 Welcome to Amsterdam. And KubeCon, CloudNativeCon 2023. Join John Furrier, Savannah Peterson, Rob Streche, and UPSCOT. As the Kube covers the largest conference on Kubernetes, CloudNative, and open source technologies together with developers, engineers, and IT leaders from around the globe. Live coverage of KubeCon, CloudNativeCon 2023 is made possible by the support of Red Hat, the CNCF, and its ecosystem partners. Hello, everyone. Welcome back to the Kube's coverage here at KubeCon in Amsterdam. It's Europe. It's CloudNativeCon. I'm John Furrier, host of the Kube at Rob Streche. Analyst in the Kube. Got a deep echo here. CTO of D2iQ, formerly MesoStreet for the folks in the industry going back in the beginning of DevOps and containers and all this great stuff happening now. Deepak, thank you for coming on the Kube. Thank you for having me. So we've been unpacking a lot of the trends here. The Kube's been to every single Kubernetes KubeCon since its existence, I think that our 13th episode here. It just gets better and better on the mainstream adoption. Still so much work to do, simplicity, a lot of challenges there. Although they score themselves great on portability, Rob and I think that Rob's more on the, they need more work on portability. But overall, Kubernetes, being more enterprise grade, the progress is there. WebAssembly is looking good. A lot of devs coming together. Infrastructure Express Code is a done deal. We're starting to see it go mainstream. Containers, virtualization moving into containers, big migrate, a lot of headroom from a business standpoint. A lot of more projects coming down the pike. You guys are the center of it. Where do you see it going on here? What's the top story at KubeCon this year from a Kubernetes standpoint? So Kubernetes, over this years, I would say the day zero aspect of Kubernetes has simplified quite a bit. It has become a no opt for anybody to bring up a Kubernetes cluster. But Kubernetes alone is not enough to run in production. I would say it's an enabler to create a platform. So that's why it's an API based and it has been very successful. What we're starting to see is people moving higher in the stack and abstracting some of the complexities of Kubernetes to simplify it. As bigger organization, Global 2000s are adopting Kubernetes, they need that simplicity to spin up a cluster which is production ready, which is secured by nature, which has observability, which has monitoring, logging. All traditional aspects of a cluster that they have always seen in another platform. Now you would see there's integration points happening in Kubernetes as well. Yeah, and I think you had some announcements that, to me, signal that Kubernetes really has a lot of momentum. It's the fact that you're going into air-gapped environments, the fact that you're going into managed service providers. I think to me that would lead me to believe that there's newer people coming in looking at these microservices built applications. Exactly. In fact, one of the CNCF surveys from two years back have showcased that all the innovations that are happening in any industries, you take blockchain, you take digital reality, genomics, robotics. In fact, the chat GPT, which I'm pretty sure is known, named in every house, is being run on Kubernetes. So Kubernetes has become the de facto platform for that is fueling innovation in every area today. So one of the rise of Kubernetes impact has been the enterprise application award that we heard earlier. That's not a really word. As it gets more mainstream, Rob was talking about the platform engineering piece becoming much more defined from a practice standpoint, where team formation, roles. You guys talk a lot about this. This is the hot trend here, KubeCon. What does platform engineering mean today? And how do you see that rolling forward? Is it a modernized version of IP with a reconfigured or what's your view of platform engineering? Platform engineering is a team that is responsible or it's a concept about creating reusable standard tools that are scalable, that are consistent, that brings in the consistency in managing the platform. There has been a trend, especially which was started when the cloud services came a thing where developers started doing their own things. And that led to correlation of different toolings, a different way of doing the same things. And it was a costly thing when they do that. What platform engineering has started doing is consolidating all those where it was standard tooling, where developers can still achieve the agility that they hope to achieve with a cloud native platform, but in a much more consistent and cost effective manner. What's the biggest impact to IT from an infrastructure's code standpoint as DevSecOps becomes day two like, more day two like as we were saying, what's the big impact? What's different about platform engineering than classic IT? The platform engineering, classic IT was responsible. They didn't have the insights into what the developer requirements are. What they used to provide is the infrastructure on the request basis. Platform engineering is, I would say much, if I had to use a word, is an aware IT ops, which is they now understand the platform through, like for example, Kubernetes. And they configure in such a way that takes away the burden from developers and standardize them into a platform so that developers can focus on their developing the application, making their life easier. Like for example, deploying the cluster, making sure the cluster is elastic, secure, easy to use. Whereas the developer can focus on their business applications. And I think what's really key to that, and I think it is that centralization. I'd like to say my former employer, Amazon, with their two pizza teams kind of started a whole movement where you're going to have, okay, eight people and a manager, eight people and a manager using different systems, different methods, different SREs, different DevOps on different infrastructure and different tooling. It would seem that the CIO is really helping lead some consolidation around it. Is that what you're seeing? Yes, exactly. So what led to, one of the, DevOps has been very good in achieving the agility and time to market. But what it missed is it led to the problems like cluster sprawl, cloud sprawl, where developers are spinning because their main focus is how to get their code running in production easily. But for the organization as a whole, it created some of the challenges, which is what is simplified by having platform engine or IT team, which has the experience and expertise to use like something we call instance platform engineering, which is using the right platform, right tooling and providing an environment for developers where they can still feel self-managed and self-serve, but under the guardrails of the platform that the platform engineering has provided. Deepak, Rob and I were talking about day one. Remember we talked about developer productivity? Oh yeah. So this gets at the heart of developer productivity because obviously platform engineering has to enable not just a service organization like the classic IT. Developers are setting the agenda. Right. So developer productivity is number one. So assume platform engineering is in place. What are some of the developer productivity challenges and opportunities today? Obviously S-bombs, supply chain is a big problem. Now you've got Genovi, putting code in the mix. We call it code pollution coming in. Stack overflow just ban chat GPT. They don't want any of that in there. It's a nightmare. So like you have all this hardened kind of mindset going on. But yet developers want more data. So how do you see the whole developer equation from a productivity standpoint to a consumption? Because we're not B to D anymore. It's B, D, B to D. Right. This is developer. I mean, there is a consumer. Right. So I give an analogy with the car analogy. It's like if we ask a driver to configure the engine first before driving the car, they'll say they can't drive the car because their whole time will go in configuring the engine itself. In developers without this platform engineering is something like that. They have to first configure their system before they can use their system, which kills their productivity. What enables their productivity is a platform that is secure. Right. That is scalable. That is user standardized tooling that is consistent and easy to use and easy to deploy. Right. So the developers can come in and deploy their application without worrying about whether it's a secure way of deploying the application. Is somebody attacking their applications? Right. All those concerns are being taken care by using a platform that provides this out of the box. How would you rate the productivity of today's developer on a scale of 1 to 10? Obviously, don't compare to old way, but like in DevOps. As DevOps becomes more mainstream, more tooling is out there. What's the... Before even I give the rank, I would say it gives a perception that the DevOps is becoming agile, but what's happening is it is creating shadow IT within those teams. Give an example. Because operations are complex by nature because of the things that you have to take care, security that we have talked about. What happens is in a typical team, the senior developers takes this responsibility of providing the operation. What ends up happening is the reverse effect of what the organization was expecting their senior engineers to be doing, which is like focusing on developing the business application instead of working on the operation. What happens is they start helping the junior engineers and it creates a model which is reversed to what the organization was expecting. And that leads to the cost inefficiency and effectiveness. And we call it as like a shadow DevOps. Yes. You agree? I would agree. But this is why platform engineering has gotten... It's becoming more mainstream is that it's bringing it back together so you're not to your... I think to your earlier comment, it's IT that is developer friendly versus an adversarial where that was very adversarial back. Even if you look at how databases were dealt with in VMware. Right. And I kind of look at everything that's old is new again and we kind of learned. It seems like, and very interested in your opinion on this, it seems like we're getting through that at a faster pace where it's coming together. Like the teams are coming together. One of the differences I would highlight in comparison with the traditional centralized IT was there was a ticketing system that used to be in place. Like the team would... If they need an infrastructure, they used to create a ticket for that team. That team will act on it. So that used to delay a lot of productivity or agility that the organization were looking for. That's why DevOps became so attractive because now you don't have to do any of the requests. You can just go and create your own thing. What Platform Engine allows it to do is they still maintain the agility and the self-serving and self-managing nature of DevOps, but they provide an environment that is being custom made for that environment. Instead of like centralized IT figuring out all the time based on request. Platform Engine is... That's why I was using the term aware centralized IT. Aware of what would be the need of developers and providing that environment. They still got to do it. Yeah. The job changes. It's faster. Agile is the key word. But the shadow IT example for devs is very interesting because they want to go faster. There's going to be consequences to this. It's got compliance, it's security. What are some of the things you've seen that's been bad that companies can avoid? Because I'm sure it's happening in other companies that no one knows about. But how do you identify it? And then what are some of the consequences or benefits of it? Benefits of Platform Engine. Well shadow IT you could argue was actually good because that helped the DevOps movement. So is there a positive or negative impact to shadow DevOps? It ultimately leads to a negative impact. That's why I was reading one article where CIO is saying after three or four years of their investment they are now looking for return on that investment which they are not seeing. Because what happens is initially it gives an impression that everything is moving fast but over time it builds up the maintenance and the operational burden of managing that platform. And that's where the organization realized that they missed on standardizing tooling. Like for example they may start with one cloud provider. Different teams would use different cloud providers. They maintain different toolings, different teams to different skillset and they lose the efficiency or the benefit of having some kind of a standardization. So the shadow IT and then it because it's not one time you launch a code and you forget about it. There's a continuous upgrade that happens. There's a continuous maintenance. And the teams are not benefited by having this cross team knowledge sharing because they do it by themselves. Yeah and I think that it's been interesting hearing. That's been a theme that we've been hearing. Especially the multiple cloud teams and things of that nature. Do you think projects like Backstage and others are going to really help bring that back so it becomes more of that, becomes the interface. And are you guys working on Backstage or with the Backstage group? I've heard a lot. We are not directly working on that but I have heard positive thing. What we have done is with our platform is something very similar. Which is like providing that instant out-of-the-box production ready environment which is multi-tenant. So that management platform is being maintained by platform engineering group but then it has the ability because it's multi-tenant it has the ability to function like a DevOps under the guardrail of what the platform has provided. For example just to give a simple example if an organization wants to provide a set of applications that their developers should be using without having their developers to select those applications then the central platform engineering team can expose those through the platform so that their developers can come in and spend those applications. So like a marketplace that you're providing. Right, exactly. And one of the interesting things is when Savannah would give me crap for not mentioning it we actually showed off your sweatshirt yesterday in our swag segment. It was a really good piece of swag so high marks on the swag for the show. Did we get our sweatshirts? I haven't gone by yet. I'll have to swim by it somewhere. We'll buy a few sweatshirts. It's a great swag. Thank you. Well thanks for coming on. It's going to be an ongoing conversation. We should dig into it when we get back to the US. Platform engineering is moving fast. It's kind of recasting into, I won't say recasting into, I think it's redefining IT certainly has. The Shadow Dev is interesting because it could be an opportunity or it could be problematic. We'll see. I mean I've always felt Shadow IT was, swipe the credit card, get on Amazon. That funded the movement. It definitely did. It's definitely where it started. Deepak, thank you for coming on. Thank you for having me. CTO of D2IQ here on theCUBE, breaking it down. Platform engineering, the new DevOps, the trends continue. Faster, more reliable and actually more portable Kubernetes. We'll be right back. The leader in tech coverage theCUBE, we'll be right back.