In this episode of Sovereign DBaaS Decoded, Vinay Joosery sits down with David Linthicum, acclaimed cloud strategist, author, and industry thought leader, to dissect the real story behind the “cloud repatriation” movement, why cost, compliance, and control are pushing enterprises to rethink their cloud commitments, and what the future of hybrid and multi-cloud looks like in a world grappling with both hype and hard reality.

Key insights

⚡ Cloud repatriation is a strategic reset—not a retreat

David emphasizes that today’s moves away from public cloud are less about abandoning innovation and more about correcting course. Many enterprises, lured by the “herd mentality,” rushed into cloud without proper analysis—and are now seeking the right balance between cloud, on-prem, and everything in between.

⚡ The true cost of cloud: much higher than expected

According to David, companies are now discovering they’re spending 2.5 to 3 times more on cloud infrastructure than initially estimated—a wake-up call that’s fueling the repatriation trend and driving organizations to reassess what workloads truly belong in the cloud.

⚡ Europe leads the sovereignty push

Heightened concerns around data sovereignty, compliance, and geopolitics are causing European enterprises to move away from US hyperscalers at twice or even three times the rate of their American counterparts. Trust, regulation, and risk management are in sharper focus than ever.

⚡ It’s all about “right-sizing”—not abandoning the cloud

David points out that the most successful organizations aren’t ditching cloud entirely; instead, they’re embracing hybrid and multi-cloud strategies, placing workloads where they make the most business sense, and leveraging the strengths of specialized and sovereign cloud providers.

⚡ Beware of hype cycles (again!)

The mistakes of the cloud rush are already re-emerging with generative AI. David warns leaders to avoid simply following trends and instead focus on genuine business value, sound cost analysis, and bespoke strategy.

Episode highlights

💡 How did we get here? The cloud herd mentality [03:00 — 09:00]

David and Vinay break down how industry hype, analyst pressure, and consulting incentives drove mass migration to cloud — with little regard for actual business needs. David recalls, “It was a classic example of herd mentality. Everybody is doing it. So should we…”

💡 Sticker shock: Cloud’s hidden costs [09:00 — 15:00]

David shares eye-opening insights on cost overruns: many companies now realize their cloud bills are 2.5 – 3x higher than planned, often cannibalizing innovation budgets and leaving less room for R&D.

💡 Cloud repatriation: Quiet trend, big impact [15:00 — 20:00]

Despite a lack of public admissions, enterprises are moving workloads off the cloud — sometimes to on-prem, sometimes to specialized providers. David notes, “There’s way more repatriation happening than we’ll ever hear about…”

💡 Europe’s sovereignty movement [26:00 — 30:00]

David explains how European businesses, facing regulatory pressure and geopolitical uncertainty, are leading the shift to sovereign and local clouds to safeguard data and ensure compliance.

💡 The future is hybrid (and nuanced) [31:00 — 36:00]

Vinay and David discuss the new equilibrium: it’s not about picking cloud or on-prem, but about strategic workload placement, operational flexibility, and never settling for one-size-fits-all.

💡 Lessons for the AI era [45:00 — 55:00]

History is repeating itself as AI hype builds. David urges: focus on value, question the crowd, and don’t let incentives or hype drive billion-dollar decisions.

To hear David Linthicum’s candid take on what’s next for cloud, why strategy matters more than ever, and how your business can avoid the pitfalls of past hype cycles, listen to the full episode now.

Here’s the full transcript:

Vinay Joosery: Hello and welcome to 2025’s fourth episode of Sovereign DBaaS Decoded. I’m Vinay Joosery and this episode is brought to you by Severalnines. So today we’ll talk about the evolution of enterprise cloud strategy, from aggressive public cloud adoption to the growing trend of cloud repatriation. To help us unpack the topic, today’s guest, David Linthicum. David has over 30 years in enterprise technology and is globally recognized thought leader, innovator and influencer in cloud computing, AI and cybersecurity.

Vinay Joosery: He’s the best. He’s the author of over 17 bestselling books, over 7000 articles. That’s an impressive number. And over 50 courses on LinkedIn learning. So now I’m really excited to have David with us today. We try to connect to a couple of years ago when you were just finishing your tenure at, as, Deloitte’s chief cloud strategy officer.

Vinay Joosery: And the timing was not, quite right. So. So thanks for joining us. David, where does this podcast find you today?

David Linthicum: It’s great to be here. I’m in, Loudoun County, Virginia, just outside of Washington, DC.

Vinay Joosery: Excellent. And what are you up to these days? Doing the independent, influencer thing? Right now, I retired from Deloitte last year, as the chief cloud strategy officer for many years.

David Linthicum: I had a blast doing that job. That was a lot of fun, but obviously limited in what I could, post and and and produce. So, started a YouTube channel called The Cloud Insider. The same name of the book. It’s doing really well. I think I’m up to 180,000 subscribers, and it’s growing quickly. And then, doing the independent influencer thing, doing some consulting and lots of project mentoring.

David Linthicum: So there’s lots of Agentic AI, generative AI, architectures and projects underway. I’m doing, you know, probably a half a dozen mentoring, contracts with different, you know, global 2000 companies that just need some assistance and going through and learning the ropes, with their first time through, an AI architecture, doing still doing cloud architecture, still in cloud punditry.

David Linthicum: I still have the InfoWorld blog, still doing LinkedIn learning courses. I love doing that. You know, have a fully mentored, course I didn’t go cloud career, so I’m think I’m up to seven jobs again. So just kind of getting things back where it probably worst retirement ever is the best way to describe it. Okay. All right.

Vinay Joosery: Well, you’re keeping busy. So in this episode, we’ll examine the the virgining cloud, repatriation movement in the light of the cloud journey life cycle, which I think is a term you coined. You know, David. So we’ll take a deeper look into those, into the decisions, the false assumptions, the miscalculations that led to today.

Vinay Joosery: We’ll talk about the costs of the headlong rush into the cloud, how enterprises are rebalancing their cloud portfolios and, what the contours are of today’s hybrid landscape. And finally, how CIOs and other decision makers can avoid the same mistakes in the future. But first, let me ask you, the repatriation Zeitgeist seems to be now in full flower, right?

Vinay Joosery: So to say, is it safe to say that the pendulum has swung from the public cloud to, you know, alternative environments?

David Linthicum: Yeah, I hope not. It never swings. Right? In other words, it’s it’s always an adjustment and a correction. In terms of moving, too quickly to a platform without understanding the requirements of the applications and the data workloads.

David Linthicum: And that’s what we’re living through right now. So I think what is back into, Vogue is the ability to kind of look at the environments and see what it’s costing us to run on a public cloud provider versus what is going to cost us to run on prem. And then looking at the attributes of the applications, you know, the predictability of the workloads, security, compliance things and then figuring out where they need to actually be running.

David Linthicum: So in other words, there was a mass movement, you know, ten years ago I was involved in some of it. I was not a happy camper. You know, warning people, this is not a good thing. This entry’s not going to, be kind to this. And moving mass, massive amounts of applications, thousands of applications, into the cloud, typically through your lift and shift and you just kind of moving the data.

David Linthicum: They can move a, moving application to the cloud in about a couple hours. You know, as long as the data sets are reasonably get it up and running and get it going. And so what we saw was just a huge amount of cost that came from that because they weren’t optum… They weren’t operationally optimized for the cloud platforms are removed on.

David Linthicum: And so as I write about in my last book, Cloud Computing Insider, we’re paying between 2.5 and 3 times the amount of money we thought we were going to pay for the infrastructure running in the cloud. And so enterprises are kind of hitting the reset button, and I’m counseling them not to make the same mistake in reverse. In other words, pull everything off to cloud, put it back on prem.

David Linthicum: But you know, look at the application attributes in terms of what they are, where they should be running is certain applications are going to be perfectly fine and probably should be running on cloud providers elastic capabilities. You know, the ability to provide kind of an ecosystem on demand, which is the case with AI, things like that.

David Linthicum: There’s a reason there’s a business benefit for doing it. But many of the applications that were just lifted and shifted, didn’t belong there. And so now they’re being corrected, not necessarily moving them back to a, corporate data center anymore. They’re being moved to CoLo’s managed service providers. What I call, you know, second tier, third tier cloud providers.

David Linthicum: Sovereign clouds, like Lito who runs, you know, a sovereign cloud in Germany. That seems to be a popular option in some of the German sounds right now. And then specialized cloud environments and, you know, getting to a much more reasonable cost, impact based on what they’re actually using in the cloud. Many of these applications is never we’re going to use the advanced features.

David Linthicum: They’re not going to be serverless. They’re not going to, you know, go all in with the AWS, Microsoft and Google infrastructure. And so this is about, this is about right sizing the cloud, I guess is the best way to put it. There’s, right shifting, is, I use that term as well, but it’s going to take work.

David Linthicum: It’s nothing where we’re just kind of hitting the reset button. I think that’s, making a mistake in the other way. Right. And with with respect to what you mentioned, lift and shift of thousands of applications, you know, what are some of the finer points that no one is talking about, right, in terms of how we got here?

Vinay Joosery: Example of classic example of herd mentality. I mean, everybody is doing it. So should we… what was the role of the SI’s and the let’s say even the, you know, the, the industry analysts in this?

David Linthicum: Culpable! I think if I can answer that in two words. So, yeah, we had, you know, just a big, you know, suddenly when I kind of went up and got into clouds, you know, 20 some odd years ago, I was CTO of a cloud startup, you know, back in the early 2000, and everybody was just kind of feeling their way through it, you know, I was, you know, promoting this as a utility model that may have some value.

David Linthicum: And I think what happened is like 2009, 2010, you know, everything shifted the definition. Cloud computing came out of this, and I was actually involved in working on that. And then suddenly everybody goes, okay, we got to have this stuff in the cloud. We’re going to make this move this smarter way to do it’s cheaper, better, more scalable, more secure.

David Linthicum: So everything needs to move to the cloud. And so lots of companies declared we’re shutting our data centers down in two years, and we’re moving, you know, everything to AWS, everything to, to Microsoft or to Google. And that’s what happened. Then suddenly it what it came from when I was getting kicked out of the UN meetings for promoting cloud or at best, promoting the look at, promoting think people look at cloud and they didn’t agree with that.

David Linthicum: And now suddenly people are getting kicked out of the meetings for not promoting cloud. And so obviously the SI’s and the consulting firms were all in on that because they’re going to make money. Any kind of change occurs. So you make money as a consultant. And, you know, I’ve benefited that personally myself. And so the cloud companies the SI’s, about $2 billion of marketing spending that was going on at the time, lots of the, thought leaders who are now influencers were promoting it as well.

David Linthicum: And, you know, people were, you know, just kind of made the shift into the cloud because they thought it was the destination everyone was going to. When I’m doing triage on these projects, in other words, when they are doing repatriation, I’m getting called in by boards of directors and they say, go find out why this happened. So it doesn’t happen again.

David Linthicum: And that’s the reason nine times out of ten that they give me, everybody was moving in that direction. We were told by, you know, the analyst firms were told by the consultants, we were told by the vendors themselves that it’s the right decision and everything’s going to be cheaper, better, faster. Get out of the data center business.

David Linthicum: You don’t want to be in that. And and for aspects of that were correct, but not the total, you know, reassignment of all these applications into an infrastructure that’s going to be cost ineffective. And I think that’s what they missed. So they were assuming that the costs were going to be much lower. And you would have to get about 50% cost reduction to make the migration in the cloud, a viable option for you.

David Linthicum: But they found out something different. They weren’t optimized for the cloud. So it’s very much like running a 1950s refrigerator in your house. It’s going to take a whole bunch more power unless you’re modernize it and redo it. So unless you take these applications and rebuild them as something that there’s something that’s going to be much more efficient, you’re going to find that these things are going to, you know, just be, huge cost suck.

David Linthicum: And so the enterprise is I think, you know, just had a, enough of it, you know, a couple of years ago and started to look at other opportunities, and not necessarily hitting the Hyperscaler button every time they wanted to move something in the cloud, and also taking a lot of those workloads and moving it out of cloud.

David Linthicum: And that becomes, the process of admitting you made a mistake, which is a tough thing to do. Fortunately, many people who make these mistakes, that the companies are gone. And so the poor people who are working there now have to deal with the aftermath. And so it’s just a correction for doing some things following the hype.

David Linthicum: And we’re seeing an analog of that today with the AI stuff. In other words, we’re moving to AI. In many cases, AI is a perfectly good, development model and architecture for building these applications. AI is going to have, you know, some influence on what the application is able to do and bring value back to the business.

David Linthicum: But, many of the times I’m saying that they don’t have a proper business case in use case for AI, so there’s using it because the hypes, they’re, they’re doing a genetic AI because the hypes there, and they want to be kind of a hero and their company for working on the newer cooler stuff and also follow where they think the crowd is going.

David Linthicum: They think that’s the right thing to do for their career and the right thing to do for the company. That’s never a good idea. It’s never going to be a single architecture framework. It’s always has to be the it depends. And so everybody hates, but it’s the it depends answer. That’s the right answer.

Vinay Joosery: Right. Okay. So we’ve read about you know, cloud cloud cost being the top issue of CIOs for several years now.

Vinay Joosery: It seems like everyone was saying there’s an iceberg ahead, but no one was doing anything right. When did signals start flashing red for you?

David Linthicum: When they were doing, God, because I was doing the cost models. And some of the applications I realized were going to benefit from moving in the cloud, but many of them are probably 80%, a lot of these lift and shift arrangements, they were applications that were going to use resources in such a way where they weren’t going to be cost effective.

David Linthicum: And so I hit the, you know, I hit the panic button on the panic button, but, you know, warned everybody, you know, that this could be the outcome here. And they kind of assumed that the cloud costs were going to go down. The cloud providers were going to fix things for them. And even if it’s going to cost us more in the in initially, it eventually is going to go down.

David Linthicum: And I just didn’t see that happening because the cloud providers have no motivation for reducing their costs, even though the cost of hardware, you know, has come down like an amazing amount in the last ten years, cost of storage and SSD storage as well. And, and so it was, you know, kind of the architecture, skeptic in me that started to talk about the fact that we may not be doing this, but no one was listening, you know, at the time.

David Linthicum: So I remember I was writing for at my InfoWorld blog, where I had for about 15 years, had for about 15 years, you know, my editor, Eric Knorr. Based on some of the, skeptical articles I was writing in, in the blog and I posted twice a week, and I do do to this day. And he said, do you like the cloud?

David Linthicum: Because there was so much there were very, you know, devil’s advocate for let’s not move here. Let’s think about this and things like that. But no one else was saying that. And I think that, you know, it was a message that people probably agreed with, but it fell under the radar screen with everything else out there. And so five, you know, probably five years ago is when it’s kind of started to emerge that repatriation was actually a thing.

David Linthicum: I saw some secret projects to move many of the applications that were on the cloud back to the cloud. And typically these were, you know, repeat offenders where they were, you know, just burning way and way more cash than they thought they would be. And the enterprises wanted to get off that, that cash, that cash extraction train.

David Linthicum: And then it slowly just kind of built up and it’s kind of a weird market because, it’s not talked about, with a few exceptions, no one does a press release. It tells them that, you know, remember how we moved AWS, well, we’re moving back? Maybe Dropbox and a few companies that are selling technology may do that because it can benefit from it, but they’re super secret projects.

David Linthicum: You sign NDAs when you do them, and you don’t talk about them, certainly not to the press, but it’s just and it my anecdotal, information is I’m probably seeing, you know, growth of in the last three years of, you know, 120% in terms of enterprises that are working on this and they’re moving to other infrastructures, are not moving back to an enterprise data center, and that buying data centers and building data centers that are using, you know, CoLo’s and managed service providers, which are fine analogs for doing that kind of stuff.

David Linthicum: You still get the same value out of that stuff. But it’s weird, you know, it’s like, “yeah, we’re doing repatriation.” You know, they’re they whisper it at conferences and I go, okay, well, no one’s listening, you know? And it’s a good thing you guys are, you know, able to admit that you have an issue and you have to go fix it.

David Linthicum: And that’s what architects should do. No shame in it. But, organizations don’t consider it something that’s going to reflect reflect positively on them because it looks like they made a mistake and they’re fixing it now, which in many cases, that’s the case. And so it’s this weird market where everybody is normalizing their applications and no one’s talking about it.

David Linthicum: And you’re also compelled not to talk about it. If you’re working for them as a consultant, which I respect, all the NDAs I sign and confidentiality and that kind of stuff. But there’s a lot of it going on.

Vinay Joosery: Yeah. Right. Right. So in terms of, you know, how we measure the cost, I mean, how we have we fundamentally measured the value of the cloud in the wrong way.

Vinay Joosery: So in terms of just cost, because when we talk about, for example, you know, capital allocation CapEx versus OpEx, I mean, if you look at Andreessen Horowitz, analysis of Dropbox cloud costs, right, weighing on that cost of goods sold, effectively, it reduced the market value of the enterprises. And that was kind of, that was another cost, in a way, to the enterprise.

Vinay Joosery: But, I mean, apart from that analysis, I haven’t, I haven’t seen a lot more on that, let’s say, being discussed. It’s more like people are saying, well, you know, we are surprised by the by the cost of the, you know, of the bills coming in. Well, there’s a whole category of, skill sets now and, and practice, which is FinOps for the cloud, which is kind of, merged out of the cost, crisis that cloud computing had.

David Linthicum: So people are putting more metrics and measurements and value measurements around, the cost of cloud computing. I’ve been skeptical with that movement and some of the tooling there, because in essence, I don’t think they’re looking at the right metrics. I made a YouTube video about it. And they didn’t like it so much, but the fan ops organizations.

David Linthicum: But the reality is, I think they need to tune their stuff and get a little bit better at it, a little bit more realistic in terms of how they’re looking at the architectures. So there was not a lot of cost analysis, almost none that went in to do that, didn’t do business cases, didn’t figure out I was doing some back of the napkin stuff where I, you know, realized that this is not going to work out as well as we think it is.

David Linthicum: But now that cost analysis is back in place. And so if you look at the recent surveys and CIOs are still focused on cloud aspects of it, but they’re going to control the cost in a very, in a very disciplined way, at least according to them. We’ll see if that actually bears out. In other words, it’s going to take a few years before the FinOps systems spin up and we get the cost governance in place, and we have the cost accounting.

David Linthicum: And the AI stuff is helping us tremendously in that and understanding how those things are doing. But then you get into how you’re going to define the value. And so people will say, we’re moving in the cloud because we’re going to get huge amounts of value from the agility of it and the ability to have instant scalability and the ability to have auto instant auto provisioning.

David Linthicum: And so we can move as fast as the market moves. And in many cases they put a huge value on that. But now we look at the enterprises, they don’t they don’t use it. In a tire plant in Akron, is going to say, yeah, we’re going to benefit from this agility in the scalability. The cloud is able to provide.

David Linthicum: So we’re willing to pay the premium to to remain in the cloud, remain in the infrastructure. And they never need agility. They never need scalability. They have a very well-defined, you know, slowly growing business. It organically grows over time. And they don’t necessarily need special it needs. So you can certainly run it in the cloud, but it’s going to cost you three times, you know, that.

David Linthicum: And running into an analog that’s going to run on prem. So we’re getting smarter about it. But I still find, a lot of lacking skills, in how people are doing the cost analytics. And even in the fin ops world. So the fin ops organizations are supposed to bring, you know, lots of discipline to it. I mean, the issue is, are run by consulting firms.

David Linthicum: You know, those are a lot of consultants that are in there. And, you know, even the vendors are in the middle of it. And so everybody has their own agendas in terms of how they’re promoting the stuff. And I’m not sure that the enterprises are benefiting from that moving as much as they should. I think they should benefit from some of the ideas that are coming out of it.

David Linthicum: But we may again be building false metrics and false value measurements. You know, that’s coming out of the cloud stuff because we’re not taking a good critical look at what this stuff means and how these applications are going to exist in the larger ecosystem. People always, you know, judge the cloud applications that are sitting on the cloud. I understand why they’re doing it, but the real value determination is your ability to exist in a larger ecosystem.

David Linthicum: Enterprises are running heterogeneous environments, say, a mainframe computer, supercomputers, you know, some of have quantum based systems. So how does everything kind of work and play well together, and how should the value metrics be evaluated based on that? That’s a harder question to to to do you need a consulting engagement really to get to the right, answer, you know, which I do.

David Linthicum: But many enterprises aren’t aren’t looking at that right now. They’re going to conferences. You know, I just went to IBM think this week, they’re hearing the rah rah stories. Now it’s about AI, it used to be about cloud, and they’re not necessarily thinking as critical as they should around how to leverage this technology effectively. And it’s always going to be an issue.

David Linthicum: It’s never going to be a perfect world where I was going to make mistakes. We’re not going to look at things in the end with the discipline that we need to look at it. But we’re just going to end up repeating the same mistakes and wasting a lot of money. I mean, all that cash that went to the cloud providers could have gone into R&D, could have gone into hiring additional people.

David Linthicum: I could have even given it to charity. You know, could gone into, you know, better investments in the company. You know, instead it was cash to just rolled out of the company that brought no value back into the company. Based on what is expanded. You can’t do that too long. You’re going to get you’re going to find yourself out of business.

Vinay Joosery: You know, and and actually, this is another interesting point, that you mentioned, right. Your your cost goes two and a half to three x more. Right? And that starts eating into your R&D budget. I mean, it eats into everybody’s budget, right? So, so, so you’re supposed to be using the cloud to experiment, to innovate, but then you are spending way too much, right?

Vinay Joosery: Even with these static applications, and you’re basically cannibalizing, you know, on your innovation budget. Do we know that cost what is that cost?

David Linthicum: It depends on the organization, you know, as to how much that money would have been, spent and how much value would have come back to the organism. It’s never a good idea to spend money.

David Linthicum: You don’t need to spend. You know, it’s a it’s an easy business, thing. But as far as how much value you’re extracting out of the organization depends on what what industry you’re at. I mean, if you’re a tire company, you’re not going to have a big R&D budget, to find, you know, better combinations of rubber or whatever like that.

David Linthicum: And so it’s not necessarily going to hurt you as much. You’re just going to lose the cash. But if you’re a financial organization or a technology organization, health care organization, where that research can really lead to the eureka moment, which takes your business to the next level. I think many times you’re missing those opportunities because you can’t fund those programs.

David Linthicum: You know, what’s what’s kind of, funny to me ultimately, is that the the enterprises, don’t understand it and and don’t underst don’t like the it depends answer that comes from it. So they’re looking for value metrics are white papers that are written in terms of how you measure the value. It just for you. It’s got to be, customized for your particular industry that you’re in your particular business.

David Linthicum: And it can be very, very different. Huge amount of spectrum there in terms of how much value can come back to the business.

Vinay Joosery: Right? Yeah, yeah. And I guess now, I mean the the genie’s out of the bottle. It’s like, well, okay, I mean, this thing’s costs maybe not everybody will will will admit. But you know, people are kind of moving away.

Vinay Joosery: And so how are these enterprises rebalancing their cloud portfolios and how long do you think it’s going to take?

David Linthicum: Yeah. They’re basically, evaluating the applications again. So they did a three year before they migrated them into the cloud. Didn’t do a very good close look at the applications, but that was part of the migration factories that the consulting firms would set up.

David Linthicum: Now they’re looking at them for what aspects they are. So they are hiring people, not necessarily to change them, but to assess them, in terms of how inefficient they are. And it’s very easy to do with some of the AI tools we have today. We can look in the code, we look in the data, look at the access IO, you know, IO optimization, you know, the, capability of the application, efficiency of the application, how much power it burns, all that stuff is available to you via these tools, which is cool.

David Linthicum: They weren’t around when we were migrating them into the cloud, or at least they weren’t as powerful. So that’s occurring. And so you are seeing these assessments of all of these workloads and these data sets to see, you know, what they’re doing, how they’re profiling, and if they can benefit from migration to the cloud or away from the cloud.

David Linthicum: And so we’re looking at alternative platforms. This is nothing more than, normalization of workloads on the, on the cloud in terms of understanding what they are and where they should exist, and reevaluating that and asking the question, should this be on a sovereign cloud of this thing only runs in Germany, and that’s where we have to keep the data sets.

David Linthicum: And we have a big compliance issue, which means it means we have to maintain better control over it and auditability and things like that. Should this run on a private cloud infrastructure, private clouds back in, you know, 2013, when the public cloud folks were exploding, was not a good alternative from the public cloud providers. They were awful. They were engineering projects to get those things going.

David Linthicum: OpenStack was tough thing to do. Not that way anymore. VMware, you know, HP, you know, all have private cloud options that are there. Excellent. You can set them up pretty quick. And so looking at those opportunities and just kind of reevaluating, you know, where these things should run. And also it’s never moving back to another platform.

David Linthicum: In other words, we’re not moving everything back to the one single managed service provider. When CoLo provider were moving these applications to any number of systems, where they’re going to be best optimized and best able to run. And it’s a, it’s going to be something that’s a fairly easy process to work. It’s just a lot of work to get it done.

David Linthicum: And you ask the question about time frame. I think it’s going to be ten years, I really do. I know people are saying what’s going to happen in 3 to 5 years. It never happens in 3 to 5 years. It took it took 15 years to migrate 40% of the workloads into the cloud. You know, removing the SaaS stuff.

David Linthicum: And it’s going to take, another ten years to start migrating them to the other specialized platforms. And so that’s, you know, that’s going to be the case. You know, it’s funny, I was remembering in an anecdote, you know, it’s kind of a a funny thing, people, the vendors are complaining to me because the enterprises aren’t moving as fast as they think they should move into AI, and they want to sell more AI services and things like that.

David Linthicum: Definitely the cloud providers. And the number one reason that, enterprises are saying why they can’t afford, to move applications in AI is are cloud costs are too high. So in other words, the cloud providers are taking away the money that they want used to, to move their systems in the AI. So they’re biting off the, you know, nose in spite their face.

David Linthicum: And, you know, many instances. And I’m like, you know, you guys know that talk to your enterprises. They can’t move to AI because your costs are too high. In other words, you’re taking too much cash out of them. They don’t have the budget to do it. And the poor CIOs, you know, now they’re getting the board of directors that are, you know, coming down and having a lot of pressure on the move to generative AI.

David Linthicum: It’s going to be core, you know, competitive advantage that we can make in their they can’t afford it because they moved all their applications into the cloud. That cost three times as much as they thought it was going to cost. And now they’re stuck unless you give them more money. They’re not going to be able to migrate.

David Linthicum: They’re not going to be able to, you know, take any, build the AI system you’re looking to bake. And these AI systems are going to be five times the cost of traditional systems. And the data in the process, you know, specialized GPUs, TB kind of stuff that is very expensive. So we’re in this kind of stagnant, you know, area these days where, I think the cloud providers are their own worst enemy as far as them moving into the cloud.

David Linthicum: I mean, AWS just missed their numbers. All three of the Mr.. Numbers last quarter. We’re going to see some weaknesses in the acceleration in that space. I think they’re going to benefit from AI, but they don’t have good reputations in the enterprises right now. They view them very much like the big enterprise software providers, you know, back in the 80s and 90s is something that’s, you know, a net negative tool for their enterprise and a net negative for their IT infrastructure, which is never where they want it to be.

David Linthicum: But that’s what the cloud providers are.

Vinay Joosery: Yeah. Yeah. So you mentioned ten years. Now let me ask you this. You know, geopolitically there’s been lots happening in these last, what about 110 days or so. Right. So the US is at odds with pretty much the whole world right now, right. Will we see a boost in repatriation?

Vinay Joosery: I mean, as an example, in Europe there is growing unease right over the dominance of of the top, you know, one first, second, third, you know, hyperscalers, which are all of us. Right. And it’s been there for several years. This, this kind of unease. But perhaps now the EU will do something, for example. Right. There’s a movement called Eurostat to build a European and like an independent European technology stack.

Vinay Joosery: What’s your thoughts on that? Would it accelerate that movement?

David Linthicum: Yeah. Right now. And I just did a video on this is posting in a couple of weeks on my insider channel. The Europeans are moving away from cloud about twice, sometimes three times as fast as the US space companies. And so if you look at the surveys, I don’t see them as much because they’re normally they’re not my clients.

David Linthicum: But they’re doing so for a variety of reasons that are normal, that are related to what the United States is companies are moving away from and some that are kind of unique to them. They don’t trust the providers as to protecting their data. They view their data exposed to US law enforcement as far as breaches and things go.

David Linthicum: And how would you like to, you know, be a European country? You don’t have a presence in the US and suddenly a US based cloud provider that you’re using, announced that, you know, feds showed up with a warrant and stole it, took your servers where you threw data on it. Possibly. That could happen. Are they? You could get kicked off the clouds and things like that.

David Linthicum: So the skeptical nature of them, they always been very skeptical. Didn’t like the Google, you know, camera cars running on the street and things like that is kind of coming out. And it’s very nationalist, which is fine. And they’re looking to kind of turn inward and to getting their own cloud infrastructure. And so we saw the rise of companies like Lidl, which is a grocery store that decided to build a public cloud provider only for Germany.

David Linthicum: I think they only work there a sovereign cloud. They only work in that country, you know, very small compared to the big hyperscalers. But they’re having an impact that provide good alternatives. And basically, you know, these applications only need processing memory, you know, CPU, storage, very basic services. They don’t need the, you know, 60,000 services with one announced every week that’s coming out of AWS, Microsoft and Google.

David Linthicum: They just know this something that elaborate user paying for that stuff even though you’re not using it. So that skeptical nature and some of it’s kind of based on paranoia when I talk to them and some of that is based on reality. In other words, they are looking at the risk of, you know, breaches occurring. They don’t want their data controlled by the, American based companies because there are concerned about something occurring with, you know, geopolitical things and not a political dude.

David Linthicum: But I can kind of see how things could go south for you, and suddenly you’re caught holding the you know, short end of the stick, you know, very much like tariffs. Now some companies are going to benefit from it. Some companies are not. And any time think government changes policies and certainly compliance, there’s going to be a shift in how that impacts some of these businesses out there.

David Linthicum: And so the Europeans see it coming. They have special compliance that’s way more strict than the US based systems. They have to adhere to. And so they’re being more conservative in making these decisions and in many instances, or they’re moving off of the American based cloud providers, they’re moving to other cloud providers like Lidl and things like that, but they’re definitely moving off of the major hyperscalers because they don’t trust them.

David Linthicum: I hear that over and over again. I thought it was paranoia. You know, the first five times I heard it, now I see some of the analysis coming out of the countries by some of the people who follow the European, you know, cloud, engagements. And there’s some reality there. Now, there is just going to be some risk if you’re putting your data and putting your workloads on, some, cloud based system that’s actually controlled by something it sets in another country, which means they’re under governance of that country.

Vinay Joosery: Right, right. And then as, as, as you mentioned this, this is about optimization, not abandonment. Right? So so the companies are they are finding a new equilibrium between cloud and on prem. So as part of moving us from the US Hyperscalers, they could put it on prem, or somewhere else or, or.

Vinay Joosery: Yeah. Or. Yeah. So local, local cloud, maybe, you know, better cost structure. And, maybe it seems like it’s a bit of a get out of jail card now. It’s like, okay, so, you know, at least for the, for the, for the companies that feel affected by these geopolitical issues, I can just say, hey. Yeah, we, you know, we gotta get out. So, Yeah. So can you conceptualize, the hybrid landscape?

David Linthicum: Yeah, I think it’s a term that’s overblown. Because everybody’s going to have a hybrid infrastructure. If they’re using a public cloud provider, they obviously have some on prem. You know, even if it’s just, office automation servers that, that, that are working, that, that are there. So it it’s always provided us with the best, degree of options.

David Linthicum: There was never going to be 100% moving into the public cloud provider, as that was, people were saying that I just that’s not going to happen. And there’s never going to be 100% movement back to the on prem system. So we’re always going to have a mix between cloud providers, typically multiple clouds, you know, and now with the new sovereign clouds and the specialized clouds and the second tier clouds, you’re probably some companies are having a dozen clouds and you deal with the SAS providers, you know, Salesforce and you know, all those, service now and all those, types of built.

David Linthicum: And you’re getting into, you know, hundreds of things that you’re managing where your infrastructure is sitting off, off prem, in the hands of a SAS provider or a cloud provider. So we’re definitely in that world. But obviously the prem infrastructure is should be some sort of an architectural option where we have engaged in owning platforms. We know where they are, we have access to them, we’re able to expand and we’re able to expand the storage systems that compute systems to run.

David Linthicum: Workloads are going to be more apt for running things on prem, and I always encourage people for maintaining that infrastructure, because it’s always great when you’re an architect to have options and sometimes the options are going to be very limited to what’s currently in-house and what skills are currently there. And so it’s much more advantageous for them to have skills in dealing with on prem systems, even mainframe based systems.

David Linthicum: And so, you know, people are concerned about, well, I got to scale up and get my enterprise, my, my, data center people back. Now, you don’t you’re going to partner with, CoLo providers, managed service providers. All that stuff is outsourced. Now, you don’t even see the hardware. You know, if you can, you can engage with a CoLo provider.

David Linthicum: They’ll find you a contractor to set up the cage in the CoLo and bring in, bring in the servers, screw them and power them up and just throw you the IP address. And you have never even to travel there. So in essence, you’re getting a cloud where things are provisioned with people and, you know, manual stuff. And so all of that infrastructure should be in place for BCDR in other words, you have to be able to look at, you know, hitting the reset button.

David Linthicum: If something occurs in the cloud, we could have some cloud outages that are devastating in your ability to have backup systems that are running on some other platform or even some other cloud are going to be a necessity. I think, for many of these organizations. And again, the best thing for an architect is you having options when you I think the fallacy people have a, you know, notion of architects who show up, you know, to a project and they go, okay, you can use any technology you want to solve the issue.

David Linthicum: That’s almost never the case. You know, it’s going to be we have an on prem infrastructure. We’re using traditional lamp stacks. You know, Linux and you know that’s what’s currently running. Those are the skills in-house that are maintaining those systems. And we use AWS because we have an agreement with them that goes on for the next few years where we bought you know, pre bought, you know, $3 million for the credits.

David Linthicum: I don’t know if they actually do that, but I’m just using that for this scenario. So you’re going to be limited by what they allow you to use. And so I always tell my architecture students you’re not the decider. So you’re going to be you’re going to put these limitations around you because some, some sort of a business engagement that coexist and you can’t fight it.

David Linthicum: You have to build an architecture in that kind of array. And so you’re going to be much better off if you have on prem options where you’re able to leverage commoditized hardware. Because especially with these AI systems that people are looking to build. I mean, you know, you think applications running applications in the cloud was expensive over the last ten years?

David Linthicum: Running AI applications is going to be, you know, ten times the cost of what we could do in replicating, these environments on prem and also, you know, getting away with using C… traditional CPUs versus GPUs and your ability to optimize those environments much better because you have a better control over them. And also, the cloud providers aren’t optimized, for AI as much as people think they are.

David Linthicum: So they’re going to be inefficient running on there as well. So it’s options. And I don’t think you have to open up too many options. But I do think you have to have skill sets in-house to deal with the on prem. And you can always allocate more money into the cloud if needed, out of court, more money into the on prem skills and those operations if needed.

David Linthicum: So the choices, the options, the ability to fail back, you know, that’s going to make a hybrid cloud, you know, way more attractive. Very what are your thoughts about the private cloud tech stack right then VMware, Broadcom Nutanix. You you work with OpenStack before there’s other ones like CloudStack.

Vinay Joosery: Yeah. CloudStack, OpenStack were the big ones that were we were dealing with like back in, you know, 2013, believe it or not.

David Linthicum: Much better than they were, and used to be. The private clouds were never a good option. And I ran into this a ton. Where they go, I don’t want to move in the private in the public cloud. I want a cloud. So I guess what? I’m going to go with an OpenStack cloud. And, you know, here’s an open stack code tree, you know, go set it up on our private servers.

David Linthicum: And it was an engineering project is probably the most complicated, massive stuff I ever went through in my entire life. Not not to, you know, take anything on OpenStack. And and it was an open source project. Those things always work out that way in the initial infancy. Now they’re better. The OpenStack stuff is fine. If you want to use it.

David Linthicum: It’s fairly turnkey to install and it’s, you know, it’s in its seventh generation. So it’s been hammered in and in much better quality. And it was ten years ago, the private cloud providers who are commercial the focus, the Broadcom, the HP’s Nutanix, things like that. Hyper converged systems are also have great public and private cloud providers in other words very turnkey, very easy to install.

David Linthicum: They’re obviously not as easy as setting up a public cloud provider, you know, virtual ecosystem, but they provide an AI ecosystem on demand. They provide application dev test stuff on demand, DevOps and DevOps. Toolchains. On demand. You can configure it however you want it. Most of the software providers and tool vendors run on those platforms and support those platforms.

David Linthicum: So the options are there and so in other words, if I’m trying to get the benefits of using, public cloud provider, in other words, easy button for lots of different things. And that’s why people use public cloud providers mostly. You’re going to see that in the private cloud options. And they’re going to be much cheaper because you’re paying for the hardware, and you’re owning the software.

David Linthicum: You may pay a license fee, but it’s definitely not going to be a cloud service where you’re getting a bill every month for how many you know, how much CPU utilizing memory, IO and network bandwidth, ingress, egress fees. Those things are terrible. And it’s going to be, I think, a better option for many enterprises and certainly for the AI stuff.

David Linthicum: I just kind of see that and looking at the cost efficiencies of running it in a public cloud provider, the hyperscalers, the big three, you know, versus the other options out there, that’s going to be a much better option. Also, specialized cloud providers like, you know, Core Weave and Lambda Labs that provide GPUs a service, they’re very less much less expensive than going with the major cloud providers.

David Linthicum: They have to be. So these are purpose built clouds out there, these private clouds that are just going to be much more cost effective options for many of these workloads that are out there, and they’re going to be much more easier to maintain, configure, set up architect build systems on than they were ten years ago. Like I said, they were a nightmare.

David Linthicum: I you know, that’s something I almost had PTSD from those days because it mentioning about, you know, GPUs service. I mean, you know, Nvidia has this neat, operator, you know, the Nvidia, the Nvidia operator, which basically allows people to, you know, easily slice and dice these GPUs and just serve them. So I guess, yeah, I mean, that’s that’s, they make it really easy to, to become pretty much a cloud provider if you, if you can buy the hardware, you know, Kubernetes on top of it.

David Linthicum: And yeah, I, I think I could become a, I mean, a cloud, you know, set up a cloud ecosystem, system, including multi-tenancy, you know, in about five days. I mean, ready even putting the billing system on top of it, which means internal billing, you know, chargeback, show back, you know, things like that. Just because there’s so much out there now, you just configure it, integrate the systems, you know, where as a huge engineering project, you know, but, you know, back in the 2000, I was, you know, taking software and turn it into multi-tenant SaaS providers.

David Linthicum: That was a, consulting firm I had called Blue Mountain Labs that did that. That was a major project. And making those things happen and looking at the same sort of configuration. Now, I would take a very short amount of time because the automation and the technologies and the tool sets that are around setting those things up, and you can even use some of the stuff that’s on demand.

David Linthicum: So in other words, it’s in the cloud. You build your cloud and you run it on prem, which is, you know, kind of a weird way to do it, but that kind of stuff exists, right?

Vinay Joosery: So am I right to say that a lot of talk is around compute storage layer, but maybe not stateful? And the operations, you know, DB is a whole, but there’s this but there is.

Vinay Joosery: But I know that’s my impression that, you know, like OpenStack had this, this database as a service project. Now the name has escaped me. I can’t remember the name now. But, yeah, they they had an open source project, but nobody uses that. And, you know, the people who worked on it x years ago just, you know, they’re gone.

David Linthicum: Yeah, Trove. OpenStack Trove, that’s what it was. And, yeah, that’s kind of an orphan project now. No one’s really working on it. And, yeah, the the stateful, the storage, the database stuff, that’s always the hardest thing to move. And so that’s kind of the lagging thing that’s that’s holding a lot of companies back. But even that’s starting to become easier.

David Linthicum: There are lots of options now, and the vendors are realizing they have to make it easier to migrate databases and stateful workloads. You know, it’s a business opportunity. But it’s always going to be a challenge. Moving data is always the hardest part, and it’s the thing that everybody underestimates, both in terms of effort and in terms of cost.

David Linthicum: So I think we’re going to see more tools and more managed services in that area over the next couple of years. It’s just inevitable because that’s that’s where the pain is. And where there’s pain, there’s opportunity for the vendors.

Vinay Joosery: That’s right. And I mean, you mentioned, you know, all these options that are out there now. Do you think that makes things more complex for enterprises, or is it just giving them the flexibility they need?

David Linthicum: Both. I mean, it absolutely adds complexity. It’s a it’s a management challenge, and it’s a skill set challenge. You know, the more environments you have, the more you have to monitor, secure, optimize, the more skill sets you need in your team. But at the same time, that’s the reality of the world we’re in.

David Linthicum: Enterprises are heterogeneous. They always have been. And I think it’s actually an advantage because you can leverage the best technology for each use case. So it’s just a matter of having the right tools, the right processes, and the right people. And that’s the hard part, not the technology.

Vinay Joosery: Yeah, yeah. It’s it’s the people, it’s the process, it’s the culture, it’s all of those things. And, you know, people still underestimate that. They think, you know, if I just buy the technology, that’s it. But actually, that’s the easiest part. It’s everything around it that that’s the hard part.

David Linthicum: Absolutely. The technology is easy. The people, the process, the culture, that’s the hard part. And that’s what slows down these projects and makes or breaks success.

Vinay Joosery: So I guess, as we kind of wrap up here, what advice would you give to CIOs or IT leaders who are maybe feeling overwhelmed by all of this? You know, the repatriation movement, the hybrid options, the AI hype cycle, everything that’s happening. What’s the best way to approach these decisions?

David Linthicum: Don’t panic. That’s the first thing. Don’t don’t chase the hype. Don’t follow the herd. Make data-driven decisions that are based on your business needs and your use cases, not what everyone else is doing. And really focus on getting the right talent, the right culture, and the right processes in place.

David Linthicum: The technology will follow. You know, the technology is easy. It’s everything else that’s hard. But if you get that right, you’re going to be in a great place, no matter where the pendulum swings next.

Vinay Joosery: Great advice. Thank you, David. It’s been a pleasure having you on. There’s a lot to think about here, and I think our listeners will really appreciate your perspective.

David Linthicum: Thanks for having me. I enjoyed it.

Vinay Joosery: And thanks to everyone for tuning in to this episode of Sovereign DBaaS Decoded. Be sure to check out David’s work—books, articles, and YouTube channel The Cloud Insider. And if you enjoyed this episode, please subscribe, share, and let us know what you think.

Vinay Joosery: Until next time, take care.