 From London, England, extracting the signal from the noise, it's theCUBE, Cover, Discover 2015. Brought to you by Hewlett Packard Enterprise. Now your hosts, John Furrier and Dave Vellante. Okay, welcome back everyone. We are live here in London, England for HPE Discover. This is SiliconANGLE, theCUBE special coverage. This is our flagship program. We go out to the events and extract the signal from the noise. I'm John Furrier. Dave Vellante, our next guest is Vishmal Shan, Senior Director of Product Management and Marketing for HPE Enterprise Storage. Good to see you, CUBE alumni. And we have Kai Viljanen, Telus Soranera. Welcome to theCUBE customer. Thank you, welcome to be here. So I got to ask you, what's your take on HPE Discover right now? Because storage has always been active at the show. But now it's part of almost every solution out there. What's your assessment of this year's new HPE Discover? Yeah, it's mostly as you already said in here about when everything is integrating and connecting together. So for example, Synergy stuff and so on. That was very interesting. See the demos on there and so on, so. So Vish, I got to ask you. I saw you on the plane coming over from the US. Yes. The European market storage. Talk about some of the dynamics here. What's compare vis-a-vis the US customers and what is the European customers looking for? Is there anything different? Can you compare and contrast? Yeah, absolutely. The storage market in Europe is very interesting because it is broader in terms of the number of countries. And so you will see a very specific set of requirements. So for example, in the US, asynchronous replication is common because of the distances. In Europe, synchronous replication is common because the distances are not so large. You also have a combination of large customers, large globals like Telia Sonara. But you also have a lot of smaller when you compare them to the US customers, but still very much on the leading edge of deployments. So I wonder if we could maybe get some more information on Telia Sonara. So you guys are a telco. Yeah, it's a- With a huge amount of data. Enough data, let's have it like this. Yeah. Yeah, Telia Sonara is a fifth largest telecom communication company in Europe. And at the moment, we have something like 31 petabytes of raw capacity in our infrastructure and source data or the written data somewhere around seven petabytes, so. Okay. That's a lot of stuff. And your infrastructure, can you describe it a little bit? Your- Yeah, storage point of view, of course. So at the moment, we have standardized it and it actually took two years. It was quite big task. And then 80% of that stuff is built with 8-speed 3-bar technology and that's our shared block services. And then in the file side, we are using a NetApp technology. And also, of course, we have a cloud infrastructure, which is like a 10% of this net capacity. And, okay, and you've only, when did you start moving toward flash? First evaluations did start like a beginning of 2013. And then, of course, the case was very different back then because everything is moving so fast at the moment. Okay, and so you started with the high performance? The- Yeah, the latest- The high performance is mostly, yeah. Okay, so you're replacing high spin speed storage, spinning devices with flash, is that right initially? Yeah. And they were more expensive at the time of flash, right? Or no? Very expensive. Okay, flash very expensive. Yeah, at that point of time. But you still went down that path. How did you justify it? Because we have like systems which require one millisecond response time, or otherwise you will lose your business on there. So then how to say the business justifies it over those infrastructure costs. So that is a 31 petabytes? Yeah, at the moment. Roughly what percent is flash? 240 terabytes. Now we have all flash infrastructure, but then of course we still have like SSDs on the standard infrastructure, almost one petabyte. So we have still a lot of work to do. Okay, so still dominant with the traditional spinning disk? Yes, yes, yes. And what do you see from a cost standpoint, from a customer perspective? I'm going to ask Vish what he predicts. What's the cost curve look like today relative to flash and spinning disk? Yeah, if we are just talking about to change everything to all flash, the class A production data. And now when all of the new compaction techniques are working and then you calculate for example, the administrator cost that you save on there on all flash, I think that we are pretty near on a threshold to change everything. So you're headed, Jordan, you're on a path to all flash data center. Yes, yes. Vish, how common is that in your customer base? You know, the all flash data center conversation is happening. I think there are variations of where people are in that life cycle. You know what the customers are telling me, Dave, is there are sort of three things they look at. They look at clearly RPEX and CAPEX savings when they go to all flash. They look at, or they get productivity savings. And then some of them are even reporting business transformation, right? Where they can evaluate more data, make better decisions in their businesses. So, Kai, what kind of business outcomes have you seen? Obviously the performance impact early on. Yeah, yeah. If I can tell like a real business benefit in here, it's a case, for example, when the customer goes to the help desk, then the help desk resource has to make some kind of a database query and the response is instant and he or she can reply to the customer and you have a happy customer. So that's the benefit on there and you don't lose the customer, for example, because of the bad service or whatsoever. Okay, so that's having an impact on your business. Have you, talk about your development environment. Are you doing lots of application development? You outsource most of that? Yeah, that's the word. We have outsourced a lot of that stuff. Okay, all right. So that's sort of project-based activity. And then, so should you talk more about what you expect over time? You headed to the all flash data center, but today you're predominantly spinning desk. How long do you think it's going to be? I mean, take us through that, your forecast, your five-year plan, if you will. What's it going to look like? Well, because it's a depending, still so many things, not just about the cost and technology. In a bigger companies, it's also like a politics and whatsoever and so on. But anyway, technology point of view, hopefully we are there in like what, at least within the next three years or something like that. That's true. Yeah, but as I said, this is for a class A products on data. And still, of course, there might be a case is when, if you have purchased RA like yesterday, it's pretty hard to justify to make it all flash tomorrow. Yes, and what do you see in terms of the reliability, the whole MTBF, the failure rates, et cetera, flash. Yeah, that's good that you brought that up. That, of course, reduced the administrator cost of the infrastructure. So when you have like, you don't have a similar failure as, or there is less failures and less often. So you don't have to administrate your stuff. And also specifically with the flash in here, that it's more of less like, you will allocate and forget style capacity. So you don't have to work with the OS guys, with the database guys, and we can tune all of the time where I place this data and et cetera. So that makes it, how to say, more cost effective, of course. Okay, so you're spending less time tuning, which makes sense. What do they do with their freed up time? Anyway, we don't have too much free time. And hopefully that time is moved from this to proactive work. To make everything more better already in infrastructure and so on, the planning work and this kind of stuff. Help them grow the business as opposed to just keeping the lights on. So, Vish, let me ask you, so taking Kai's example, how would you sort of summarize where he fits in, you know, the broader customer mix that you talk to that you see? So, you know, I would say Kai's experience right now is very typical of the many customers we have deployed with Flash. I hear consistently that they're surprised by the productivity gains, where, you know, it's simpler and easier, less intervention, and so that's a nice pleasant bonus. And I think, like many customers as well, they start off on the performance curve and then that goes better than they expect and then they move it to more mainstream use cases. Why HP? How'd you come to that decision? When we did do a lot of evaluations, including POC stuff, the demos, this kind of stuff, there was a couple of benefits in there because I can't say that they were much more better in the performance or whatever. It was about the same line on there. Everybody could deliver very good performance, but then the other things in here, such as that they have same administrating on the Flash than in other stuff in our infrastructure because it makes like a multi-country cross-border organization, it makes it much easier to administrate everything and much more, how to say, fast and cost-effective on there. And also in here that they were, from the beginning, they did have all of the same storage features that they have on a shared storage such as a snapshot, remote copies, et cetera. And for example, all of these new players in the beginning, they didn't have that. They didn't have a very good performance, but they were missing these kind of traditional storage features. So that was a big factor. I mean, that's a big part of HP's marketing is we have the fully baked three-par stack and you're saying that was a major factor in your decisions. Talk about more why. I mean, because you don't have to change your processes or comfort level. Was it a reliability factor, sleep better at night? Yeah, of course it was mostly about that that when you have like just one group of people working and administrating like in several countries, several DCs, you want that everything is built quite the same way. It's standardized. And as you said, it makes the processes more fluent easier and you don't have to have very many different processes for example, allocating stuff or whatsoever. You know, I've been talking to three-par customers for a long time now. And I kind of gave up asking, but I'm going to ask you again. So I used to ask, you know, what did you like? You know, what would you do differently? The products were so transformational. People always loved them, but I'll ask you, because it's relatively new, all flash. Anything you didn't like, anything you'd like to see improve, what's on HP's to-do list in your opinion? You can't say cheaper. Yeah, because that's the stuff that everybody has to do. We're working on that. Yeah, yeah, yeah. I guess that they started pretty well because they did do like, as I said, these traditional storage features also on there. So I can't say anything which was done, how to say, very bad. It's the same answer that I always get. I don't need to say anymore. Let the customer do the talking vision. That's what I always say. My final question is, what are you looking for in the future? A lot of this show is predicting and teasing out the future, the machine, software, management software, a lot of new stuff coming around, all kinds of storage, protection, all kinds of great software. What do you guys have in terms of requirements? Is there a corner that you'd have to look around to prepare for the future? About the future, if you're talking about- Maybe next year. Yeah, but longer term. Yeah, well, of course. Three to five years. We want to move from the technology-centric to service-centric provider in our side, and our infrastructure has to change on there. So therefore, of course, we need that new kind of IT, or next generation IT from the vendor, and also how to implement that in the service catalog, our services, and so on. So I want to move on that, that we are selling a service, not IT, to forward in here. So that's like a combination of things. And also one thing that make it, how we can handle more efficient is because we have a lot of traditional, we have cloud, telco cloud, what's our big data infrastructure, so how to orchestrate that, operate this, everything's all to say. What would you say are your top two or three challenges as an IT organization, not necessarily storage-related, but security, digital transformation cost? At this point of time is from moving from the legacy infrastructure, so this cloud and so on, because as a telco, we have like not one legacy, we have even like phase two legacy stuff, even like 10 years old stuff and so on. So it's a lot of struggling to migrate this everything on cloud native stuff, for example. So agility, and then are you looking at things like NFV as well? Yes, actually, we have telco cloud build top of that already, so. Open stack, any open stack activity? We're looking for that. Looking for that too, okay. Well, thanks for sharing the insight on theCUBE. Appreciate it, another happy three-part customer, HP. You guys doing great, as always. Testimony Dave tries to get something out of it, always. We've been trying for 10 years. Thanks so much for joining theCUBE. Thank you, Dave. Thank you. Bringing you more action live from London, we'll be right back with more after this short break.