Deloitte Consulting Chief Cloud Strategy Officer David Linthicum deciphers the evolution of cloud computing; how it has changed the way enterprises deploy IT.
Join the DZone community and get the full member experience. Cloud computing has come a long, long way from the early days of mainframe timesharing technology and virtualization. Over the past decade alone, we’ve seen cloud computing experience tremendous growth, providing ubiquitous solutions for both consumers and enterprises. This trend is expected to continue with the rise of cloud-based services for machine learning, artificial intelligence, and edge computing. However, the transformation to cloud computing remains a very challenging prospect for many enterprises. Enterprises need to consider factors such as cost, security, legacy and on-premise technologies, and the value of using cloud-native vs cloud-agnostic services. In this round of Coding Over Cocktails, our David Linthicum sheds some light on the evolution of cloud computing; how it has changed the way enterprises deploy IT; how enterprises balance the demands of the cloud and legacy systems; analyzing your company’s capabilities of going cloud-native; and the future of the cloud. Kevin Montalbo: Joining us today from Australia is TORO Cloud’s CEO and Founder, and Cocktails’ co-host, David Brown. Hi, David. David Brown: Good day, Kevin. KM: And today’s guest is an internationally recognized industry expert and thought leader, who is currently the Chief Cloud Strategy Officer for Deloitte Consulting. He has written more than 13 books on computing, published more than 7,000 tech articles, and spoken at more than 700 conferences. He is also a Gigaom research analyst, a cloud computing blogger for InfoWorld, and a frequent contributor to ‚IEEE Cloud Computing,‘ Tech Target’s SearchCloud and SearchAWS. Joining us to share his knowledge and expertise is, well, we didn’t plan this, but another David! David Linthicum all the way from Virginia. Welcome to our podcast! David Linthicum: It’s great to be here, lots of Davids in the world I found out. DB: A fair share of David Browns as well. KM: 13 books and 5,000 tech articles. That’s a lot of content. For someone who works on the content side of things, I’m just curious, how did you get into content publishing? DL: Boy, I started way back when I was in my 20s. I used to publish articles for the local PC Sig, remember those days, special interest groups, days when the internet wasn’t around. We used to gather together once a month and exchange ideas pn how to build and deploy PCs. I used to publish wire diagrams and lots of technical articles then. A little challenging because I’m severely dyslexic, never spelled well, and barely made my way through college and high school. I started to train my brain in order to produce additional content. Actually the use of a computer and word processor kind of changed the game for me because it could in essence go behind me and check my spelling and my grammar and things like that in the early days of stuff. I found out that it was a career enhancer. The more I published, the more people recognized my name, and more jobs I got, and the more money I made. When you’re in your 20s, you want to go out and have a good time, buy a car, things like that. That kind of meant a lot. It just progressed my career. In other words, I started writing books, the articles started to go to PC magazine. I was one of the editors there, it was a great amount of fun because it was testing and deployment. In PC magazine’s editor’s choice, I was involved in that than any number of enterprise software books, database programming design, case trends which was computer software engineering trends. Just basically anything that came along that I thought was interesting as a topic and subject area. That’s kind of the way I learned how the subject area was. In other words, you can’t do everything. As a human being, I can’t be an expert in every programming language, in every architecture, in every database, but I can learn a lot about a lot of things just kind of basing the fact that I’m writing about them. I know a lot about a few things, I know a little about a lot of things and I do that through research and writing. It allows me to an essence, kind of figure out how things are shaping and working together. It allows me to identify emerging patterns and kind of what I do for a living now. It’s figuring out where the balls are gonna be kicked and figuring out what the next generation technologies are gonna be. In my publishing history, last I published was in 2009, I’ve kind of fallen out of it right now, people are actually trying to get information was ultimately just revealing new concepts that didn’t exist before. EAI, Enterprise Application Integration, that was my term that was basically represented by the same book Cloud and Service Oriented Architecture, just basically building on those. Then I got writing into columns, I’ve been writing the cloud computing column for InfoWorld for the last 12 years and any number of other publications as well. I do a lot of radio and TV and podcasts like this. I just really enjoy it. KM: You have witnessed the evolution of cloud computing over the last 20 years. How has the cloud changed the way enterprises provision and deploy IT? DL: Well, ultimately it provides a different consumption model at the end of the day. People always talk about the revolutionary nature of cloud. When you think about it, we’re still dealing with databases, storage, compute cycles, and many of the same platforms we leverage on-premise. If you look at what cloud computing is in the way it evolved to be, it’s an essence in the ability to configure a data center virtually. Certainly, the infrastructure is the service provider. I can put it in a storage, computer, database, or I can put it in platforms, I can put it in Linux, Windows, and C++, and anything I need to configure what infrastructure I need to support my application. The great thing about a cloud is you can do it in about an hour versus if you’re a traditional enterprise, you have to go cycles. It’s gonna take six months to a year for the standard gold 2000 company. Revolution is really around the consumption model and how it’s gonna free us up from actually being innovative and creative and not necessarily putting IT as a limitation on the business. You talk to the CEOs out there, they’re gonna tell you that “it takes me two years to integrate a company I buy because IT is going to basically spin around on particular integration technologies that they need to make these things run”. Cloud computing is a matter of spending up what you need. It’s a matter of core integration technologies. It’s a matter of layering in different security platforms and doing so you really kind of at the speed of need. So, that’s the revolution of it. And the second generation of it that we’re seeing now is the ability to leverage advanced technology that doesn’t exist on-premise. I mean, some of the AI stuff you can’t find on-premises, you know, some of the machine learning-based systems can’t find on-premises, certainly the serverless computing stuff you can’t find on-premises because that would be weird. Ultimately we’re able to weaponize technology that prior to this was unaffordable and obtainable. And right now, probably five years ago, the cloud kind of crossed the chasm where they have better technology than we do on-premise, whether it’s security, governance, management, monitoring. And that’s because if you look at the way the R&D dollars are being spent within the technology companies, they’re investing in building cloud-based products, not necessarily on premise-based products. So, those two kinds of waves occurred. And I think the next wave will be the sync and migration of legacy systems moving forward and kind of dealing with that whole thing. But, the revolution I saw, certainly back in ‘99 when I got into the cloud game, was the fact that this consumption model is going to be more aligned with what business needs and business even understands what they need and that turned out to be the right thing. But of course, it took, you know,11 years to do that. DB: Many companies still have significant investments in IT infrastructure which is on-premise. How do companies balance the demands of cloud with legacy on-premise systems? DL: I’ve been in cloud for a long time, and I live in Northern Virginia, and this is the data center capital that seems like, of the world. And I’m watching data centers go off, probably, there’s probably 10 projects right now within five miles from me. And so what happened, suddenly the ability to leverage cloud and share resources would get us away from traditional data centers. Well, the reality is much different. The reality is that there’s a certain amount of infrastructure that we’re going to have to own. Either that’s going to be in managed service providers, colo providers, around private data centers. So we’re going to hit a saturation point in cloud migration. Cloud guys don’t like me to say this, typically between 70 and 75%, where the remaining 30% can’t migrate anywhere. They’re more valuable to the business. A lot of things we have in the cloud and therefore we have to make these things, cohabitate what I say, work, and play well together. And right now we don’t have a lot of very good middleware layers that are able to span on-premises in the cloud. We’re getting to security, we’re getting to governance, we’re getting management monitoring, which have this duality of roles where they’re able to in essence monitor systems that we own, monitor systems and managed service providers and CoLOS as well as in the public cloud. That’s new to everybody. Everybody has a tendency to kind of want to leverage whatever native tools and technologies are there moving forward. I’ve been spouting for a long period of time and it’s actually got me more ire than anything else is the fact that I don’t think these legacy systems are going to completely go away more so than you think, they’re going to be a big part of our infrastructure moving forward. I think we’re going to be able to find analogs, other ways to run them. CoLOS and managed service providers are good options for that right now. Everything’s not going to be able to live in the cloud. Even if we have platform analogs that exist in the cloud, that doesn’t necessarily mean that a data set or an application workload should exist in the cloud. We’re finding that the economies are really kind of driving a lot of things back to the data centers. Some of the Silicon Valley-based companies where they’re putting out a social media platform or they’re putting out a gaming platform and they did so in the Amazon cloud or whatever public clouds that are out there, found that it’s more cost-effective to basically move it back to their existing data center that they own. Because you think about it, it’s really kind of a single pattern of workload. In other words, they’re running a game, they’re doing one thing. They’re not running 10,000 applications like the typical enterprises. And if they’re able to own the hardware and software and able to optimize that and optimize the network bandwidth as it goes between them and the user is able to do so at a decreased cost, that’s where you’re going to be. And I think we’re going to see a lot of that. We’re going to see normalization in the market where people moved out to cloud probably ill-advised and they’re going to normalize back to the data centers, their private data centers, or CoLOS, things like that. And then we’re going to see certain platforms move out the cloud. You know, one of the things that I tell my clients when I meet with them, it says, you know, I’m not a cloud bigot. I’m here to basically mediate the best solution to the problem and the best architecture, which is going to be optimized in a certain way. And sometimes that’s on-premise. Sometimes that’s other platforms, sometimes it’s cloud, private cloud, edge computing, IoT based computing, all of the above. It’s really becoming that way and figuring out how we’re going to look at this from an unpartisan way since we have an election coming up, where you’re actually going to pick out the right solutions for the problems that we’re looking to solve and not necessarily move in a certain way, what I call the managed by magazine crowd, just because everybody else thinks it’s trending to move out in this direction. DB: It has a lot of pain points associated with it. If you tried to make that move in and you’re not ready for it, or it really wasn’t designed to be made that move. Are there any sort of use cases that come to mind where people shouldn’t try to move on-premise to the cloud? DL: Obviously, if you have a significant amount of security risks, that’s there and you’re worried about, for example, data sovereignty issues, in some of the countries out there have very strict data retention rates, auditing rates, things like that, where it’s almost impossible to move it out in the cloud. If you have to audit these systems in a certain way, in many instances, the cloud providers just don’t provide a way for you to do that. You can’t walk through their data centers, for example, and take the serial numbers off of their servers. You don’t know where their data centers are. And that’s a good thing. You don’t know what servers are running on. That’s a good thing. You don’t know who’s maintaining the servers. So it’s those kinds of restrictions are there, that immediately eliminates cloud. The other thing I say, which is a little different and different than most people have an opinion if you’re overly concerned about the control aspect of it.
