: Well, one customer that we have is an outsourced IT group, downtown Chicago, and they've got about six people in their shop - a couple of administrative people, and then some techs that are running around, putting out the various fires that tend to happen at companies. And they've implemented an iSAN in their office. They've gone to their customers and actually set up VPN access into their network, which then allows those customers to use our agent to back up to the outsourced IT group's iSAN. And they're actually using it as a backup to the backup. And the reason they're doing that is because of the number of customers they have, they can't be running all over the place doing test restores on the tape that the customers may or may not be changing out on a daily basis, so it doesn't get done. And what's happened is that customers have had data availability problems, or they need to restore data because they've lost data, or they've corrupted it, or what have you, and they haven't been able to do it because they haven't been able to verify that the backups are happening, and verify the reliability of those backups over a window of time, like a year, or even two years.

So what our solution provides them is the ability to have their data replicated. Now it's truly offsite, they know its offsite, because it's at their site. They have the ability to quickly restore the data back to their customers, even in the case of a large amount of data because they can take a USB hard drive, plug it into their iSAN, download the data, and run it over to the customer, which breaks the whole type speed barrier, which is one of the problems of online backup right now. And they're able to operate at a very low cost, a dollar a gigabyte for the backup.

GCJ: What about partnerships? Are there any partnerships you'd like to talk about?

Felker: Right now we're only in talks with companies, trying to see if there is a valid partnership with us and a potentially larger person who's also in the space. But nothing concrete at this point. The one partnership I can mention is to say that the Globus Consortium has been very helpful.

GCJ: So what about next year for RevStor? We're wrapping up 2006, any big announcements or developments or product roadmap features you'd like to tell our readers about to keep an eye out for?

Felker: Absolutely, one of the biggest product roadmap pieces - and there's really two big product roadmap pieces that we'll be putting in place over the next six months for our next product release. The first is live data access. So right now, as I have said, this is for archival, retention - archival data, record retention. The next version of the product allows there to actually be something where you can double click on a file and it'll come up in the associated application. It'll truly be a networked drive. And that'll be out in May of 2007. The second thing that'll also be available then is true NAT bypass - so Network Address Translation bypass. So that you will have a relay server network in place that will allow iSANs to communicate directly with one another and share information back and forth, which will allow for things like inter-company - intra-company disaster recovery types of situations, or even inter-company disaster recovery situations. And then also it'll allow for collaboration between either offices or between companies in a very secure and managed fashion.

close window