 Hello, welcome back to theCUBE's coverage of AWS re-invent 2021. I'm John Furrier, your host of theCUBE. We're here Michael Danielo, platform architect at VMware's Carbon Black. Michael, great to see you. We're here at re-invent virtual hybrid in-person. Great to have you on theCUBE. Thanks for coming on. Yeah, thanks a lot, glad to be here. So one of the big stories that we're tracking obviously is workloads, all cloud for all workloads. Obviously the data is a big part of things. But under the covers and optimizing cloud for the application developers, this modern application movement is more and more at the top of the stack. People just wanted to code, infrastructure is code. You're seeing DevSecOps as a big trend that's driving all new microservices, all new greatness for developers. But still there's an optimization question. I want to get your thoughts on this is what you do. Take a minute to explain what your role is at Carbon Black around this cloud optimization. Yeah, absolutely. Yeah, so my name is Michael Danielo. I'm a platform architect at VMware Carbon Black and work across all the different engineering teams and our main objective is to develop scalable platform tools and that includes yeah, cloud security automation pieces, pipelines, cost optimization, like we'll be talking about today, developer enablement tooling and observability tooling. One of the big things about instances is that, do I have enough instances? Obviously Elastic Cloud is amazing, all kinds of new resources there. But talk about the AMD portion of the instances. How do we identify these instances? How do developers understand it? What's in them and what's the selection criteria? Take us through that whole process of the Amazon Web Services and the AMD instances. Yeah, sure. So essentially we're leveraging a lot of our instances to run our EKS clusters, which is a managed service from AWS to run our Kubernetes clusters. And we identified that we could take a bunch of those instances and gain some cost optimization benefits by selecting from Intel to AMD processors. And initially we had measured out to be roughly a 10% reduction in cost just for selecting that instance type. But yeah, we actually learned, we gained quite a bit more. Developers are always like, I want more power. And this is what the whole idea of cloud is. Cloud scale has been a big competitive advantage. But also there's a cost aspect of it. What's the balance between maximizing performance and cost optimization? Because now people don't want to, they want more power. They also don't want to have a lot of extra spend. This is kind of one of those things they talk about in cloud where it's been so successful, cost is important. Yeah, yeah, for sure. And it's got to be easy to get that cost optimization benefit. Otherwise you're spending all your cycles and burning that money there in the human capital and in the team, in the engineering effort. So luckily this change is a one-line change. We use Terraform for our automated provisioning layer and we were able to make that one-line change. And then developers didn't have to make any application changes, which was great. So it was a no-brainer for us to pursue this. Talk about the EC2 instances that leverage AMD-based processor, the EK, as you mentioned that earlier. What is that all about? What's the benefits? What's in it for you guys? Yeah, for sure. So essentially the workloads that are running on these instance types are our actual carbon black cloud application. So all the backend systems that support our customers. And so in that use case, we're spinning up all of our containers that are running our applications and essentially that's our use case for those instance types. How did you come to use the AWS EC2 instances on AMD? Did you have an evaluation process? Did you just go select it? I mean, take us through that migration aspect of it. Yeah, sure, yeah. So originally we're looking across the board. How can we do better cost optimization, right? And that goes across every different AWS resource, but we targeted this one specifically. We worked alongside with their AWS TAMs and representatives to basically find out, hey, is this financially worth the effort? And we did reach that conclusion with some analysis, basically targeting these instance types and doing some analysis on that cost optimization, optimization specifically, and ended up being the right thing to target. What was the ease of use of the switch? Take us through that. Was it the heavy lift? Was it seamless? Take us through the impact there on the moveover and what was the results of that? Yeah, so that, I mean, that's the greatest thing. Like I said before, I mean, we had to make just a single line change just to change that instance type in our config and then roll that out across our regions. We did slow roll that in order to make sure that those changes in our development environments didn't make any performance hits or we didn't run into any snags with the applications themselves. But yeah, that's the greatest part about the story from my perspective is the ease to migrate over and to switch to these instance types. And then you just immediately gain that cost optimization benefit. You know, I love about what your job is, platform architect, kind of that word, kind of had a lot of meeting even 10, 15 years ago, but now at the cloud, it's almost like you're always kind of finagling and managing and massaging and nurturing the infrastructure to enable more new things are coming online as well, more high level services. So you got a fun job, but it's always evolving. How do you stay on top of it? What's the impact been for your customers too? As you start deploying some of these new instance capabilities, take us through kind of the day in the life of what you do and then what's the impact of customers? Yeah, sure. So, you know, like you said, there's quite a bit now to look at, you know, you got to stay on top of different blogs and keep connected with your network to see what your other colleagues are doing across different companies. You know, you go into conferences like AWS re-invent, right? To keep on the cutting edge here. But yeah, I mean, that's essentially, you know, one of the key aspects is just trying to look at all the different aspects, all the new technologies that are coming out, making sure you're making the right choices there and trying to get the most bang through buck while you're at it. What are some of the big factors that you see in cloud native as you start to look at what customers are doing? Obviously with Kubernetes, you're starting to see that platform develop inside the industry, as well as de facto kind of orchestration layer. But now as customers start to look at it, they want to have more ease of use there too. At the same time, they don't want to have to do a lot of front end work. They want to get instant benefits in the cloud. Obviously, whether it's from a security standpoint or just rolling out a modern application. Okay, so as having all this infrastructure under the covers, how do you look at that problem and how do you solve that or capture that opportunity? Yeah, and I think that's why we're seeing a movement here on platform teams. It's kind of a newer terminology, usually a band of developers and SREs come together and say, well, we've got a lot of different things to look at. We're onboarding applications to Kubernetes and we need to make tools so that developers don't have to think much about the transition and the underlying platform. And so that's one of our success metrics on the platform engineering teams just to almost be non-existent, to just have everything flow through our systems and then have just a high ease of use to onboard the applications to the new platform. You know, it looks like you have some great success with the AMD-based instances. Can I ask you a question because I wanted to figure this out. How do you identify an AMD-based instance when you're making the selections? Yeah, sure, it's as easy as just the A after the name. So for us, it was the C5.4 XL. And if you want the AMD one, it's just the C5A.4 XL. So I guess technically I said it's a one-line change. It's actually a one-letter change. So quite easy there. Yeah, it's almost like back in the old glory days of command line, one quick update. The customer aspect of this is also important to want if you don't mind while I got you here. What are some of the things that you're hearing from your customers from performance standpoint that they're looking for? Obviously the cost optimization is key, but as they look to deploy more power and more performance, what are some of the things that your customers are looking for from Carbon Black? Yeah, so I mean, we are a security company, but we're really a data company because we have 8,000 customers. We process over a trillion events per day. We ingress over a hundred terabytes of data per day. And so our customers need high level of performance. And if we can't provide that with low latency, we're not successful. So that's why performance on the underlying systems that are running our applications is super critical. Yeah, you're like a trailblazer over there. I mean, the work that you guys are doing with the data is amazing. And that's a big theme at Reinvent this year is that data is a huge part. We look at the success of the cloud growth on this, I call Gen2 cloud happening. This whole modern movement is all about how people handle the data at scale because cloud scales here and now you got processing all that data. The trailblazing that's going on, there's this new wave of kind of, I was called first generation trailblazers, but you guys are doing that. What's advice would you have for other architects out there and kind of the mainstream enterprises who were like, I want to take advantage of the path that you guys have plowed through. What's your advice? Yeah, I think one of the key things in a place where we've had a lot of success is creating standards, making sure that we're choosing technology wisely and making sure that your company isn't building the same solution in silos. And that's a huge pattern that I've seen in my career. And if you can negate that, you're going to be in a great place. So choose the right technology, container first, cloud native first, push forward and then make sure that everybody's kind of on that same ship running in the same direction. Well, great case study on this AMD based instance migration. Was there any uplift in experience that you've seen on the switch and the performance cycle? Can you just talk about that? What does it mean to upgrade? What benefits are you seeing on the performance? Yeah, so I didn't kind of hit on this yet and I really wanted to. So we, yeah, so upfront, the instance itself is 10% cheaper. However, we found out that we had to run far less instances because of that performance increase. So we ended up saving roughly 30%. And we continue to scale out. So at first it was a couple hundred instances, now we're in the thousands and we're gonna keep ramping up to over 10,000s, tens of thousands. Let me get this right. So it's single line change, letter change, instance change. So you get not as many instances and you save money. So you get cost optimization and higher performance. Yep, they say if it's too good to be true and it's not, in this case it actually is. So why is it so good in your opinion? What did you discover? What was the big revelation that went down this path? That's good value proposition. Yeah, for sure. I mean, so initially we were just chasing that initial BC2, 10%. And then as we kind of push it forward, we're looking at the metrics at month-to-month costs. And we're actually saying, well, as we kind of swap over from one instance type to another, we're actually paying less. And then once we fully swapped over, it took five or six months to get to the same amount of cost as we continued to scale upward. So it's been a great story. It is a great story. It's a super nuance, but it's super important to know these platform benefits. I got to ask you on a personal question, if you don't mind. We love covering cloud. We've been covering Amazon for a 90 year at Reinvent. Just love covering all the action and tech as this just total awesomeness environment. Cloud scale, innovation, capabilities. It's like surfing a big wave. But there's a bigger wave coming and we're seeing it now. I want to get your thoughts on this. As you look to the next big wave, which beyond cloud now, cloud scale, data, new architecture, new architect is rolling out with Edge, basically distributed computing at large scale, and tons of security challenges, right? How do you look at this next big wave coming? You stare at it and say, wow, this is going to be huge. And how do you ride that wave? What's your mindset and how do you look at that? Well, first of all, I'm just extremely excited about it. The further this thing grows out, there's definitely more complexity, but just a whole slew of fun problems to solve. But when we look at these different problems and solving them at scale across multiple regions, it gets pretty exciting, right? So I can say one example of this is our security of our cloud, not the security product. And we've developed automation for prevention and auto remediation in our pipelines. It's been such a success story. And these type of technologies did not exist even a couple of years ago. And we've been able to take advantage of them. So there's going to be a lot more of that, where that came from. So yeah. Michael, great work. And again, you're truly a trailblazer. And this is a great, you got to do it. You got to screw your own cloud and stay on the cutting edge and ride that wave. Congratulations on the cost up cloud optimization and the success with AMD based instances. Congratulations. Thanks. Thanks. Okay, this is theCUBE's coverage of Avis ReInvent 2021. I'm John Furrier, your host of theCUBE. Thanks for watching.