 Live from Las Vegas, it's theCUBE, covering HPE Discover 2017, brought to you by Hewlett Packard Enterprise. Welcome back, everyone. We are here live in Las Vegas for SiliconANGLE's CUBE exclusive coverage of three days of wall-to-wall interviews here at HPE Discover 2017. I'm John Furrier, your host with Dave Vellante, co-host of our next two gasses, Rod Bag, VP of Analytics, customer support, data center infrastructure at HPE, formerly Nimble, now HPE, and Paul Sabin, senior network and infrastructure manager, Baker, Botts, LLP. Guys, thanks for joining on theCUBE. Appreciate it. Thank you for having us. Thank you for having us. So we talked before we came on camera about all the great stories. Nimble, obviously now part of the fold here at HPE Enterprise, new customer stories. Let's get right into it. Tell your story about how Nimble puts you out of a job. That's my favorite one, go. Okay, so when I started, or when we bought Nimble Storage, I was the singer storage engineer. So we purchased it, we brought it in-house. It was up within an hour. I was already starting to carve out lones. At that point, I'm using the RESTful APIs to carve out the rest of the 200 lones that we needed presenting it to the hosts. And by the end of it, it ran itself between InfoSide and the fact that the product just is so easily automated. I kid you not, true story, at the end of the year when we're doing our self-evaluations, my evaluation said, and congratulations, you don't need me anymore. My position is obsolete. And the management came back and said, Paul, you're absolutely right. We agree that we don't need this position anymore, so we're going to promote you to the senior network and infrastructure team. So I manage that now. So you got promoted, but this is a trend in automation. This is the DevOps. This is the programmable infrastructure world we're moving into with hybrid. Ron, this is big deal. Yeah, yeah, exactly. And InfoSide, as we see it, plays a big role in that. The product is simple and being able to automate that, but InfoSide giving our customers sort of visibility at a very deep level into how the systems are performing and what we do on the backend to drive availability really takes a lot of pain off of our customers. And I'm not sure that we put everybody out of work, but we certainly make life easier so that they can focus on the business aspect. Yeah, I mean, you automate those tasks the way that really should be automated. And that's a cool thing. Take a minute. I'd like you to take a minute and tell me what the product is and what you guys are doing just so we can get that out there as context and then gentlemen, this is more stories. Yeah, so from an InfoSide perspective? Yeah, so InfoSide is our predictive cloud analytics platform that uses machine learning to predict and prevent problems from occurring to our customers. So we're not disrupting their business. And so we collect somewhere in the order of, but maybe 25 million pieces of information from every array and the virtual environment every day from every single array. All of that gets into a big analytics database where we have a team of data scientists working with our support engineers and our product engineers to build wellness rules. We have about 800 health checks that are really looking out at every part of the infrastructure for our customers and really avoiding issues for them. So you take the data across your entire install base. Yep. Yeah, I'm sure you take care of the data so it's not all. Oh yeah, it's all secure and not always. And then use that as predictive to prescribe or both or how are you? Yeah, both. So our real goal there is that if we know of an issue that's either we found in our labs or maybe one customer has experienced it, really we're doing everything we possibly can to analyze that issue across the entire install base. So we're learning from peers and applying those learnings across the install base and preventing other customers from hitting that. And the system is auto-didactic in this sense. It learns and then applies, is that right? Yeah, so we do machine learning semi-supervised in a lot of cases. So where we've seen an issue and we can train the models and then it will look out for those sort of issues across the entire install base. I like the notion of wellness. Yeah. It brings some of the people in a relationship. We also heard terms like self-driving storage. Yeah. Playoff testing. Yeah. But this is again the trend that really is needed. Share other stories that you have because this is really where IT is going as it moves to a different kind of application and consumption model for you guys. Right. So, well, kind of touching about what he was talking about. When you're as a storage guy, what's the number one thing that us storage guys have to do is we have to prove that it's not the storage that's the problem. So usually what happened was in the old world, I would produce some statistics of, okay, and here's the IOPS that we're producing. Here's the latency during this time. So based on this, it wasn't me. I don't know who it was. I'm just going to tell you it's not me. In the new world. That was the finger pointing world. Yes, it was. But with InfoSight, it's like, hey, I can tell you, but you're also welcome to go here as well. But let me show you VM Insight where it's going to show you not only what was happening at the storage, but let me take you all the way down to the host and then the VM and we're going to find this problem. And yeah, it turns out sometimes it's going to be the VM that's all of a sudden taking whatever reason, adding a huge amount of latency. And that is something that, there's no more finger pointing at it anymore. So all of a sudden we're all on the same team. It's like this kumbaya thing. That's awesome. It's good for the cohesiveness of the team, but also it's time savers too. When you reduce the steps to do things, you get your weekends back, as you guys say, and before we came on camera. Tell the story about how you had to do all this work on the provisioning or on the replication side. Sure. So when we deployed the arrays, we decided it was a business decision to go ahead and put the production arrays into our production data center and then we would do the DR at a later time. So I've got all of my data live on production and they say, okay, now we're adding our nibble storage at our DR site. Paul, how much replication bandwidth do we need? And so same story in the old world. You go and you pull your statistics from your replication technology, you put it into Excel spreadsheet, you figure out, okay, here's my peaks and if I want to just say if we fall behind just a little bit, this is what we can do. And so usually what happens is I say, guys, in my best guess, based on what I can see from my limited scope, because my eyes are bleeding at this point. From the spreadsheet, you're in a spreadsheet right now. Yes, exactly. You're in spreadsheet hell. I'm in spreadsheet hell. And so what I do is after about a weekend's worth of work, I put in this recommendation and I usually fluff it because I could be wrong in my statistics and so this is what I end up creating. Yeah, you don't want to be under, you want to be over. Exactly, I'm always trying to do that. So the firm, hopefully nobody's watching at the office but sometimes they may be overpaying for something because I just don't want to make that chance in the new world. This is actually the coolest thing ever. So I'm on InfoSight and I go to this little drop down and it's like the tool planner, okay, what's that? And it's like, we're actually going to tell you what you need for bandwidth based on your actual real data. So then I'm pulling like, okay, based on this time, what is the replication if I want to do it every hour? And what if I want to do it every two hours? So then I just take that and I turned it into this report that I got to present to the executive team and they're like, oh my goodness, you have certainly stepped up. How many weekends did you use on this one? And you know, I'm not going to tell them, it took me five minutes in InfoSight to be able to create this report. Now they know. Yeah, but now they know. Well, you already got promoted, so that's good. That's true. So, Rod, can you talk about the decision to acquire Nimble? I mean, what was the genesis? Obviously there's a portfolio component. Yeah. Tuck-ins, fill in some gas but there's this other sort of IP piece. Maybe take us back. Yeah, so certainly there was the portfolio fit with the storage platform. So that was obviously a big part of it. I think the other obviously big part was InfoSight. So the idea that what we're doing there with our customers and improving the availability of the systems and the operational performance of the system and keeping it close eye on that to make sure it's optimized. So all of that value prop around InfoSight was a big part of the decision, I think. We are working on extending InfoSight into the HPE product line, starting with 3-part. So we are working already with that engineering team to be able to bring some of these features out as quickly as we can into the 3-part world as well. So what is that, I mean, from an engineering standpoint, is that sort of the requirement there is to point InfoSight at the 3-part data? Yeah, exactly. So 3-part does collect a lot of data already. Yeah, it sure does. So really we're just pulling that data into our pipelines and so on with an InfoSight and taking advantage of some of the machine learning and algorithms and so on that we already do, things like DMVision would be possible and so on in that environment as well. It's interesting. 3-part customer. Back maybe 10 years ago, 3-part was sort of the gold standard of what we used to call the hero report. That's right, yep, yeah. And people love that. Thin provisioning, what impact it was, how much you saved, et cetera. And then that predated the whole big data analytics. Yeah, exactly. So when Nimble started, they could have started with that premise right around that time. I mean, when I first saw it, I was like, wow, this is magic. Yeah, exactly. I mean, that was the premise was to really apply data science to all of that data that was coming in and really transform the support experience for Nimble and I think that's the other big element for HP as well. I mean, there's lots that we do in our support organization that to be honest, is quite enviable by a lot of storage and high-tech vendors. Well, I mean, you guys took a different approach, I think what's really notable for me, which I'm impressed with is everyone talks about this, but very few put it into action is making the user experience center of the value. I mean, all the things you're talking about, the benefits, is really centered around your experience, right? Saving you time, making your life easier, shifting the automation that could be automated with the right things and moving you to a higher value thing. So Paul, what's your thoughts on this as it goes forward, this world's evolving? We're hearing the messages here, simplifying, hybrid IT, you get cloud right in the door, step multiple clouds are going to be the end game on all this, so all said and done and whole new infrastructure is going to be out there. What's your view of how that user experience for the practitioners will evolve? What's your vision? How do you even see it playing out? Putting out a job again. Ha ha ha, that's a true story. So the firm decided that they were going to bring in some people to help us look into what cloud we should or how we should utilize the cloud because even from us, we're trying to keep ourselves agile as a law firm because if we can provide our services in a better, more meaningful and faster way, that gives us a competitive edge. So we brought in this team and they went over all of our IOPS and at the time it was under the different storage system so it took at least 20, 30 hours of my time to get all of these numbers that they wanted. And then they created this report for us which I thought was really meaningful and valuable. The last line was you should do cloud where cloud makes sense. So that was it, solid advice, money well spent. And that's what Meg's basically saying in the keynote, the right mix of cloud versus on-prem. Certainly law firms have proprietary information they want to secure. I guess my question really fundamentally is a provocative one. I'd love to give you thoughts on serious questions. You can laugh at it a little bit but with AI bots coming, you can almost see these kinds of legal tasks being automated away. So you might be next promotions and take it over the firm. That's where big data can come in. So how are you guys looking at that as a firm because I'm sure the lawyers are saying, hey, you know what, I can shift my value to higher yield activities where that makes sense. You guys talk about that at all? We do and I actually use the example of NASA. I really love NASA, I'm a huge fan. And NASA decided they declared we're going to go to Mars. We're going to do this, how are we going to do this? We have to let go of our operational stuff. I mean, we can launch the shuttle all day long. We're comfortable with that. We can go into the space station. We're comfortable with that. But now we've got to go new. And the way we have to do that is we have to drop this stuff. Let's let other people do this. Let's let the InfoSight team start handling a lot of that work for me. And now I'm asking my team, guys, I want you to start dreaming. Get out of the operational work, start dreaming out loud. Let's figure out ways we can deliver value to our attorneys to free them. And let's let them just, again, take that same freedom. And with the business intelligence and the machine learning, you're right, that their document management, which is their bread and butter, is their document production. Even that's getting scrutinized or transformed through this machine learning. And so you could take this as a way of saying, no, there goes my job. Or you can say, no, now I've got the opportunity to do something even better and cooler and really bring the value. And stretching, that's that whole stretch goal, having that moonshot, in this case, Mars example. It's the stretch and leverage, right? That's the concept. How do you apply that storage? Because now HP's got the composability, they got synergy, they have all kinds of now glue layers kind of developing. We heard Antonio Neary in the press and analyst Q&A with Meg Whitman talk about, most of their acquisitions have been in software, except for like maybe one or two over the past couple of years, have been software. So hardware, software kind of blending? Yeah, I think so from the storage perspective, certainly. I think that's happening. I think from the InfoSight perspective, where we see that going is again, today when we put a lot of effort into our recommendation models, and that's an area that's very much in the deep data sciences realm. So when we come up with those recommendations, we do things where we can prevent people from hitting issues, and that just sort of happens automatically. But some of these things are, something needs changing in their environment. So maybe there's a QoS policy that should be applied on the array to optimize performance because of some peak workload during Christmas, something of that nature. So that's still a last mile problem for us because you've got a human at the other end that's got to go in there and fix it and hopefully do it right and not ignore it and everything else. I can see the headline now, storage wellness comes into HP, but this is really interesting. So you have the concept of self healing, right? So that's where we want to go with that. I mean, that really is the thing we're working towards in the vision is how do you go and do that, change those QoS policies for the customer where we could inject, let's say a change control within their change management system. They can go hit a button, which we orchestrate that change for them. It's all documented and well controlled. It's not just storing the data, it's being data driven for the data being stored in the self-driving order. Rod, Paul, thanks so much for sharing the stories and congratulations on the promotion. Thank you. And congratulations on, and you guys got great. By nimble get promoted. Come in the queue, get promoted. Birds of a feather. Appreciate it, thanks for having us. More live coverage here from theCUBE here at HP Discover 2017. After the short break, I'm John Furrier with Dave Vellante. We'll be right back.