 Live from the Moscone Convention Center in San Francisco, California, it's The Cube at Oracle Open World 2014. Brought to you by headline sponsor QLogic with support from HGST, violin memory, and MarkLogic. Now, here is your host, Dave Vellante. Good afternoon from San Francisco, everybody. This is The Cube. The Cube is a live mobile studio. We go out to the events. We extract the signal from the noise. This is our fifth year at Oracle Open World. This year, a lot about cloud, a lot about hardware and software engineered together, but we're going to talk about an emerging trend around data, copy data, copy data virtualization. We're going to talk to folks from Actifio, or folk in particular. David Chang is the senior vice president of products and he is also a co-founder of the company, David, welcome to The Cube. Thank you very much. It's great to be here. So, David Floyer and I, and some of our other Wikibon colleagues, we went in, we've been sort of following Actifio from the early days. We know some folks over there, but we went in for a deeper dive a couple weeks ago, right, David? And got into it and it was pretty interesting. But a lot of folks don't know about this space, so I wonder if we could go back to kind of beginning the sort of idea behind Actifio. You start around copy data management, copy data, extending copy data services, and really have extended that now into new use cases. But take us back to the beginning. What was the sort of genesis for the company when you and Ash started it? Absolutely. When we initially started Actifio, unlike a lot of different approaches in terms of taking a great new piece of technology and then finding places to apply the technology, we kind of went the other way around, right? So we did some fairly detailed analysis of the Fortune 500 IT budget. And one of the very interesting sort of facts we came up with was actually quite a surprise for Ash and I, was that for every $10 people were spending on data, actually $1 goes to the primary storage system, and the other $9 actually went to all the copy data use cases. And for Ash and I, we were primary data guys. Excuse me, what's EMC's market cap? No, that's actually why EMC's been so successful for so long. Yeah, absolutely. So that was sort of the interesting observation we had, and it was quite a bit of a surprise for both Ash and I because we thought the primary data subsystem usually is the big dog in town, but in actuality the reverse was true. And that was actually, even though it was true, but it was a fact that a lot of enterprise IT back then didn't realize they were spending so much money on copy data. And this is really the genesis of where copy data came from. Yeah, the sort of Trojan horse of copies. And of course, copies, you know, the whole idea of snapshots when they first came out. Oh, that's great. I can take a snap here, populate a database, do some test and dev, do some backup, do some remote applications. Holy cow, I got all these copies. I have no way to get rid of them, no way to manage them. Absolutely. You evolved, and initially, if I understand it, one of the initial use cases was backup. Yeah, absolutely. Utilizing your snapshot capabilities. We've been talking about a better way to do backup for how long now, David? Five years. You know, the better part of a decade. And so that makes a lot of sense, but one of the things that we learned at our site visit to you guys is test dev is this exploding use case. Yeah. And that along with some other things that you guys are doing in the service provider space. So I wonder if you could talk about the evolution of the use case. Yes, absolutely. So, this case says. So interesting, we have around, probably over 500 enterprise 2000 customer these days. And we just recently did a survey of all of our customer base, and over 60% of our customers are actually using our capability, not only doing backup, but actually repurposing their backup for test and dev situations, right? For taking a copy of their golden data and reapply it for analytics, for copy data, for all these different use cases. So in many ways, our customers were ahead of us in terms of leveraging our capability for this whole test and dev and dev ops space. And naturally we're, as any good company would do, we're following the leads of our customer and bringing down this path. So talk about what you guys built, the architecture, kind of your secret sauce. How do you, how should we conceptualize the solution? Yeah, so one of the primary issues in the past with a siloed approach is that, typically you have an application running on some enterprise storage capability, and then you have a siloed backup infrastructure, you have a siloed replication infrastructure, you have a siloed test and dev infrastructure, and analytics infrastructure, and so on and so forth, right? So if you, not only the cost of actually acquiring this hardware and software is substantial, but the tax of your primary infrastructure to pull this data off of your database, for example, five or six different times, was really the key to this problem that we have in the industry. Now with Actifio, the way we actually do this golden copy or this copy data is we maintain this application, this format, in a application format that the application can directly consume. So we're no longer doing any type of format translation, i.e. from block or files, into a tape streaming format, if you will. And that really unlocks all these different use cases based on this golden copy. So with the golden copy that we now have, we can now apply this for instant restore. You know, if you have, we recently have a customer that has a 50 to 60 terabyte database, and anytime anything goes wrong with that database, there were no other option besides Actifio in terms of making this immediately available, and they would turn their customer to run directly on the Actifio golden copy versus have to deal with their production copy per se. So once we make this till and have this full copy in application directly consumable high performance format, now we can take that golden copy, apply it to all these different use cases at a substantial decrease in cost for our customer set. So this is where the simplicity comes in, the complexity drastically reduced with an Actifio infrastructure. So what are the components of this? So you got a virtualization engine in there, you got a catalog, you got snapshot technology, you've got a platform. Talk about the secret sauce a little bit that you guys have developed. So the secret sauce is really in the architecture. Once you apply the Actifio architecture within a fairly large infrastructure, then all these sort of benefit comes into play. And we next, we married this very capable copy data platform per se into a set of user interfaces, SLA driven architecture that really allow our customer to simplify the management of their application data. Now they can assign a single SLA to a application data that only comprises of local protection, but now you can add primary replication, backup replication, long-term archiving, then sort of construct now a workflow that includes data masking and then virtual copies to anywhere from two to 20, 30 virtual copies, while giving you the full benefit of the full copy with drastically reduced capacity and time to actually do this ETL load. And you've got a metadata capability engine that allows you to understand where the copies are, what versions are where, so that you can essentially have an accurate record of everything, right? And really the other secret sauce is in the deep application integration, right? So again, we have this very capable platform, but we've done the hard work of having very in-depth integration with the likes of Oracle, with the likes of Microsoft Exchange, with the likes of Microsoft SQL. So all these different sort of verticals enable us to apply this capable platform into all these vertical application use cases. And that really is what results in the simplification, the application-centric nature of what we bring to the table. So you're actually affecting the application value itself by being able to put in new versions of software quicker, being able to apply fixes quicker, all of that. You're actually lowering the friction for putting in new functionality, is that right? Absolutely, right. So how do you sell that value? Who do you sell that value to? It's a very interesting question, right? So we typically, the typical people that within our customer set in a large corporation sees value either directly from the application layer or where the application infrastructure meets up. And that's typically the place we need to have that conversation to be successful at an account. You're absolutely right that it's very tricky and it's actually, it's very key for us to be at the right level to have that discussion because the value that we provide is so comprehensive. It all boils down to the location where the infrastructure and the application meets. Right, okay. So there's a lot, storage is one of the things that's really changing a lot now. The advent of flash and other technologies allow you to share a lot more data than you used to do because you, the disk drives are terrible, weren't they? I mean, the access density to disk drives was very small, so you were forced to make copies. Yes, yes. You had to, but is there a role for making a logical copy actually on one physical copy? Is that where you're going with this type of technology? I think so, I think in many ways the advent of storage advances like the SSD storage devices really enable you to put a lot more test and dev copies on the same set of SSD drives. But what actually I think the industry is missing is the ability for that to be done easily at the application tier. Right. So we really enable that seamless transformation from physically going through a workflow process of clone after clones into a single golden copy and then able to map out these virtual read-writeable copies to anywhere you have in your environment, regardless of the protocol that you want to leverage to actually get there. This is really also in an area that Actifio brings a lot to the table. So where does your stack end? So you're creeping up, aren't you, into the whole of the operating system and you're touching large parts of the whole stack. Absolutely. How far up are you going? We touch everything within a modern IT data center, whether it's wide area network, it's high performance networking, storage networking, IP networking, virtualization databases, tier of application on top of that, we touch everything. But we clearly see the delineation as being clearly for us in terms of copy data version production. So what we want to do is enable your production infrastructure to be as fast as possible to support the very challenging application requirements of today. So we want to stay out of that as much as we can. And what we want to focus is really in this whole copy data space where really what we believe we'll bring to the table is this intimacy with the application. Taking the infrastructure that we have and making all the integration available that's possible deeply into the application. And this is really the strategic direction that the Actifio product line is going into. Now, we're at open world, you guys got a big booth, a lot of orange running around. So talk about your play in Oracle. Absolutely. So Oracle is a enterprise database. We have again, 60% of our customers running our capability on very large and very capable databases. We also have a huge footprint in the service provider space. So we see our instantaneous recovery capability that the ability for us to run at near production performance with your copy data as the foundation piece of building a very large database architecture today. You have to have that to survive in today's world in the 24 by seven world that we have today. So that's absolutely a portion that we think we can add a lot of value. The other portion is really the acceleration of this workflow in terms of getting copies of your data to the sort of the test and dev of the environment to do the QA infrastructure. And we think we bring, you know, we recently, not recently, we have a customer that's been in production for about two years now, a very large service provider. And this particular customer has, you know, so far by conservative estimate within a given year, save about 2000 hours of test and dev time in terms of accelerating that database and making that available to their development staff so it could be ahead of their competition. So those are all sort of very tangible, substantial savings that we're able to prove. And that's currently in production. Okay, and then now talk about how a customer implements. They bring in an appliance. How does the, you know, data get populated? Maybe talk, take us through sort of a typical implementation. Yeah, absolutely. So we deliver our capability in both a physical appliance or a virtual appliance format, right? The reason why we deliver in physical, it's all about time to value, right? So we come in within a hardware stack within that first day, literally two or three hours, we could be up and running, protecting your environment and making sort of test dev copies off to the rest of your environment. That's as simple as it gets. It's really about time to value. You want to make sure the champions, our champions within our customers can really justify their active purchase as quickly as they can. It's about time to value everything. So once that installation is done, you know, depending on the infrastructure we're protecting, whether that's virtual machine, VMware, Hyper-V, Oracle, typically give us a credential that we can actually tap into the data and within that first day we're off and running. That's how simple it gets. And you also have cloud-based products, for the sort of mid-sized enterprise. They're relatively new. Can you talk about that a little bit? So that's primarily driven from our virtual appliance. So again, the deployment profile is absolutely the same. We take, except we deliver that in a virtual appliance format. So for the mid-sized customer, download our virtual appliance, install it within their ESX environment. And again, within the first hour or so, they're seeing substantial value. It's really as simple as that. So talk a little bit about that value. So they had the sort of initial beachhead was better data protection. Yes. Right, and so that manifests itself in what? Saving money from less copies, simplified management, less arms and legs, managing all this stuff? All of the above. And what did people do? Yeah, all of the above. What did they do with that freed up time? What did you see in the customer base? I don't think I have ever met an IT person that says I don't have enough things to on my list of things to do, right? So really with the Actifio product, our customer are enabled to move or retask their critical and ever-thinning set of resources onto a set of activities that directly correlates to their business value versus sort of the bookkeeping, making sure things are running. We're so, you know, we're in the 21st century now, right? So just keeping things running is no longer good enough of a goal for the IT organization. It's all about what value you can bring me to enable me to have a strategic advantage over my competition. And Actifio really frees up the people resources and also the compute and the IP resources to allow them to do that, to retask those resources to an element that they can explain to their CEOs in terms of enabling the next set of revenue that need to get from their customers. And Test Dev is really, you know, taking off from what I understand. Absolutely, absolutely. Can you talk about why, how people are using your capability in a Test Dev environment and how it compares to what they were doing previously? We see a lot of customers that have very sort of, I would say existing best practices where they use tier one capacity, right? And clone that tier one capacity however number of times and make these available regardless whether that performance envelope is really required or for the duration of how long they actually need that copy. So, Actifio really brings to the table that knob that our customer can find too. If you truly need a copy because you're running really heavy analytics associated with that data, Actifio can give you an entire clone of that copy very efficiently while not avoiding the ETL process of fully having to recopy the entire size of the database from A to B before that can happen. We can actually just refresh the deltas on a daily basis. And once, so if that performance is what you need, fine. If you need a flexibility of putting on SSD type capacities and make that and make 16 different virtual copies to all your Test and Dev QA people, we can facilitate that. If you want to have just, if you cost cutting is sort of, if cost is the only picture for you, we can make that truly the golden copy and make again 15, 20, 30 copies all give you virtual rewritable information on top of that single spindle set of spindle you have. So again, the, we really, Actifio really leaves, gives our customer the ability to kind of tune that knob to exactly what performance and cost envelope they want to be in. Yeah, so my, in that example, my dummy data, my test data can be today's data. Yeah, absolutely. So imagine a traditional case where you have a separate stack that's doing data protection. A separate stack that's doing replication and a separate stack that's actually doing Test and Dev. Imagine sort of the tax or the load you're putting on the CPU, PCI bus network and through the fiber channel network and finally the tax you're doing from an IO perspective on that high end storage subsystem. With Actifio, you only have to extract those changes once, right, so that frees up all this hardware to do to service the true application performance requirement that your application vendor, your application buys really need and really frees up your infrastructure to service that while you can retask that copy for multiple use case. And that's really where we think we, we can provide substantial value to end user from that fact, from that perspective. Your end goal is to have one true copy of the state of that company and then all the records of what made that happen. Otherwise you have a multiplier effect in terms of the tax in your infrastructure. This is why we're seeing the discrepancy between, most customers, their primary data is actually growing at a fairly reasonable rate. They're doubling, let's say, every 12 to 18 months. But once you add all this multiplier tax of the rest of the infrastructure, this is why the capacity is out of control. This is why your network traffic is out of control. This is why your wider area network traffic is out of control. These are all the fundamental reasons why we are in the state of mess, as you could say, that the enterprise IT have today. And really the solution to this problem is not by bolting another cool solution on top of it, is to dress this inefficiency at the root of the problem. And I really think that's what we really come to the table in terms of having that problem solved. So if you, but also if you've solved that problem, don't you, aren't you able to add more value on top of that by having true data that you can actually compare and contrast and having access to it and knowing about it. Imagine if your test and dev group is actually working on data instead of a week old, two hours old. It actually brings up a whole different perspective in terms of the agility of your business. And that gives you such a high competitive advantage over the rest of your competition that was previously not possible before. Due to the fact that it was the cost it takes to actually realize that capability you want to have. So David, where are you guys at as a company? I know you've got some Rockstar VCs, you raised a bunch of dough, can you maybe give us some idea of headcount, margins by products, whatever detail you want to give us. But where are you at? What's the plan for the future? I can't use the excuse of we're in a quiet period. Yeah, yeah, yeah, right, right, right. Maybe someday you can use the excuses. IPO in the future, I wonder if you could talk about what you guys have said publicly about that. You know, listen, I'm a technology guy. Focus on what's best for our customers and our focus is really to take one customer at a time and delighting the customer. So they either very happy with Actifio and will be great Actifio references and that's really, if you look at about sort of the amount of money we've raised and that's the area that we're investing heavily because we understand that we're in the enterprise game. There's only globally 2,000 accounts that we, first of all, we're interested in getting and they see also substantial value from Actifio. And our goal is to make sure every single one of our customer is happy and delighted with Actifio. And if we can focus on that, I think the rest of this, whether it's IPO, successful exit, I think we'll fall in place. Excellent, all right, David Chek, thanks very much for coming to theCUBE. It was really a pleasure having you. Thank you. I know you're super busy. I appreciate you coming by. Thank you, sir. Appreciate it a lot. All right, keep it right there, buddy. We'll be back to wrap up day two of theCUBE at Oracle Open World right after this. Right back.