 Okay, welcome back everyone to theCUBE's coverage of Red Hat Summit 2021 virtual, I'm John Furrier, host of theCUBE, got two great guests here from AWS, Bob Wise, general manager of Kubernetes for Amazon Web Services, and Peter Ulander, head of product marketing for the enterprise developer and open source at AWS. Gentlemen, you guys are the core leaders in the AWS open source initiatives. Thanks for joining on theCUBE here for Red Hat Summit. Thanks for having us, John, good to be here. So, you know, the innovation that's come from people building on top of the cloud has just been amazing. And you guys props to Amazon Web Services for constantly adding more and raising the bar on more services every year, you guys do that. And now public cloud has become so popular and so important that now hybrid has pushed the edge. You got outposts with Amazon, you see everyone following suit. It's pretty much a clear vote of confidence from the customers that hybrid is the operating model of the future. And that really is about the edge, right? So I want to chat with you about the open source intersection there. So let's get into it. So we're here at Red Hat Summit. So Red Hat's open source company and timing's great for them. Now part of IBM, you guys have had a relationship with Red Hat for some time. Can you tell us about the partnership and how it's working together? Yeah, absolutely. Why don't I take that one? AWS and Red Hat have been strategic partners since, shoot, I think it's 2008 or so. In the early days of AWS, when engaging with customers, we wanted to ensure that AWS was the best place for enterprises to run their Red Hat workloads. And this is super important when you think about, what Red Hat has accomplished with RHEL in the enterprise, it's running SAP, it's running Oracle, it's running all different types of core business applications, as well as a lot of the new things that customers are innovating in. So having that relationship to ensure that not only did it work on AWS, but it actually scaled, we had integration of services, we had the performance, the price, all of the things that were so critical to customers was critical from day one. And we continue to evolve this relationship over time, as you see us coming into Red Hat Summit this year. Well, again, to the hard news here, also the new service, Red Hat OpenShift Service on AWS, known as ROSA, the A for Amazon, Red Hat OpenShift, A for Amazon Web Services, Clever acronym, but really it's on AWS. What exactly is this service, what does it do and who is it designed for? Well, let me jump in on this one. Maybe let's start with the why, you know, why ROSA customers love using OpenShift, but they also want to use AWS, they want the best of both. So they want their peanut butter and their chocolate together in a single confection. And a lot of those customers have deployed AWS, have deployed OpenShift on AWS. They want managed service, simplified supply chain. They want to be able to streamline moving on-premises OpenShift workloads to AWS and naturally want good integration with AWS services. So as to the what, it's our new service jointly operated and supported by Red Hat and AWS to provide a fully managed OpenShift on AWS. So again, like a lot of customers have been running OpenShift on AWS before this time, but of course they were managing it themselves, typically. And so now they get a fully managed option with also simplified supply chain. So single support channel, single billing. You know, we were talking before we came on camera about the acronym on AWS and, you know, people build on the cloud, it's kind of like it's no big deal to say that, but I know it means something. I want to explain, you guys to explain this on, because I know I've been scolded for saying things on theCUBE that were kind of misspoken. It was easy to say, oh yeah, I built that app. We built all the stuff on theCUBE was on AWS, but it's not on AWS. It means something from a designation standpoint. What does on AWS mean? Because this is OpenShift service on AWS. You see other companies have their products on AWS. This is specific designation. Can you share, please? Yep. John, when you see the branding of something like Red Hat on AWS, what that basically signals to our customers is that this is joint engineering work. This is the top of the strategic partners where we actually do a lot of joint engineering and work to make sure that we're driving the right integrations and the right experience. Make sure that these things are accessible and discoverable in our console. They're treated effectively as a first-class service inside of the AWS ecosystem. So there's not many of the ons, if you will. You think about SAP on VMware Cloud on AWS and now Red Hat OpenShift on AWS. It really is that signal that helps give customers the confidence of tested, tried, true, supported and validated service on top of AWS. And we think that's significantly better than anything else. It's easy to run an image on a VM and stuff it into a cloud service to make it available. But customers want better. Customers want tighter experiences. They want to be able to take advantage of all the great things that we have from a scale availability and performance perspective. And that's really what we're pushing towards. And I've seen examples specifically where when partners work with Amazon at that level of joint engineering, deeper partnerships, the results are pretty significant on the business side. So congratulations to you guys working with OpenShift and Red Hat, this real testament to their product. If I got to ask you guys, pull the Amazon Playbook out and challenge you guys or just create some commentary around the process of working backwards. Every time I talk to Andy Jassy, he always says, we work backwards from the customer and we get the requirements and we're listening to customers. Okay, great. He loves that. He loves to say that. It's true. I know that I've seen that at AWS. What is the customer work backwards document look like here? What was the need and what made this become such an important part of AWS? What was the, and then what are they saying now now that the product's out there? Well, OpenShift has a very wide footprint as does AWS and some working backwards documents kind of write themselves because the customer demand is so strong that there's just no avoiding it. And then it really just becomes about making sure you have a good plan. So it becomes much more operational at that point. And Rose is definitely one of those services. We have so much demand and it's as a result, no surprise that we're getting a lot of enthusiasm for customers because so many of them asked us for it. What's the response been so far? What's been the reaction? I'm asking demand, that's, I kind of got the sense of that, but okay, so there's demand now what? What's the use cases? What are customers saying? What's the reaction been? A lot of the use cases are these hybrid kind of use cases where customer has a big OpenShift footprint. What we see from a lot of these customers is a strong demand for consistency. In order to reduce IT sprawl, what they really want to do is have the smallest number of simplest environments they can. And so the customers that standardize an OpenShift really want to be able to standardize OpenShift both in their on-premises environment and on AWS and get managed service options just to remove the undifferentiated heavy lifted. Peter, what's your take on the product marketing side of this where you got OpenSource becoming very enterprise specific and Red Hat's been there for a very long time. I've been a user of Red Hat since the beginning and following them and Linux, obviously is Linux where that's come from, but what features specifically jump out in this offering that customers are resonating around? What's the vibe here? And John, you kind of alluded to it early on, which is I don't know that I'd necessarily call it hybrid, but the reality is our customers have environments that are on-premises in the cloud and all the way out out to the edge. And today, when you think of a lot of solutions and services it's a fractured experience that they have between those three locations. And one of our biggest commitments to our customers is to make things super simple, remove the complexity, do all of the hard work, which means customers are looking for a consistent experience environment and tooling that spans data center to cloud to edge. And that's probably the biggest kind of core asset here for customers who might have standardized on OpenShift in the data center as they come to the cloud, they want to continue to leverage those skills. I think probably one of the, an interesting one as we headed down in this past, we all know Delta Airlines. Delta is a great example of a customer who, joint customer who have been doing stuff inside of AWS for a long time. They've been standardizing on Red Hat for a long time and bringing this together just gave them that simple extension to take their investment in Red Hat OpenShift and leverage their experience and again the scale and performance of what AWS brings them. Next question, what's next for Red Hat OpenShift on AWS in your work with Red Hat? Where does this go next? What's the big to do item? What do you guys see as the vision? I'm glad you mentioned open source collaboration at the start there. One thing to point out is that AWS works on the Kubernetes project upstream as does the Red Hat team. So one of the ways that we collaborate with the Red Hat team is in open source. One of those projects is a new project called ACK, Amazon Controllers for Kubernetes. And this is a kind of Kubernetes friendly way for customers to use an API to manage AWS services. And so that's one of the things that we're looking forward to as that goes GA rolling out into both Rosa and to our other services. Awesome. And I got to ask you guys this while you're here because it's very rare to get two luminaries within AWS on the open source side. This has been a huge build out over the many, many years for AWS. And some people really kind of don't understand kind of the position. So take a minute to clarify the position of AWS on open source. You guys are very active in projects you mentioned upstream with Kubernetes in other areas. I've had many countries with Adrian Cockroft on this as well as others within AWS. Huge proponents, web services. I mean, you go back to the original Amazon. I mean, Jeff Barr was saying 15 years ago some of those APIs are still in play here. APIs back in 15 years ago, that was kind of not mainstream at that time. So you had open standards really made Amazon web services successful and you guys are continuing it. But as the modern era is very enterprise like and you see a lot of legacy, you're seeing a lot more operations that are going to be driven by open technologies that you guys are investing on. Take a minute to explain what AWS is doing and what you guys care about in your mission. Yeah. Why don't I start and then we'll kick it over to Bob because I think Bob can also talk about some of the key contribution sites but the best way to think about it is kind of in three different pillars. So let's start with the first one which is around the fact of ensuring that our customers' favorite open source projects run best on AWS. Since 2006, we've been helping our customers operationalize their open source investments and really kind of achieve that scale and focus more on how they use and innovate on the products versus how they set up and run. And for myself being an open source since the late 90s, the biggest opportunity yet challenge was the access to the technology but it still required you as a customer to learn how to set up, configure, operationalize, support, and sustain. AWS removes that heavy lifting. And again, back to that earlier point from the beginning of AWS, we help customers scale and implement their Apache services, their database services, all of these different types of open source projects to make them really work exceptionally well on AWS and back to that point, make sure that AWS was the best place for their open source projects. I think the second thing that we do and you're seeing that today with what we're doing with Rosa and Red Hat is we partner with open source leaders from Red Hat to Redis and Confluent to a number of different players out there, Rafauna and Prometheus to even foundations like the LF and the CNCF. We partner with these leaders to ensure that we're working together to grow the overall experience and the overall pie, if you will. And this kind of gets into that point you were making, John, in that the old world legacy proprietary stuff, there's a huge chance for refresh and new opportunity and rethinking or modernization, if you will, as you come into the cloud, having the expertise and the partnerships with these key players as enterprises move in is so crucial. And then the third piece I like to talk about that's important to our open source strategies is really around contribution. We have a number of projects that we've delivered ourselves. I think the two most recent ones that really come top of mind for me is what we did with Babelfish as well as with OpenSearch, right? So contributing and driving a true open source project that helps our customers take advantage of things like an SQL, a proprietary to open source SQL, a SQL conversion tool, or what we're doing to make Elasticsearch, the opportunity or the primary open platform for our customers. But it's not just about those services, it's also collaborating with key industry initiatives. And Bob's at the forefront of that with what we're doing with the CNCF around things like Kubernetes and Prometheus, et cetera. Bob, you want to jump in on some of that? Sure, I think the one thing I would add here is that customers love using these open source projects, but one of the challenges with them frequently is security. And this is job zero at AWS. So a lot of the collaboration work we do, a lot of the work that we do on the upstream projects is kind of specifically around kind of security oriented things because that is what customers expect when they come to get a managed service at AWS. So some of those efforts are somewhat unsung because you generally do more work and less talk in security oriented things. But in projects across AWS, that's always a key contribution focus for us. That's a good way to call out security too. I think that's being built into everything now. That's an operating model. People call it shift left, day two operations, however you want to look at it, you got this nice formation going between under the hood kind of programmability of the infrastructure at scale. And then you have the modern application development, which is just beginning to programmable DevSecOps. It's funny, Bob, I'd love to get your take on this because I remember in the 80s and during the Unix generation, I used to pedal software into the table like, here's a copy of Unix, don't tell anyone. People in the younger generation don't get the fact that it wasn't always open, okay? And so now you have open and you have this idea of an enterprise that's going to be a system management system view. So it's, you got engineering and you got computer science kind of coming together. This SRE middle layer, you're hearing that as kind of a new discipline. So DevOps kind of has won. I mean, kind of knew this for many, many years. I said this in 2013 on theCUBE actually, that reinventing, I just recently shared that clip. But okay, now you got SecOps, DevSecOps. So now you have an era where it's a system thinking and open source is driving all that. So can you share your perspective because this is kind of where the puck is going. It's an open, open to open world. It's going to have to be open and scalable. How does open source and you guys take it to the next level to give that same scale and reliability? What's your vision vision? Yeah, the key here is really around automation and what we're seeing, you can look at Kubernetes. Kubernetes is essentially a robot. It was like the early design of it was built around robotics principles. So it's a giant software robot and the world has changed. If you just look at the influx of all kinds of automation to not just the DevOps world, but to all industries, we see a similar kind of trend. And so the world of a IT operations person is changing from doing the work that the robot did and replacing with the robot to managing large numbers of robots. And in this case, the robots are like a little early and a little hard to talk to. And so you end up using languages like YAML and other things, but it turns out robots still just do what you tell them to do. And so one of the things you have to do is be really, really careful because robots will go and do whatever it is you ask them to do. On the other hand, they're really, really good at doing that. So in the security area, I think the research points to the largest single source of security issues being people making manual mistakes. And a lot of people are still a little bit terrified if human beings aren't touching things on the way to production. AWS, we're terrified if humans aren't touching. And that is a super hard chasm to cross. And open source projects are really playing a big role in what's really an IT wide migration to a whole new set of not just tools but organizational approaches. Peter, what's your reaction to that? Because we're talking essentially software concepts because if you write bad code, the code will execute what you did. So assuming it compiles like the old days. Now, if you're going to scale large scale operations that has dynamic capabilities, services being initiated and terminating, tear down, up, started, you need the automation but if you really don't design it right, you could be screwed. This is a huge deal. This is one reason why we've put so much effort into a GitOps that's, you can think of it as a more narrowly defined subset of the DevOps world with a specific set of principles around using kind of simplified declarative approaches along with robots that converge the desired state, converge the system to the desired state. And when you get into large distributed systems, you end up needing to take those kinds of approaches to get it to work at scale. Otherwise you have problems. Yeah, just adding to that. And it's funny you said, DevOps is one. I actually think DevOps is one, but DevOps hasn't changed as the cloud move, right? You know, the reality is it was founded back what quite a while ago, it was more around CICD and the enterprise and the closed data center. And it was one of those where automation and run books addressed the fact that every pair of hands between service requests and service delivery created an issue. So that growth and that mental model of moving from waterfall agile to DevOps, you built it, you run it type of a model, I think is really, really important. But as it comes out into the cloud, you no longer have those controls of the data center and you actually have infinite scale. So back to your point of you got to get this right. You have to architect correctly. You have to make sure that your code is good. You have to make sure that you have full visibility. This is where it gets really interesting at AWS and some of the things that we're tying in. So whether we're talking about GitOps, like what Bob just went through or what you brought up with DevSecOps, you also have things like AIOps, right? And so looking at how we take our machine learning tools to really implement the appropriate types of code reviews to assessing your infrastructure or your choices against well-architected principles and providing automated remediation is key. Adding to that is observability. Developers, especially in a highly distributed environment need to have better understanding, fidelity and touch points of what's going on with their application as it runs in production. And so what we do with regards to the work we have in observability around the Grafana and Prometheus projects only accelerate that whole concept of continuous monitoring and continuous observability. And then kind of really adding to that, I think it was last month we introduced our fault injection simulator, a chaos engineering tool that again, takes advantage of all of this automation and machine learning to really help our developers, our customers operate at scale, right? And make sure that when they are releasing code, that is not just great in a small sense, it works on my laptop, but it works great in a highly distributed, massively scaled environment around the globe. You know, this is one of the things that impresses me about Red Hat this year. And I've said this before on other covers, events I've covered with them, is that they get the cloud scale piece. And I think their relationship with you guys shows that. I think DevOps is one, but it's the gift that keeps giving in open source because what you have here is no longer a conversation about the cloud, moving to the cloud, it's the cloud has become the operating model so the conversation shifts to much more complicated enterprise or and or intelligent edge, and whether it's industrial or human or whatever, you know, you've got a data problem. So that's, it's about a programmability issue at scale. So, you know, it's interesting is Red Hat is on this bandwagon, it's an operating system. I mean, basically it's a distributed computing paradigm, and essentially a lot AWS concept as a cloud now goes to the edge, it's just distributed services via an open source. So what's your reaction to that? Yeah, it's back to the original point, John, where I said, you know, any CIO is thinking about their IT environment from data center to cloud to edge, and the more consistency automation and you know, kind of tools at their disposal to enable them to create that kind of, you know, I think you started to talk about it infrastructure, the whole as code, infrastructure is code. It's now almost everything is code. And you know, that starts with the operating system, obviously, and that's why this is so critical that we're partnering with companies like Red Hat on our vision and their vision because they aligned to where our customers are ultimately going. Bob, you want to add to that? No, I think you said it. I think you said it all well. You guys are crushing it. Bob, one quick question before we go, I got you here. You mentioned GitOps, I've heard this before, I kind of understand, can you just quickly define from your perspective, what is GitOps? Sure, well, GitOps is really taking the, as I said before, it's a kind of narrowed version of DevOps, sure, it's infrastructure as code. You know, sure you're doing things incrementally, but the GitOps principle, it's back to like, what are the good, what are the best practices when you're managing large numbers of robots? And in this case, it's around this idea of declarative intent. So instead of having systems that reach into production and change things, what you do is you set up the defined declared state of the system that you want and then leave the robots to constantly work to converge the state there. That seems kind of nebulous. Let me give you like a really concrete example from Kubernetes. The way the entire Kubernetes system design is based on this. You say, I want five pods running in production and that's running my application. So what Kubernetes does is it sits there and it constantly checks, oh, I'm supposed to have five pods. Do I have five pods? Do I have five pods? Well, what happens if the machine running one of those pods goes away? Now suddenly it goes and checks and says, oh, I'm supposed to have five pods, but there's four pods. What action do I take to now try to get the system back to the state? So you don't have a system reaching out and checking externally to Kubernetes. You let Kubernetes do the heavy lifting there. And so it goes through a loop of, oh, I need to start a new pod and then it converges the system state back to running five pods. So it's really taking that kind of declarative intent combined with constant convergence loops to fully production at scale. That's awesome. Well, we do a whole segment in state and stateless future, but we don't have time. But I do want to summarize real quick. We're here at the Red Hat Summit 2021. You got Red Hat OpenShift on AWS, the big news. Bob and Peter, tell us quickly and summary why AWS, why Red Hat, why better together? Give the quick overview. Bob, we'll start with you. Bob, you want to kick us off? I'm going to repeat peanut butter and chocolate. Customers love OpenShift. They love managed services. They want simplified operations, simplified supply chain. So you get the best of both worlds. You get the OpenShift that you want fully managed on AWS where you get all of the security and scale. Yeah, I can't add much to that other than saying, Red Hat is a powerhouse obviously on data centers. It is the operating system of the data center and bringing together the best in the cloud with the best in the data center is such a huge benefit to our customers. Because back to your point, John, our customers are thinking about what are they doing from data center to cloud to edge and bringing the best of those pieces together in a seamless solution is so critical. And that's why AWS. Guys, great, thanks for coming on. I really appreciate it. I just want to give you guys a plug for you and being humble, but working the CNCF and the standards bodies has been well known and getting the word out. Congratulations for the commitment to open source. Really appreciate the community. Thanks you, thank you for your time. Thanks, John. Thank you. Okay, CUBE coverage here covering Red Hat Summit 2021. I'm John Furrier, host of theCUBE. Thanks for watching.