 from Gillette Stadium in Foxboro, Massachusetts. It's theCUBE, covering VTUG's New England Winter Warmer 2017. Now your host, Stu Miniman. And we're back. This is SiliconANGLE Media's presentation of theCUBE. We are the worldwide leader in live enterprise tech coverage. Here at the VTUG talk about a broad spectrum of technologies, talking about virtualization, cloud computing, developers, and one of the keynotes this morning was my guest that's joining me, Dewey Sasser, who is the Cloud Solutions Architect at Align Software. Talked about AWS and what it's like to be as a user of public cloud and your consultant that works on the AWS offerings. So thanks so much for joining us. Thank you. All right, so Dewey, I mean, I noticed a lot of cats in there. Does that mean that, you know, public cloud I'm going to get some claws and some people will be allergic to it and it'll be challenging or is it just because, you know, cats and cats are great? Well, I think it's actually a feature of working with developers. You're going to get the claws. Yeah, absolutely. It was funny. I told you the Docker keynote, which also was talking about developers, oh wait, there's the demo was a catgift that comes up and you refresh it every time and it's a different catgift. So that I could go in and hack at that. But at this conference, it's been interesting to watch the maturation of the last few years, you know, people that, you know, understand virtualization, things like VMware and, you know, getting more and more comfortable with public cloud. Talk to us a little bit about, you know, how long have you been doing public cloud? What kind of things are you helping companies with? So my history is I was actually a software developer and then I moved into private cloud very much what this conference has historically done. And the last few years I've segwayed that into public cloud. Basically I was here today to talk about the social aspects of public cloud, how to adapt your company to it and what are the particular challenges you have in a cloud environment. You started counting some of them in a private cloud environment and with general virtualization, but a lot more when you move into public cloud and everything becomes easily accessible to developers. Yeah, I mean, I'm an infrastructure guy by background and, you know, for at least the last decade we've been talking about, you know, we need to get out of our silos, we need to re-chain, you know, rethink how the organization is made up, especially from the infrastructure level but from an organizational, you know, area. I need to think about services more. I need to think about, you know, architecture more. I mean, I think the big challenge of our industry is really about building distributed systems and the impact that that has across the board. And what I saw in your presentation was a lot of that is right. Who owns what, you know, what are common services, what are done by, you know, various parts of the organization and how that plays out. Maybe you can, you know, give us a thumbnail of how some of that looks. So the key from my perspective is you have to make sure that the responsibility and authority are always aligned and you have to put the actual power to do something in the hands of the people capable of doing something. So there's no point in waking an ops guy up in the middle of the night for a code problem. There's no point making a, working a code guy up in the middle of the night for an ops problem. So the key here, I think as we move into public cloud, the basic momentum of dev ops and agile processes and public cloud are really coming together to meet in one spot. And in order to deal with this, we need to change it so that people can be responsible for what's going on in a minute to minute basis. Whereas we used to do it in a week to week basis. Yeah, so, you know, Dewey, I completely agree that we need to get kind of, you know, the responsibility and, you know, what you're doing all aligned, you know, how do you handle, I mean, those communication gaps or everybody's like, no, that's not me. That's, you know, that group and that falls. You know, when I go agile, you know, that goes away with the waterfall world or, you know, it's not a silver bullet, right? It's not a silver bullet. Fundamentally, they're all people problems, communication problems and waterfall, agile, all of these are just different ways of solving the problems. All right, great. So, you know, you're, you know, I know you can't talk about the customer itself, but you know, mobile is involved heavily, you know, streaming type of things are involved. Some of the things, I know when I went to AWS re-invent the first time, gosh, you know, four or five years ago, it's, you know, lots of those, you know, it's new applications, you know, gaming, you know, is heavily involved out there. From an application standpoint, is it, does it help to be kind of that new type of application that's starting from the ground up? Or, you know, what about, you know, companies that have legacy environments? How do you see public cloud bridging some of that gap? So, there are different techniques you use when you're going in with a new environment, going with all the buzzwords and restful services, Phoenix server patterns, which is basically a pattern where you burn down the server. Instead of debugging it, you just torch it in place and put up a new one versus the old style of doing things where you really cared about each individual server. You would go and debug the problems, you would fix it up. Public cloud can be used for both of them. It really comes into its own for the Phoenix server, for the agile, for the elastic stuff. There's so much you can do there that you really can't do in your data center without way more investment than it's worth it for almost anyone to do. But for legacy products, yes, the cloud also works. I've also done a number of legacy product migrations into the public cloud. And it's just different tools and techniques, though you can take advantage of some of the elasticity and some of the dynamic provisioning you get from the public cloud and that too. Yeah, so when you engage with a customer, do you have to kind of understand their maturity of their organization? I know you've worked with Colos in the past. You know, certain people have certain skill sets and understand certain things. Moving to the cloud, is there a certain amount of education training? Do you gauge if they're ready for a certain type of environment or do you just kind of help them say, here's the new way and you're going to catch up? You absolutely have to pay attention to the individual customer. Amazon actually has, I think, a five-step model of cloud adoption that I found pretty accurate. People start with a trial, you know, let me get my credit card and try a machine there, smooths and dev workloads, smooths and non-critical production workloads. It's a pretty good model. It varies, most of the customers I talk to, you don't have to explain public cloud anymore. They've heard about it. They know it's a thing. You need to explain to them what it is and where it's good for their problems and where it's not good for their problems. All right, maybe what are some of the pitfalls there? What kind of things should people avoid or make sure that they don't try to, you know, shoehorn into the public cloud or things that they need to prepare themselves for? That's really interesting. I would say that at this point, the public cloud is pretty much mature enough to handle any workload. The question is, is it cost effective for you? There are a few workloads, not terribly many, that if you already have an investment in a COLO facility, might not be worth moving into the public cloud. On the other hand, if you're willing to do a little bit of work, you may be able to actually achieve some cost savings there. Build test is a great example, which traditionally has been on a dedicated rack, but if you actually look at the time spent using an individual machine, taking advantage of public cloud elasticity even for the more traditional test environment can really get you a good cost savings. Yeah, the term that I know AWS has used it, but we've been using it for a long time is companies need to get rid of what we call undifferentiated heavy lifting. I try to poke a little bit more and say for the enterprise, there's lots of stuff that you suck at that you really shouldn't do. There are very few companies out here that are really good at building data centers. And if you're an enterprise out there, you're not one of them, because the ones that are do that, they're real estate investment people. Like companies like Equinix know how to build data centers. There are a few companies from the physical layer, the slab, the power and cooling. You mentioned don't build another data center because any company that I've built with that's on the enterprise side that built a data center said, I hope that's the last time I ever have to do it because boy, I learned way too much and I know I didn't do good. The PUE is going to be much better going to either a Colo or a cloud provider. And as you work up the stack, there's things that I don't need to be the expert. As God, we all spent, how many decades do we spend putting out our email server? It's like, well, there's services, Microsoft's pushing everybody to just say, buy it as a SaaS model. So I guess the question I have for you right is, what does in-house IT need to have as their core skill set? And what stuff should they just really be able to shift over and put that onto platform, put that onto vendors, put that onto consultants? So a lot of the fundamentals remain the same. Networking is networking, whether it's virtual or physical. Security is also security. There are some different things that you care about in the cloud, but the mindset is still the same. Those still approaches still work well. Understanding of your application, what's important to your application, what's important about the security of your application, what your customers needs, those are all important things that you don't want to outsource. How to configure a particular network in a particular environment to achieve your goals, those are something you can work with a vendor to do for you. Yeah, as a matter of fact, you use the term outsourcing and I'm saying the old outsourcing model is mostly going away. I don't just want to take my mess and think that somebody else can do it for less, right? That didn't work, it failed most of the time. Companies need to understand what's strategic and skill sets that I need to have and I need to have a good understanding of platforms and vendors and solutions that offer consultants and systems integrators can help along that line, but I can't be blinded to say, okay, you go do this and I'm not going to think about it. There are services that I will consume, things like SaaS or services on public cloud. Amazon's been doing a lot to move up the stack, everything redshift and many other services up there where I'm not outsourcing it, I'm leveraging standard offerings that are out there in many ways it replaces the old shrink wrap software that's out there. The best way I've heard it put is don't outsource your core competency. So a company needs to really understand what their core competency is and I think what Amazon is doing, going up the stack with redshift with Elastic Container Service with Lambda is you can really focus in and tighten your efforts to just your core competency to what really delivers value to your customer. Let Amazon worry about the rest. Yeah, so I'm curious, I think back a couple of years ago there were certain people in IT that would look at something like Amazon and say that's a threat to my job. Oh my gosh, what do I do this public cloud? I don't understand it, they're trying to take my job away everything will just be automated. How do you look at that these days and the relationship between IT and public cloud? So I've never found a successful IT person who doesn't want to learn new technologies. It just doesn't happen even if you're trying to stay in a co-load facility you need to follow progress and public cloud is just progress. It's a larger step than we've experienced before but I also don't really know any out-of-work IT people who have been unable to make that transition. I can see how they would look as a threat to their job. It is a threat to the old way of doing things but the PC was a threat to the mainframe, the web threat to the PC. Things change, we're still smart people solving problems and we're still going to need smart people to solve problems. Yeah, I love that, right? We need to always keep learning, stay curious, keep working on it. This last question I have for you is, does it all go public cloud? I know you're talking about AWS and the like. There's a lot of different solutions out here. Many people building on-premises as that spectrum of all-cloud, most public, most staying private, where do you see things today and playing out over the next few years? I think very much what you said with you don't build your own data center in the last 10 years people don't build their own data centers, they'll co-locate. They'll rent space from someone who has built their own data center. I think over the next 10 years we're going to see more of that with public cloud. People are not going to even co-locate, they're just going to go and buy it from a public cloud. The pricing will change, what public cloud's offer will change but I think that's the way we're moving is over time we'll see fewer and fewer people managing their own physical hardware in an environment that is dedicated to them. Dewey Sasser, really appreciate you coming here, sharing with our audience everything that's going on. Public cloud, huge trends, so many shows we'll be covering this year. AWS re-invent every year. I'll be at the Google cloud event at the beginning of March and lots more on siliconangle.tv and siliconangle.com. We'll be back with lots more coverage here from the VTUG, you're watching theCUBE.