Cloud computing has grown from emerging disrupter to the very foundation of today's enterprise IT, and yet the pace of change in the cloud sector shows no signs of lagging.

Hybrid cloud has given way to multicloud -- or is that just hype? The concept of "cloud native" is now au courant, offering its own myriad challenges. Emerging technologies from microservices to kubernetes to edge computing are prompting big shifts.

These many and constant new developments beg the question: what do I need to know to truly be current with cloud in 2020?

To provide insight, I spoke with a leading cloud expert, Bernard Golden. Golden had held number top tech positions; most recently he was Vice President, Cloud Strategy, Capitol One. Wired magazine dubbed him "one of the ten most influential people in Cloud Computing." He's the author of Amazon Web Services for Dummies, a bestselling cloud computing book.

Golden: It's the sort of thing that happens when a new way of doing things comes available and there are incumbents who are experts at the old way of doing things. And then the benefits of the new approach become so manifest, so evident, that to remain competitive from a business perspective you have to shift to that.

"I use auto manufacturing, auto industry, as an example. Basically, before Henry Ford mass assembly, cars were built the way you and I might do it. Everybody did everything. You didn't have to build a very elaborate factory. It was operating expense heavy capital, expense light. Henry Ford turned that on its head. He came up with a way of manufacturing cars that was far more productive, far less expensive.

"So the operating expense dropped enormously, but you had to have much more capital to be a competitive automaker. The net result was cars got a lot cheaper. So people started buying them, so it drove the size of the market. But to be a car manufacturer also that was a completely different game. And to respond and be competitive in that marketplace, you had to be able to have tens of millions or hundreds of millions of dollars of capital.

"And so, something like 200 car companies went out of business between 1913, 1925 because that changed. And cloud native, I would assert is an analogous kind of technology shift. And when [cloud native] companies like MTailor, all of a sudden grow to hundreds of millions of dollars a year of revenues, you have to respond. Because then people go, "Why can't I have something that's perfect for me?"

Golden: The question for most enterprises is, How much do I have to transform what already exists, and how do I free up enough investment to do this transformation? That is a huge challenge.

"I talked about the car manufacturing. Most car companies couldn't make that transition. They just couldn't find their way to funding that high capital investment, low operational, low OPEX model. They just... They surrendered. They said, "We are gonna go out of business." So, that's the whole thing about application inertia and the budget crunch. Where will enterprises find enough money to fund these changes? How can they find the talent to do it, which is typically expensive talent? And what do they do about their existing environment that so much of budget is tied up in?"

Golden: [The four companies are the three cloud providers, AWS, Microsoft Azure and Google Cloud, with the fourth being VMware].

"Five years ago, VMware presented a kind of an ultimatum to their customers: 'Use us or use those cloud providers.' And I didn't think that was a very smart strategy, because it sort of cut them off from all of this innovation and wasn't great customer relations.

"They've really changed that around to now saying, 'Yeah, you've got a lot of our stuff.' And they're basically the de facto infrastructure environment for most of the Fortune 2000. And all those companies are like, 'How do I deal with application inertia?' which we just talked about. How do I get to digital transformation? And VMware is now saying, 'We have a way, a pathway for you.' And part of it is they've said, 'We will run our environments in these cloud infrastructures.' So that they enable their customers to get out of owning their own data centers and tying up so much capital and investment there. And part of it is, they're enabling application monetization, they just brought Pivotal in."

Golden: "I think that most enterprises will be multicloud. They will use multiple clouds. There's no doubt about that. But the vision that there's some way to magically make applications somehow, 'Oh, I used to love this provider but now, you know, the sales rep really ticked me off. I'm gonna move all my applications' I think that's just too simplistic a view.

"A lot of companies say, Well, if you use containers or Kubernetes that will happen.' But I estimate that's maybe 25% of all the capability to actually migrate an application. You gotta look at how do you handle identity management, how do you handle security, how do you handle all the billing that's associated with a provider. Because most enterprises aren't gonna be just buying the list price off the website, they're gonna have a customized contract that they've negotiated."

Golden: "Just what proportion of the overall use cases will those [edge computing] use cases represent? I'll give you an example. I just had a new thermostat installed and it has some smart stuff, like it tracks whether people are in the house. And if nobody's in the house for a certain period of time, it'll turn things down, so forth and so on.

"Well, so it's got some sense capability. It sends a message back, probably once a second, to the cloud-based application by the thermostat provider, but that's probably a few hundred kilobytes. So does it need a local kind of compute capability and so forth? No. At the same time, there's other kinds of applications that, for a variety of reasons, latency, the difficulty of communication, whatever it might be - teah, they do require much more locality of the...[processing]

"But to think that everyone's gonna be running out and having a little sort of mini data center in their office to run whatever their edge thing is, it probably isn't... I think it's important to have a realistic perspective on that."

Originally posted here:

Cloud Computing 2020: Five Key Trends - Datamation