 Hi, this is Yossi Kulbhartiya and welcome to a brand new episode of T3M, our topic of this month. And the topic of this month is open source. And today we have with us, Satya Sankaran, Chief Operating Officer at Cloud Casa by Katalajik to talk about this topic. Satya, it's great to have you back on the show. Satya, it's a pleasure being here always. And I was just going to say, what a timely topic. What a fantastic topic you're running this month. Open source is just everywhere. And if people didn't know, I think OpenAI and ChatGPD is just kind of making this household topic these days, right? So wonderful to be a part of this discussion. What are the forces that are driving the adoption of open source? I mean, if you look, you know, in the cloud, open source has kind of become defective. We were at KubeCon a few weeks ago and all you see there is open source. So talk about what is the driver, what is the, what is driving this adoption of open source cloud? Yeah, I would say two main things, Satya. What's driving adoption of open source is really one, trust. We human beings make a lot of decisions based on trust. And a lot of times we even supersede other factors with what we trust the most. And what I think open source is completely cracked open is that it can get people's trust. Source code is visible for everybody to use, you know, no government is out there saying you cannot use open source products, right? Everybody trusts it because they can see it operates in transparency. And that is a primary driving factor. And then comes community. You could trust something, but unless it's popular, people aren't going to use, right? And this goes hand in hand when more and more people using open source, it becomes just a de facto standard that people are going to use, because everybody else is using it and there is a significant amount of trust. And if you don't trust, you can look into it and see what it does and what it doesn't do. And that drives a lot of the adoption I think we see today. When we look at open source, we talked about Kubernetes ecosystem, just because it is open source doesn't mean the company should just lean towards that, you know, open source. When it comes to the open source market, it could be fully, you know, market where you can download things for free from GitHub or GitLab or you can go with the vendor. You still have to do the same kind of review as you would do with any proprietary software. So talk a bit about when enterprises, they do either make a switch or, you know, start something new and they are looking at open source solutions. What are some of the criteria they should look at before they make a decision? Yeah. Great question. I mean, in a recent webinar, we ran some polls so up there, we asked people, hey, what is your preference? Open source or closed source commercial solutions? And of course, majority in a lion's share of people on the webinar picked open source. That tells you that's the preferred model if you can find a solution in that model. Right. Second question we asked, we kind of tried to dig a little deeper. We asked them, do you want open source free and completely free solutions? Do you want open source with paid support? And would you go with, you know, commercial where you have an existing relationship and so on? And you suddenly saw a lion's share of the same users pick open source with paid support is the majority preference because, look, not everybody is going to know, even if you operate in complete transparency, you still need skill sets to understand what an open source solution does. And what you need, how you need to configure it to make it work for your environment. And you're not going to have users that are an expert at 100 open source technologies. That's just not a plausible scenario. You get to, if you are one of those users that gets to focus on one open source project and that's what you need to build expertise on, great for you, good for you. But if you are like the majority of IT infrastructure folks or developers that are using a combination of 10, 15 open source projects to make something work, then you're not going to be an expert at everything. And that you need the right level of expertise and support to make sure that open source actually works for you. There's no piece of software as a return to work for everybody one way. You're going to have to be able to configure it, you're going to be able to troubleshoot it. I mean, you're going to have to have all of these capabilities. And I think that's one of the models that you work on top of open source to deliver for users. It is a win-win. I mean, of course, everything could be free, but somebody's got to pay these developers and their work and effort at some point. And what you will see is that with open source, it creates the win-win approach where, yes, you are essentially getting a large portion of the technology for free, but you're only paying for some support and services on top of it. Or if you're running it as a managed service, you're getting paying for some of that convenience. So I think that creates the win-win model. People are looking for, at the end of it, you end up paying less for software and technology and you get a more trusted, more widely adopted technology for your use. And that's win-win for a customer perspective. I've been covering open source for a very long time and what I also see is that one of the most important, not the most, but one of the important kind of constituents of open source is also players who are offering commercial solutions based on that open source project. Like we talked about Veloro, where you folks, you know, bloggers for Veloro was announced at KubeCon. So talk a bit about the importance of this whole ecosystem because without this ecosystem, I mean, I can go back to the farming, growing potato is easier, but without the whole ecosystem of truckers and restaurants and it's used, I love that analogy. So talk a bit about the rule of commercial aspect of open source. Before we go into that, let me explain at least how, this is not my idea, but it's something that stuck with me the first time I read this, right? How do you kind of care, you know, what are the various open source topologies that exist? I think it's important to understand that before we talk about, you know, what model works best on each of these topologies, right? This is based on a book by Nadia Iqbal, I think is her name and a book called Working in Public, right? And she characterizes, you know, these open source projects into four types, really two axes, contributor growth and user growth, right? And if you have, you know, low contributor growth, low user growth, that basically becomes a pet project, right? You have your own project and you're still building something on your own and you're open sourcing it. Nothing really happens other than, you know, you are building this in public. Right now, I think pet projects are most useful for finding your next gig. If you want to find the next job, you know, pet projects are amazing. We do, every person that is hiring these days looks at your GitHub profile and these pet projects can be a huge way to advertise, you know, your skillsets, primarily, to do more eventually for a company, right? Then there is the other side, which is that there is low contributor growth and these are effectively clubs. There is low user growth, but a high contributor growth. So that means not a lot of users, but everybody that is contributing to this project actually becomes effectively a user, right? So this, I think Nadia calls them clubs, right? And these are both models that aren't really great at commercializing, right? One is a toy, great for hiring, the other is a club. You may have a pet project idea or you may be into, you know, some automation of, you know, something in your house that only a few people have. It could be a game that, you know, one of some of your few friends play. You know, those are things that are, you know, primarily good for these types of open source projects. And then the two projects, two types of projects, which is really around, you know, high user growth category, right? And high user growth, but low contributors. Not many contributors, but lots and lots of users, right? That's really what Nadia calls them stadium projects. And why is it called stadium projects? Because you are a football fan, you're a Manchester United fan, for example. You know, you've got a small group of players playing and then there are millions of people watching you, right? And it's stadium because it works very much like that sports team stadium model where, you know, there is a small group of players doing what they do best. And then there are lots of users, lots of people viewing and getting utility out of those. These are great projects to deliver, you know, support for, right? Because, again, only a few people know what they're doing. And if you want the rest of them to be able to pick up, there are a lot of support ecosystems that goes around it and so on. And then there is the fourth model, high contributor growth, high user growth. You know, Linux is a great example. Kubernetes is a great example, you know, so many contributors and so many users. When you see both, it effectively creates its own ecosystem where, you know, multiple vendors come in to, you know, plug many gaps that are there in the ecosystem. So you really have to look at, you know, what open source project that you're trying to add value to, where it fits in this kind of topology. And what we've picked up and tried to add value to is Velero, which kind of fits that stadium project. It's quite not an ecosystem in itself yet, but it's got an enormous amount of users. And it is focused on a very specific use case with, you know, about 200 plus contributors, you know, creating that open source project. And that's where we're trying to add value. We're trying to add value through two ways. One is to provide support for users that are using Velero. And two is to provide that as a managed service so that you essentially don't have to manage a backup infrastructure. If you're not going to manage your own primary infrastructure and putting it all on cloud, you know, what are the odds that you don't want to do manage your own backups, right? That's really the thought process behind these involvement of us in open source. I also want to talk a bit about the role of foundations because if you look at open source, there is code which is, foundation there is code which is owned by certain companies. Both are right way to do open source, but sometimes when the code is owned by companies, sometimes their competitors are worried that, hey, they may get locked out, the license may get changed, or the company might have more influence over the code. So they feel more comfortable leveraging code base which is available through some foundations. And there are some great foundations like Apache is there, Linux foundation is there, Open Infra foundation is there. And these foundations have helped accelerate the adoption of open source. So I want to hear from you to tell our viewers the role that you have seen these foundations have played in making open source, you know, more popular and it's getting adopted across the board. Yeah, it goes back to trust, right? And when you know it's owned by an open source foundation, you're going to know it's going to remain open source. You know it's going to be committed to community use of that project. And of course it comes with a level of sponsorship to make these things work, right? But that need not be the only way you make it work. I'll go back to the example of, you know, Valero and how beautifully it works. Valero is not a CNCF project. Valero is owned originally by Heptio which was acquired by VMware and VMware has continued investing in this project. But VMware has stayed true to the open source ethos with respect to Valero. And why we carry so much trust in this ecosystem is because they've openly invited other maintainers into this ecosystem, right? So you now have not just maintainers from VMware but also maintainers from Microsoft, maintainers from Red Hat. Red Hat actually now has most number of maintainers outside of VMware themselves. It's a large team that is building this project up. And I think we've also seen Suza play a role in this. We've seen Dell play a role in this. You know, we're contributing heavily to Valero these days. And so all of this kind of plays a role in that you are kind of operating it with the right ethos and now that multiple maintainers exist and a large contributor base exist, even if you changed the license model of this project down the road, it is now becomes really easy for us to fork this and basically run with the rest of the maintainers, this project set. So it creates that barrier to go change something completely. So foundations are great because it allows you, gives you a certain level of guarantee that this is how the project is going to be. But apart from foundation, the next best model is really shared development. If one company is doing all the work behind open source, you know, those companies don't exist for public benefit unless, you know, they're registered as one. Their job is to monetize it at some point and it comes from adoption many times. But in a model like Valero, you have multiple maintainers playing a role in lots of people contributing and that essentially future proves how open source gets used. And the other thing I wanna highlight is a hundred million Docker Pulse just this week, and it happens so nonchalantly. Can you think of a software or a technology that's been pulled a hundred million times and nobody notices, right? Not VMware, not the maintainers. And I happened to be doing an interview with Block and Files that morning and I just went in to gather proof for some of the things that I'm saying. And I opened this and I just saw that it ticked up to a hundred million Docker Pulse. Again, that comes from a large number of people putting their trust, especially in a mission critical workload like Valero. And that's only possible because, you know, so many are contributing and so many are using. And that's the way forward in this ecosystem. If we look at cloud costs or if you look at, you know, cataloging, what will you see that companies like yours play in this whole larger open source ecosystem? I think there are very, in an open source ecosystem, there is contributions and there is, you know, supporting, right? How can you contribute to an open source community? Of course you can code, but coding isn't the only way to contribute to an open source community. And in fact, a lot of companies go through stages where, you know, they make lots of bug fixes first before they get into core part of the projects and contribute to core part of the projects. So you can do a lot of bug fixes. You can own the core. You can add feature sets. I mean, these are all kind of falling into, hey, I'm going to write code to support this open source community. The second type of community, especially in a high user growth project you're talking about, is there lots of users picking up this new project, right? And often there is a Slack channel that you have to maintain. There is a forum you have to support. You know, somebody still needs to hand hold these people into using the solution right, using the technology right. You know, we do a lot of that today with Valero. In fact, if you went to the Slack channel, you know, it does over 3,200 users I think at this point and you will see in the last few months, you know, practically our team answers every question that comes through in that forum and Slack channel. Again, it's all part of getting people to successfully adopt this open source technology and kind of only when they adopt they become part of the community, right? You kind of have to be a welcoming member of that community. There is one way to contribute and we do a lot of that today because again, we built a three year expertise on Valero today and we do that. In terms of commercial support on top of just, you know, contributing to the community, of course, I think one of the things and a knock on open source is that open source gets a lot of things right, but there are a lot of things that are objective or gotten right almost immediately because again, you get the wisdom of the crowd, but a lot of things that are subjective, a crowd doesn't help you, right? You don't get to how many open source projects have really good UI is an example, right? And that's because UI is very subjective. No one person, you know, likes it, you know, the same way and there is no real configurable UIs, you know, in this ecosystem. So what this unfortunately does is it raises the bar for skill set to adopt certain open source technologies because there is no democratization of those feature sets. There is no UI, there's nobody can come in just click through a bunch of buttons without having to, you know, loan YAML files and so on. So what we are also doing here is building a service that allows people to kind of manage this project open source of Valero across multiple clusters from a centralized UI. So it's a lot about making Valero more accessible to users so that they can run Valero for mission critical workloads because when your servers are done, when your clusters are down, you don't wanna be reading through 400 pages of manual to figure out how to restore your clusters. You wanna go to an interface from where you can pick a recovery point, you can ask it to do what it needs to be done. A guided recovery path becomes very important and again, that's something we do. And the last piece is really commercial support. A community support is wonderful when you're getting onboarded but when you're running your mission critical workloads on top of it, do you wanna rely on someone on the other side of the world answering your slack questions? Or would you want to be in a position where you wanna pick up the phone call and call a specific expert and say, hey, I'm running into this problem, I need this problem resolved and have a certain level of SLA and SNO that you are running into. And that's something again, companies like us can contribute to. Satya, thank you so much for joining me today and talk about open source. It was an interesting discussion and as usual, I would love to have you back on the show. Thank you. Awesome, thanks Supna.