 Hi everybody we're back this is Dave Vellante of Wikibon.org I'm here with Atstu, Stu Miniman who is my co-host today. We're live at Edge, this is theCUBE. Eva Helen is here and she's I believe the CTO of Sambolic, the most interesting company that you've never heard of. Am I right? Are you the CTO? No I'm not really the CTO I'm fairly technical. I am one of the co-founders and definitely the president. I know you're an alpha geek. I am. I have written code. So your president, founder, co-founder. Yeah and I hand a lot of the sales side of what we're trying to do at Sambolic. So you're dangerous with all that technical knowledge out in front of customers. So anyway, welcome to theCUBE. Thank you so much. Thank you for making some time to come by. So Edge you're here. You got a shout out yesterday from Ambush Goyal. That must have been pleasing. You shouted back. I did shout back. He kind of asked for that. Yeah that's good. So what are you doing here at Edge and what are you guys doing with IBM? Well first of all we're extremely happy to be here because our relationship with IBM is fairly new. We have a lot of joint customers but those have kind of passed under the radar until now. Now we're trying to basically look at who's actually using IBM hardware to see how are they using the Melio software together with these IBM deployments. But the primary reason I came here was because we're launching a SQL server appliance with SystemX this week. And it's been in the works for a long time and it's a very exciting solution and we're going to bring it to market through partners. Yeah so tell me more about the, you announced that one today or yesterday? It was SystemX did a whole bunch of announcements yesterday and we were one of those. So what we're seeing is, you know, I'm meeting with a lot of people from the channel that are actually selling SystemX and one of the things that they're looking for is complete solutions that can accelerate the SystemX sales. So what we've done is we've taken storage servers and maybe I should just tell you a little bit about what Sambolic does before I start. Let's start there. I mean I know the company a little bit. So we set off camera. You are the software defined and software defined storage. It's back up. So Sambolic, you guys have been around for quite some time. You've built out an amazing robust storage stack. One of the best in the industry. So tell us about the company. Yeah. So we started the company 13 years ago. So we've been around for a long time and we came from distributed computing, working in the telcospace initially. And we've basically continued to develop the software suite continuously as we started and just build on top of what we started from the very beginning. So it's a complete data management suite that we have that consists of three layers, volume management, cluster file system and application clustering. And as software defined storage has now all of a sudden become the hippest of all hype words, it aligns really, really nicely with what we do. So we've not reinvented anything. We've just continued to build on what we had from the very beginning. So those were software only. There's no appliance. There's no specific server that comes with the software. We just install on all the servers in a cluster and give those shared access to the storage to provide high availability and scale out for applications. So in the SQL server appliance, which we're launching this week with IBM, we're basically taking storage servers, the hard drives out of those using the volume manager to stripe across, sharing it out with the file system and providing application clustering for SQL server from the top. So we cover all the three layers in that stack and it's software defined storage because there is no external storage device. Okay, so it's software only, but it's also software, what makes it software defined? So you abstract the underlying hardware. And then you allow granular access to services that are in your stack through an open API. Okay. And then you are sort of device agnostic, right? We're completely device agnostic and hence the reason we haven't looked so carefully at all the types of hardware that our customers are using because it's an array of everything. But we are storage agnostic and brand agnostic. We're protocol agnostic, server and hypervisor agnostic and now also storage media agnostic. And so that's why it's very exciting to be part of all these flash sessions that are going on during these days because what we can do is we can mix and match, we can combine spinning disk and flash in the same environments and with a component in the file system we can intelligently place the data a little bit more advanced than some of the other tiering tools that are out there. Yeah, so you've got a quality of service component that is more advanced, it's somewhat unique in that regard. But because I've seen the demo and you've got a veriting card you've got a fusion card, you've got spinning disk, you've got a melanox, you know switch and everything's going in within Finneband or TenGiggy and you've virtualized that entire infrastructure and then your file system is smart enough to place data in the right place. Yeah. And I think a lot of the hardware vendors and a lot of the technology that IBM has come a very long way. I think all the stuff that they're doing with TMS is extremely exciting and they've really made a huge leap forward but there's still some things that we can add, some value that we can add to that. Talk about that a little bit. Well, specifically in even if you're doing tiering, even if you are talking about the economics of the data and how are you placing it and so on, there's still multiple copies, there's still pools of storage and data that needs to be managed individually, data that needs to be moved around. The whole idea with our product suite is to create one pool of the storage that can be accessed from multiple places. So you're providing the high availability from the file system layer. All of a sudden you have a much more stretchable backend and that can be made highly available on premise in one location but you can also stretch that across sites. So in some cases what we've done with IBM is we'll have multiple locations, let's say two locations with XIV storage from IBM in both locations and then we use our RAID, the mirroring in our file system technology to stripe across those two sites to get high availability across sites as well. So there's all of a sudden a much more flexible backend and we're creating a more agile infrastructure based on the hardware components that IBM and other vendors have. So you've got a full volume management, data management, storage management stack in software, RAID, snapshots, replication, the whole deal. Yeah and I think one of the most important things is how we can scale out because even with the good flash technology that IBM is providing now, there's still not the ability to scale out seamlessly and that's one of the features that we add so nicely. So you can actually scale up to 65,000 storage devices, volumes that are 18 million terabytes and up to 2048 nodes in a cluster. Our typical customer environments are not that big but that said you can still scale out very very easily. And how automated is it in terms of making changes, dealing with hotspots? Does the system predominantly do that? Is it so-called autonomic? Is it really simple to move stuff around? Well, first of all it's very intuitive to use the software. We've spent 13 years on trying to develop something that's very easy to use. So the installation process is basically six next buttons and you're done. There's no training courses that you need to go for multiple days to learn how to use the software. It's extremely intuitive. Then a lot of the processes can be automated. You create settings and you were talking a little bit about the quality of service and the quality of service is something that you can tweak and make sure and guarantee that either an application file or a process is highly available. So it is a very intuitive software and it's really not an extra management layer. This is something that goes in as a piece of the infrastructure and it's not requiring any extra overhead or anything like that. So we don't affect the performance, which is obviously today one of the most important topics. We do not affect the performance negatively in flash environments. So you must have loved it last summer when VMware came out with the whole software-defined data center and then of course at EMC World you saw the announcement of Viper. We're here at Edge, big theme on software-defined storage. NetApp today announced Clustered ONTAP, Data ONTAP, 8.2 the press release, the whole bottom half of the press release was all about SDS. You must have seen that go, wow, this is great. This is what we do. Well, it is kind of funny how the market is coming to us because we've been at this for such a long time. Now this is still today. I'd like to compare it to fashion. It's not very fashionable to be ahead of your time and that's been the story of symbolic. We've always been ahead of our time and it doesn't really pay off. Now the technology and we are sort of better aligned and we find and think that our time is yet to come and software-defined is still a lot of talk. We can actually do that today and we are doing it today. I don't think that it's going to be over very soon either. This is just the beginning of it. The platform that we have developed is incredibly solid and good and stable and we have over 800 customers using it. Now we're just expanding that to other storage media. I have to agree your comment on software-defined storage is still relatively new especially for the big guys. Back for us a little bit, what have you seen in those customers? What's the driver as to why they've moved to that? What's been compelling to have customers move to this kind of new paradigm? I think that we're seeing the organizations are very siloed. There will be people that are dealing with storage, people that are dealing with applications, people that are dealing with the networking. The ideal customer I should say are the ones that are actually talking to each other where they actually sit around a round table every now and then and talk about what is the future, what is it that we need, where is it that we can save money, how can we actually get better utilization out of the hardware that we already have. I think that the application IT managers and the storage IT managers as they are meeting that's when software-defined storage and software-defined everything is going to become a lot more interesting. We're seeing that it's a process where symbolic typically has gone into customers talking to them about an application, VDI, SQL, file or web serving. Now all of a sudden we get an opportunity to talk to everybody at all the different layers of the infrastructure and they understand the value of putting a software that improves availability and scalability of applications by tweaking down at the storage layer. So are you finding yourself getting into the customer environment through that application owner or line of business and that's moving the storage guys or the storage guys part of the discussion at the beginning? No, the storage guys usually come in later. They have their own hardware, but they're also the ones that get beaten up by the application guys if the performance is not good enough or if they don't have enough space for their applications or whatever it might be. But right now, like I said, the ideal customers are the ones that drag in the storage people already from the very beginning. So let me ask you, so you're now seeing all this everybody come to software-defined, they're trying to extract their function out of their middleware, controllers, wherever it is. It took you a decade plus to develop this software. How long do you think it's going to take the industry? They're using more modern tools, you guys use them now as well, but decade ago it would take longer to build that stack out. You didn't have as many resources as a large company, but as we all know, throwing programmers at the problem doesn't necessarily speed things up. In fact, oftentimes it can slow things down. So what in your estimation, how much of a functional lead do you have? How long will it take? For example, an EMC or an IBM, if in fact they wanted to or an HP to build out that type of storage stack for the software-defined world. Will it be a year, two years, five years? Well, there's still so many layers and there's so many components of it, so I think that we're going to see sort of a gradual increase, right? Not everything is going to happen at once. It's not like we're ever going to be done. So there's going to be a little bit in six months, a little bit more in a year, a little bit more in two years. But I don't think that there is a sort of universal solution that anybody's going to come up with, because customers are always going to want more and more and all of us are just going to try to continue to respond to those requirements. So I don't, I can't say. So it's the depth and quality of what you're doing versus the tick-off item that, oh yeah, we have the snapshot. Well, we have more of a, you know, symbolic as more of a holistic view of the whole data center. What does it look like? What is it going to look like in two years from now, in five years from now? And we believe very much in the way that the public cloud providers have developed and designed their data centers. Scale out. Yeah, with, you know, commodity hardware, intelligence in the software, they've had to design their own and write up their own software to accommodate the requirements that their customers need. And we believe that we can help to take that concept to the enterprise. So you've a hyperscale mentality for your little slice of the world that you're bleeding into the enterprise. Yeah. Excellent. All right, Eva, well, listen, thanks very much for stopping by the Cube. It was great to have you on. It was good to meet you again. Thank you. All right, keep it right there, everybody. Thanks. Thanks for watching this program. We have Jeff Jonas coming up. If in fact, he's still awake. He just got in from Australia. He was doing an Iron Man competition. But he's a fantastic interview. So keep it right there. Right back with Jeff Jonas right after this.