Mainstream Adoption of Cloud-Native Applications – Peter Smails, SUSE

This is Textron TV. Hey guys. Thanks for the throw.

We're here with Peter smells who is General Manager for Enterprise container management business unit for suse there's a mouthful for you and we're gonna be talking about what exactly is the progress we're seeing around Cloud native these days and what should we expect in the future Peter? Welcome the show. Thanks for having me.

I feel like we've all been on this Cloud native Journey for quite some time and yet if I go talk to organizations. Yeah, I would say the percentage of applications that they're running that are container based in Cloud native based is still relatively small compared to the Legacy monolithic stuff that we've seen in the past. So, where are we on this adventure from your perspective?

I think that's an accurate. I think that remains an accurate assessment, you know, I mean it was it was it we're still easily sub 10% And so the wave I think we're collectively riding. I say we as in just the overall community of Enterprise container management Technologies and solutions are ultimately gated to you know adoption right of either, you know, net new containers, you know, net new applications getting created transformational Legacy, but That to me is it's just that's opportunity.

Right? It's still I mean the market is still practically speaking nation, and I know we'll come back to this but it kind of speaks to it. It aligns that sort of my word is maturity.

We are still riding this wave of Volition of the market and sorry there's some interesting trends that we're seeing related to that maturation, but you're absolutely right that we're still. I don't say early days because that sounds too, you know like sandbox C and you know non-production type stuff and we're definitely not in the early days, but given the number of applications that exist and given the level of infrastructure maturity that's required to take this Legacy, you know, take all these Legacy applications or some percentage of them, you know Cloud native, you know to have you left and that's gonna take time but we're definitely seeing the market sure and clearly it's in indicative in our business growth and that kind of stuff. What are you hearing from folks in terms of the challenges that they're running into because it seems like you know, we aren't getting some hands-on experience now and a lot of the feedback you hear from folks as well.

This stuff isn't easy. Yeah, but if we had a dime for every, you know, sort of every technology thing that folks ended up saying at some point in time, it wasn't easy again, the more people that are saying this isn't necessarily easiest also an indication of adoption if you will because there's just more, you know, they're more people talking about it, but In terms of what worse, you know from our vantage point we're seeing I'd say two things sort of jump to mind in terms of what I mean. Number one.

I don't think anyone's ever said that you know kubernetes was where you know container management was easy and that was part of what you know, kubernetes wasn't an engine for that's why Rancher, you know, our Flagship product has been so successful because it basically makes kubernetes and containers easy. That's the whole premise of the product but We're kind of seeing two things happening right now. Number one is and I think this is a strong sign of maturation is that security is probably the hottest topic.

That we talk about now and as you know, we have a you know, we have a worse in that race with our new Vector Technology that we've had in the stable, you know over the last, you know over a year now since the acquisition in 21 and so security has become a front and center conversation and not interested enough necessarily is just an add-on, you know sort of like, oh you have containers now you need security. It's like security. It's just so front and center that it also creates kind of its own entry point.

So I wouldn't I wouldn't say that that ties directly to the Simplicity comment that we were just talking about before I think just more from a business and strategy standpoint. It's absolutely the level of interest that that we're seeing. Yes, but specifically with new Vector but in general with just the conversation around security is you know, is is that that's a big one.

That's that's definitely front and center and so lots of conversations going on that front and obviously we're doing you know, as we talk more about this. We're always doing lots of fun, you know, lots of lots of fun stuff on the security front with the new Vector Tech. The the second thing that I would mention is the continued convergence if you will or closer and closer alignment collaboration between say it Ops and devops and I think this speaks directly to your earlier comment about containerization of applications.

Look the reality was, you know, rancher in particular sort of where we came from and where we cut our teeth was largely on the itop side. You know, how do you know people trying to deploy kubernetes to service? How do you make it easy to stand up a container and to deploy the container and then manage multiple clusters and you know, that's where Ranch really cut its teeth and where you know, it kind of gained it's it's you know, it's it's Fame.

But the real heavy lifting that's gonna drive that increase that that wave we were just talking about in terms of of additional application development is on a devops front. So increasingly more conversations that we have. We're getting closer and closer to the devops communities the devops folks within the organizations that we're talking to are more vocal.

They're more involved in the discussion. You know, they carry they carry an increasing amount of cloud because they're started there that many more projects. Maybe they're not in production yet, but there's that many more initiatives.

Around more, you know, you bring the more and more people in fact interesting metrics. It's like if you look at it, there are 22 million developers in the world. Yes, give or take I heard that that number may be a year too old but there's something on the order of like 22 million developers.

I'm not saying that all of them are doing you know containers and kubernetes, but where we're seeing what's backing up our view on this this increased focus on the devop side is things like Rancher desktop, which you're sure you're familiar with which is literally that is our manifestation of Rancher, you know, it's not about the cluster management. It's about the ability just put that developer experience right in your desktop make it easy to stand up containers just stand up a cluster Etc with Grant your desktop that things growing like a weed, you know, open source project tons of adoption there and I think that just speaks to the fact that more and more people are taking their tires. So, you know if I look at it and macro Right.

Now we have, you know north of you know, something you spit ball on your 40,000 plus, you know active users of Rancher, you know on the management side. Practically but then then compare that with 22 million developers, you know, if we can get one percent of those people using Rancher desktop, if you just you know, making that number up now you're talking about hundreds of thousands of people, you know that are now developers focused on container technology. And so that's another so with things like Grant your desktop.

We're seeing really strong adoption of the Rancher desktop stuff. We're seeing a lot of interest in our opinio technology, which you probably that's another open source project, which is like really focused on sort. It's a pass so it actually works in collaboration with ranchers.

So you sort of give people more opinionated way to just package things up and simplify the the deployment, you know, less cicd work kind of mask all that. So any and all things related to Making it easier for the developer folks to be building those apps developing those apps Etc would be that'd be the second the second Trend and then one more and then I'll send I promise I'll shut up. The the other interesting thing is particularly on the edge, you know, you've you've I'm sure you've heard us about Harvester.

You know Harvesters is our age is another open source project. You know, it's Cloud native HCI, but where we're seeing interest there is largely on the edge where folks are. You have these environments where they want to essentially build out their Cloud native their ability to manage these Edge deployments in the cloud native way, but the applications that they're potentially running could be a POS.

It could be who knows what the application could be. But some of these applications are still, you know, their VM based their traditional applications. So they haven't been you know, they're not Cloud native apps per se so we see organizations that want to manage Cloud native, but then they're but then they've got this there in this tweener state where they've got some potential Legacy applications that they're going to run continue to run not literally, you know, turn them into microservices but basically run these apps in a containerized environment and that's where stuff like Harvester comes into play.

So that's I'll pause there but I think all of those two to answer your question very very concisely security definitely continues to pop up because we're and we're seeing that in the interest in new Vector in the discussions, you know, the developer experience continues to become increasingly, you know, increasingly important topic and we're sort of seeing that through things like Rancher desktop interest in opinio interest and then sort of this convergence. If you will of sort of traditional Legacy and Cloud native in our world, you know large and the edge side, you know, we're seeing that through things like, you know interest in our Harvester technology. But I do I think that all speaks to kind of the evolution of the market.

And we underestimate what it would take to bring developers along because we've been at this now for I think you know seven years maybe or depending on where you got started in the thing, but maybe we assume that developers were gonna play at a much lower level of access and then it turns out maybe they don't want to deal with a wall of the animal files. Yeah, what and I well I'm an internal optimist. So I don't know like I can't speak to whether we we do whether we overestimated or underestimated whichever way that you look at it, but I think that the At the same time as you know, like the more you the more you give the more that you're asked for, right?

I mean, obviously the the people obviously want this to be is if we look back in 10 years five years, whatever they don't know exactly what the timeline is, then you know, she and I you know Cloud native development environments and so forth will be as common as go or you know, they'll be just it'll be all it'll be completely embedded in the in the in the DNA of the developer community. At least. That's the way I think we're collectively riding.

As we kind of ponder this are we in danger of now going to the Other Extreme where kubernetes and containers is the yeah. It's the fashion platform of the day, but maybe a lot of the things we're building aren't fit for kubernetes and we gonna have to figure out how to use everything from serverless Frameworks and one side to traditional virtual machines on the other and maybe we need to have some better advice for folks about what to use when yeah and more. Can you expand on that like more specific you be talking about like serverless versus containers like what you know, like what were you going with that?

Well, there's a lot of folks who are starting to talk about a lot of the applications that are starting to show up on kubernetes, maybe not ideal for that platform and we need to have a conversation amongst ourselves about which platforms to use for a while when And what do you what's your from your Vantage Point? What give me an example if you would of like you talked about stateful are you talking about highly transactional like what you know, what what kind of what are you hearing? Well stateful is one example where it is a religious argument almost between folks about whether or not you should use kubernetes for a stateful application.

They're also folks who just think that a lot of the applications to quote Kelsey Hightower recently or garbage, so Now this on the first one. I mean look, I don't think I don't think we'll have I mean, I think we all know this there's not going to be the entire world is not going to go completely to you know, Cloud native not every yes, right not every application is going to necessarily be built, you know containerized or go become container native. I think that would be you know, that would be a stretch so I completely I generally agree with the comment.

I don't agree with sort of the blanket. There's a discussion to be had about whether or not you want to be running stateful applications and that environment because there's a different set of requirements. I think the juries out to a certain extent we've got plenty of people that are running stateful applications.

I mean, this is one of the reasons things like why we develop Longhorn You know is because you need if you're gonna run stay flaps, he persistent storage and you know you the adoption numbers on longer are are fantastic. So that's not to say that every stateful application is is getting containerized but I think that the, you know, we're on the stateful side. We're building the technology we think is fit for purpose to support the requirements of those applications to your second point though.

I'm interested to hear obviously Kelsey is a very interesting guy with a very, you know with great insights and such. So was your comment was his comment that there are a lot of Sorry could container-based new container apps that are garbage like because he's basically was he saying that there's old apps that are garbage that you'd never want to move over. Was it more about what's getting built new?

That's kind of garbage well to be specific. He was basically saying a lot of the newer applications probably should go to a serverless framework and shouldn't be on kubernetes in the first place because they're just not they don't require that level of sophistication and complexity and the underlying management platform is built for something that is highly distributed. That would be deployed to cross a hyperscale or kind of environment.

That maybe is bias, but that's his point of view. No, but I would concur with that. I think that may and I well I can't concur with whether it's something as garbage or not.

I won't you know, I'll leave that. I'll leave that to Kelsey to apply his judgment on that, but I think that the What he just explained if kind of ties back to the roots of what we've been trying to do from day one with the Rancher stuff, right, which is that ultimately it is about abstraction. I mean, if you remember the you know, we still talk about this now.

It's sort of like Computing everywhere. You know, like ultimately it's this notion of computing everywhere, you know, you leverage contain you leverage kubernetes everywhere in order for you to compute everywhere. So right ones run many, you know and our resident with the ranch or stuff has always been about multi.

It's oh, you know, there's the value isn't in the distributions event, you know, the distributions are going to commoditize they're going they need to be everywhere anywhere you want to compute, you know, and the cool thing about that is with things like Kate your ass like I mean playing strange automobile satellites drones like tanks there. I mean kubernetes is you know, it's not literally everywhere but the applications are really myriad. But I do agree with Kelsey that the that, you know, we're kubernet kubernetes is not.

K through S is lightweight admittedly kubernetes overall when you is not light if you will or it's not super simple you can you can use tools like Rancher and the things that we deliver to make things simple, but you know, the technology was largely designed to create a complete abstraction layer for you to be able to compute. Ultimately compute anywhere. So again, I don't know specifically with Kelsey was talking about on the on those on those.

I'm sure there's certain applications that it's complete waste of time to try and containerize them because there's no point but you know the but the but there's still I mean ultimately there is still a I think it ties back to the fact that there's supply and demand now, which is we have an increasing number of people that are using containers that are you know that are investigating containers. We're still relatively speaking. We have a lot of maturity, but we're still early like we haven't, you know, the hockey stick continues largely on the developer side and I feel like that's You know, if you accelerate the developer side of things and and to your point, that's not an it's not like this.

Oh my God this the first year we're talking about devops, you know, it's not a new discussion and we've had if you remember Rio we've had multiple cuts it at trying to do pass and trying to make it simple and when you go opinionated You know, that's that can be hard because you you have to predict. Yeah, we got it right with Rancher because we've just made everything so simple. I think the developer side is just because there's such a myriad universe and community of developers.

It's harder to to find that common ground. And so I think there's just it takes longer. It's it can be complicated.

There's gonna be more thrashing. But again, I come back to things like, you know Rancher desktop just in general like the more we can watch these those communities grow and mainstream then there's obviously give me more and more need and more and more questions requirements from additional developers, but that that's all goodness. You know, if you just look at an adoption LED go to market model, that's all goodness because it's not There's no.

The issue isn't well, nobody's interested in ranch or desktop, you know like that. That's not the problem, you know kind of thing like the prop the challenge that we have collectively is making this stuff simpler continuing to make it simpler on the Rancher side. You know from a management infrastructure management side and then continuing to simplify things in mature things, you know on the dev side.

So what ultimately will decide the issue as to which container platform people will use which distribution which passes what are they going to look for from providing? That is a very good question. And I and I think to a certain extent, you know, you know our you know, our approach has been oh as always been very what we call open interoperable and that's not to say that we don't care.

If you will, but the nature of of like we don't I mean honestly we can tell you like as it relates to the distribution. We've always said use the distribution that's purpose built. Like if you're running in AWS then use the AWS distrke like we have always and that's why for example with Rancher.

We invested in providing full life cycle management. So our opinion has always been used the distribution values and management, you know use the distribution that is purpose-built for the environment or multiple distributions. We don't know you could use a myriad of different ultimately the to a certain extent.

Rancher customers or our view on that is that they don't in terms of distributions and things like that. They don't need to make that decision. They can make that decision based upon business requirements and what they learned about the different Platforms in terms of what applications I think that's going to be an ongoing discussion of You know, what is the I mean, there's so many questions.

It's compliant. There's certain applications that will never move over. They're gonna be traditional Legacy on-prem apps, you know that are never going to go up for security reasons for compliance reasons for air gapping reasons.

Whatever the case may be. They're gonna be security issues as to why they'll be compliance regulatory reasons why they won't do it. They'll be certain applications that just for technology performance, you know reasons may not make it so I don't think it's a I don't think there's a I don't think there's a singular sort of Playbook.

If you will Lisa's of today that says yeah, there's certain guard rails. Obviously, you can provide good Insight on but I think a lot of that is continuing to evolve, you know, there's no sort of ultimate decider on that if that makes sense. All right folks here in here.

It's a Brave New World. There's gonna be a lot more Cloud native applications for sure, but nothing else that right. Nothing else seems to go away.

Hey Peter. Thanks being on the show. Thank you for having me.

Look forward to chatting with you again. Take care. All right back to you guys in the studio.