 So we are back, we're back at Procota Live 2014 from the Santa Clara Convention Center, Santa Clara, California, the heart of Silicon Valley. My name's Jeff Frick, you're on theCUBE, where we go out to the events, we find the smartest people in the room, we get them on theCUBE, we ask them the questions that you'd like to ask them, separate the signal from the noise and bring you the information to our community. So I'm joined in this segment by Vinay Jusari, co-founder and CEO of Several Nines. Welcome to theCUBE. So Several Nines, you can't see it over at the booth there, they got all these mylar balloons, nines are floating all over the place. So it's not like two nines or four nines or eight nines, it's just several nines. Several nines, absolutely. So basically, people want more nines, they go to these conferences to hear about how they can get more nines, well we've got them on the roof, lots of nines. So talk about, give us a little background on the company, what do you guys do, what are you building, why are you here? Okay, yeah, so what we do is automation and management tools for database clusters. And what we've seen is over the past, I mean we've been working for the past 10, 15 years with database clusters, and recently there's more and more clusters being put in production, but the problem is that they are pretty difficult to operate and run, and what we do is really we automate all that. So operational people, they come to us to deploy their clusters, and we also have management tools to help them sort of add nodes, remove nodes, clone their clusters, do upgrades and all that stuff. And the team behind this has been around for about, I mean we've been working in the MySQL kind of community for the past 10 years or so, and we co-founded several lines three years ago. Okay. Today we have just over 80 customers, paying customers. Oh, great. So it's going pretty well. So again, just to make it clear, we're here at the Prokona Live, it really focuses on MySQL, but your guy's solution recognizes the reality that people have different database infrastructures for different applications, and so you support kind of a unified control plane over multiple clusters of different types of databases, right? So. Absolutely, and what we see is that people, I mean the first thing is they're taking their databases, they might be deploying in multiple data centers, and what we've also seen is there's many MySQL users where they're putting chunks of their workloads onto different databases. So for example, we support MongoDB, shorted clusters, replica sets, and through our tools, people can actually have one pane of glass to actually manage their clusters, whether it's MySQL or MongoDB or MariaDB, and that's across multiple data centers, geographies, everything for one pane of glass. And what were they doing before? What are they doing before you guys come into room? Before it was very difficult, lots of homegrown scripts, a lot of scripts, I mean, people building their own kind of frameworks to try to do that. Now if you're big, then probably you can put some people to do that, but in general it's pretty hard, because you have to go and program specific scripts to do a lot of things, and what we do is it's one unified interface for multiple databases, and it allows people to do that. So are most of your customers startups, are they enterprises, are they service providers? What's kind of your customer set look like? We have the whole full range, actually. We have everything from large telecom customers that we're not allowed to speak about, to people like large IT security companies. Actually, AVG.com, that's one of them. They are listed on NASDAQ, and we have people like Ping Identity, very big in the US, and we also have very small startups, so it's all over the place. 25% of our base is Telco, and another like 15, 20% is like e-commerce, gaming, web, and then we have a whole range of other type of industries. So talk a little bit about how the move to the cloud and the explosive growth of mobile have impacted your customers and the demand for your services and your business. Absolutely. So a couple of things. I mean, the interesting thing with the cloud is that it's very cheap nowadays to actually throw a service out in a data center, and the problem is that with more and more clouds around, these VMs that people deploy on, they tend to be a little bit ephemeral. They kind of, instead of doing bare bone, if you're using cloud servers, they might go down at any time, so you need to understand, I mean, you need to architect for failure, and that's one of the main things you need to do with the cloud. And then on the other side, when you talk about big data, there's a lot of data, there's a lot of kind of things being collected, and you have to manage many servers. So what's good for us is that we help people architect these database clusters on commodity sort of VMs that are not very reliable, so we allow people to build a very reliable system, resilient, across multiple data centers, across different clouds, where the actual availability of a single cloud server might be not very good, but on the whole, the system has many nines up and running, and the same thing with our focus is clusters, so we don't do single instance servers. Our focus is help people deploy systems with tens of nodes, and that's basically what our tool sets are built for. Tens of nodes or hundreds of nodes and thousands of nodes? It could be hundreds of nodes as well. I mean, just to be kind of honest, I mean, there are some companies out there doing hundreds or thousands of nodes, but then there is many, many more companies that's doing tons of nodes, and that's where we see a lot of people using us. Sure, okay, great. Now it's great, Randy Bias, the CUBE alum, one of our favorite guys to have on, had a presentation about comparing, I'm going to screw this up, Randy, pets versus cows, and you treat your pet different, which is a classic kind of single server environment versus if you've got, shooting me called cows, you'd be livestock, where you're working with the cluster, and the greatest slide in the deck is, software fails, hardware fails, people fail, so really designing for failure at the beginning of the process enables you to have these types of system to have many, many nines, so I mean, are people getting out, are they designing for failure on the front end, or are they still kind of babying these things and trying to take care of it with backup and other things as opposed to kind of a continuous state of a system as opposed to worrying about the individual components? Well, I think people are, people are actually getting into clusters and actually looking at those as systems of many, many components rather than having one, because if you look at how much, cloud adoption there is, all the workload that's being thrown into the cloud, all the CIOs looking at the benefits of utilizing that, so there are a lot of economics out there that's actually forcing people to consider that, and what we've seen is, look at the interest, for example, for Prokona extra DB cluster, right, and for Galera, I mean, there's a massive interest out there, many companies are trying it, and many people have been in the mindset of having maybe a few servers or like a simple replication setup, but I think that we've seen since last year there's a huge push to actually go all on into clusters, and that's, I mean, we've got tens of case studies on our website, and we see a huge amount of downloads, basically. Because in theory, everybody should be on a cluster, whatever a cluster is defined in terms of the underlying infrastructure, whether it's one real box or two real boxes or in real boxes, but it just seems like a much better kind of design philosophy and implementation philosophy. Yes, and what we've seen is, I think the IT world has matured also in terms of clustering, because I was out as early as 2001, actually, selling clusters and working with clusters back at Ericsson, and then I joined MySQL in 2003, and that was, we actually, we sort of visited a lot of people that actually were considering clusters, but you could see how this technology went from being pretty complicated to like around 2008, 2009, there's a whole bunch of companies that actually came to life. We talk about MongoDB, Cassandra, we talk about clusters of Hadoop nodes, so it's been a shift in the past few years towards more and more clustering, and I think that the technology's getting built, and what we want to do in this is to actually have the operational tools to automate things based on these clusters, so that's kind of where we see the market is lacking, and we probably would need a couple of more players like ourselves out there to actually build those, because I think that that's where, there is a huge market for this, that now that you have multiple cluster database types, now people need tools to help them deploy them, manage them, scale them, monitor them, that's kind of where we are heading as well. And what do you think happened that was a tipping point that kind of flipped the switch more towards a clustered point of view than not? Was it just simply the technology wasn't ready? Was it a mindset within the IT world? You know, what kind of flipped the switch, or did some fast mover kind of push it over the edge, and everybody else had to slowly get caught up? Yeah, that's a good question. I think that, I mean, what we've seen is, in terms of the MySQL world, if you go out there, you would probably find 40, 50 or more different types of clustering, right? There's many, many different types, and I would say Galera has really changed things, in a way, because it's fairly easy to work with, and it sticks to InnoDB, right? So, everybody, most people out there use InnoDB as their storage engine, so there's a lot of things that you actually stick with, even if you move to like a PXC cluster, you don't have to change your schema, maybe some modifications, so that kind of helps people move to a clustered environment, and then the other thing that has changed, I think, is that the IT industry, more and more workloads are being placed in the cloud, right? I mean, we know how Amazon is growing, and we've got users that actually go and sort of use this environment, but in that case, they have to have some kind of clustering, because in that environment, if you're not prepared for failure, then you will fail, and you will fail big time. Does your product work with a hybrid solution with some of my resources inside Amazon, and some in another location? Absolutely, so basically, today, you can use our product to automate deployment and management of your clusters, whether they are on-premise bare-born machines, or on Amazon, or in a private sort of open-stack environment. So these are the kind of free-main things that we do, and it's very easy to, basically everything is presented in one pane of glass, so you could have clusters in Tokyo, Japan, so Paris, London, US, everything is presented in one place, plus the fact that obviously we do also multiple types of clusters, MongoDB and MySQL. So I want to shift gears a little bit. You're a founder, you're a CEO, you've got limited resources like everybody does, you have to put those resources to bear, and you're playing in an open-source space. So talk about kind of the strategic vision that you have in terms of playing in an open-source world, but when you've got to allocate resources to, are my engineers working on contributions to the open-source project, or are they working on ancillary stuff that wraps around it that's our proprietary code? How do you prioritize that? How do you kind of manage in that world where you really have kind of a split personality of priorities, or is it a really split? Yeah, I think it's, I believe that open-source is a great way of distributing your product, and I think that the nice thing with it is if you use it as part of your development organization, as part of your development process, it can be extremely powerful. So like for example, if I take our company as example, we've just completed three years of operations, we have a great product, we have more than 80 paying customers, we're going to hit 100 customers by this summer, and that's all self-funded, right? But what we've managed to do is really leverage this community out there, because we come out with these versions every six weeks, and a few features at a time, and we talk to users, we really help all of our users all the time to actually do these things that we're supposed to do, and we listen to them, and sometimes we do things that are not really in our core product, we help them with some other areas as well, we do a lot of sort of education, and that's really, it kind of helps us because we are kind of working in tandem, this community is testing our product, they're telling us what they don't like, and then we're kind of building new features, and then fixing the problems, and we're kind of interacting in that way, and what we've seen is that we've been able to bring a product to market really fast, good quality that people are running in production, and the product, if you think about it, it's a management product, automation product, for mission critical clusters out there, right, for telco companies, so this stuff usually would take much longer to build if you were doing it in a proprietary environment, even if you have like 10 times more engineers and more QA, I think it would have taken much longer time, but I think we've been able to squeeze that time and do it really fast just because we've been able to work in tandem with the whole community out there and move things forward. So what is the model, right? Obviously you're open source core, but do you have proprietary code wrapped around that, do you have extensions that sit beside it, what's kind of your business model, because clearly you're going three years, you've got 80 customers, you're self-funded, congratulations, the best funding is revenue. So how is your structure in terms of a business model? Right, so we have some stuff which is open source, and then the main product itself is closed source, but it's freemium, so some stuff is free forever, and then if you pay, then you get all the added enterprise features and support, right? So we have a combination of other open source products and free services that we offer, and that attracts a lot of people, and then some of them decide to use our management product, the free version, and then finally we kind of sort of convert some of them to actually use the paid version, and today we have over 7,000 organizations that actually are using our products. Okay, so 7,000 organizations use some combination of the free products and then 80 paying customers that are set up to the enterprise version or however you package it. Absolutely, and I think the kind of difficult thing is building a product which is freemium is kind of hard because it needs to be very useful so that it does attract people and people get value from it, but you can give the whole farm away, because otherwise they wouldn't pay you for anything. So it's kind of a balancing act where how do you make it useful enough, but how do you actually have value in there that people are willing to pay for? Right, so we're back to the farm again, right? If the milk is free, why should I buy the cow? So let's shift gears a little bit. So we're here at Percona Live. The show's been going on I think for 10 or 11 years. Percona took it over from O'Reilly I guess three years ago, so why are you here and what's kind of the vibe here on the floor for the folks at home that weren't fortunate to be at Santa Clara Convention Center today? Right, so we are here because obviously we are, I mean I've been coming to this show for the past 10 years from MySQL, and I've seen kind of how things change a little bit after MySQL kind of got bought by Oracle and then Percona took over, and I see that things are actually starting to go up again. That's my impression, that's my personal impression. And I think the vibe here is great. I think we had a great announcement with this web scale MySQL, with some of the biggest largest. It's a new open source project, right? Absolutely, and all this good stuff is in the MySQL community. I don't know any other database community out there that's actually having all this innovation, all this activity as this community. So we want to be part of it obviously, and that's the biggest open source community that we know about. And obviously we're also keeping track of a lot of other developments in the NoSQL space, right? And we've also seen a lot of Hadoop there in this conference, so I can see that it's very, very, this is a good reflection of what people are interested in because you have people out there with massive installed bases of MySQL, and they're also wanting to leverage new technologies, complement MySQL, and to be more competitive. So we've learned a lot, met a lot of nice people, great people, knowledgeable people, and really looking at how do we kind of improve now that we know these things. And what does your guys play with Hadoop? So we've been looking at it, we don't support Hadoop today, we've been looking at it. I mean, we get asked by our customers to actually support different other databases like Hadoop, Cassandra, et cetera. And obviously being a small team, we can't do everything. But definitely we, there is definitely a good play with MySQL and Hadoop, so that's to be determined. Okay, fair enough. Yeah, it wasn't on your list. I was looking for the elephant, I didn't see the elephant. So we come back for Kona 2015, we get you back in the queue. What are we going to be talking about that change? We might be talking about free X more customers. I mean, we announced this great partnership with Prokona. We actually, Prokona sells extra DB cluster and they bundled our enterprise products. And I mean, I think that's going to be huge for us, for our business. So really looking forward to sort of increase the cooperation and get even more footprint. And this is going to be great for all the sort of customers from Prokona as well, all the value they're getting. So I think that we'll come back in a year with a much bigger customer base and probably another NoSQL database added to the zoo. Okay, good. Added to the zoo. We're really hung up on animals on this segment. I don't know why we went down that path. But today, Josary, co-founder and CEO of several nines. It's not four nines, it's not five, it's several nines. Thanks for coming on theCUBE. We look forward to meeting with you again next year and hopefully having a lot of those things come true. So, Jeff Frick here on theCUBE at Prokona Live 2014, Santa Clara Convention Center in the heart of Silicon Valley. We've been here for two days of back-to-back coverage. You can catch all of the videos on siliconangle.tv. Of course, everything is always archived on YouTube. It's youtube.com slash siliconangle. Go to siliconangle.com for the latest tech news and wikibon.org for free open source research. Again, Jeff Frick signing off for the crew from Prokona Live. We'll see you at our next event. Have a great day. Thanks for tuning in.