DevOps Chats: Kubernetes is the New Compute, with Rancher Labs’ Sheng Liang

Transcript

Alan Shimel: Hi everyone, this is Alan Shimel for DevOps.com and you’re listening to another DevOps Chat. I’m really happy to be joined by a good friend of mine who I’ve known probably more years than we want to admit to, my friend Sheng Liang who is the CEO of Rancher Labs. Sheng, welcome.

Sheng Liang: Thank you very much. I’m glad to be here.

Shimel: I’m glad to have you here, my friend. So, Sheng, the news with Rancher, I guess we call it a momentum release, right, you recently put out.

Liang: Yeah.

Shimel: Why don’t you just share some of the highlights with the audience?

Liang: Yeah. Every January, early February, we tend to look back the year before and see how we’ve done, how the whole industry has done, and it’s just been remarkable. I guess the top-level thing is our business, our recurring revenue grew 169%, so that’s like high three-digit growth. It’s way exceeded our expectation. That itself is good news for us, but it’s not really–I don’t think an average person in the industry needs to care about it as much.

What I want to, Alan, spend some more time talking to you is some of the underlying technology changes and industry movement that’s kind of enabling us to thrive in this environment.

Shimel: Absolutely. Let’s be fair, congratulations. A hundred and sixty nine percent year-over-year growth is nothing to sneeze at. You’re right, there is a little bit I think of a rising tide lifts all boats when we’re talking about the Kubernetes market. Nevertheless, 169% is still 169%. You and I both come from that startup world and congratulations.

Liang: Thank you.

Shimel: It’s something to be proud of.

Liang: Yeah.

Shimel: Also, for our audience out there who are not familiar with Rancher Labs and Rancher OS, it’s a great startup story. Sheng, your background, cloud.com, all the things you’ve been involved in, you’re the perfect person to have led this. But Rancher Labs came out there originally with the Container Revolution, right?

Liang: Yes.

Shimel: With an OS for running containers. You were agile, nimble, adaptable enough to see Kubernetes and say, “Hey, we’re not gonna–we’re shoveling sand against the tide, right, if we think we’re gonna beat Kubernetes. If you can’t beat them, join them.” You incorporated Kubernetes into the Rancher Lab system.

Liang: Yes, and in a fairly unique way as well. I’d love to share the story a little bit if you don’t mind.

Shimel: No, you’ll tell it better than I will.

Liang: We’re just over five years old, over 200 people now. We operate in 14 countries and we’re open source business. It’s 100% open source. Every single product we ship is available free in source code form, Apache open source, and it really worked well for us. I know in the industry there’s actually been a lot of talks about “Is open source even a great business anymore in the context of cloud?” But we continue to believe so, at least especially in the industry we’re in, and we can talk a little bit more about that.

But let’s talk about the business and market and Kubernetes. The great opportunity we see with Kubernetes is Kubernetes for once has become an industry standard for compute. Now, think about it. We all know compute these days is everywhere. Computing is on cars, in ATM machines, on your phone, certainly on the server, in the cloud, but really, before Kubernetes where the world was getting into was quite challenging for a lot of organizations.

Because these clouds as good as they were, as good as they still are, they’re just different. If I’m a CIO, say if I wanna be efficient, I wanna maybe go all in on one cloud whether it’s Azure or Amazon or Google, but until the day comes, now I merge with another company and now bringing a whole different set of infrastructures, and I don’t really wanna worry about that because my whole point is develop the apps, transform the business, help the–all the stuff I’m doing to worry about infrastructure is just plumbing, and that’s the true power of Kubernetes.

I don’t think enough people in our industry are promoting Kubernetes this way yet, so I’d like to make a strong point with you, that in my mind, Kubernetes is the TCPIP of compute. Now, think about it.

Shimel: It’s the stack. It’s the [Crosstalk] as you say. It’s the protocol.

Liang: It’s a stack and a protocol because before TCPIP if you and I could remember back in–

Shimel: I do remember.

Liang: There were a handful, almost a dozen different network protocols, and after TCPIP we had 3Com, we have Cisco, we have Yahoo!, we have Google. All that stuff happened, internet. Revolution happened because finally–

Shimel: I remember when you had to add the TCPIP stack separately. It didn’t come with Windows, so you had to add that in. That was an add-on. Early open source stuff.

Liang: That’s exactly what’s happening because–this is a great point you brought up. Someday people will say, “Oh, I remember there was a time I had to talk to vendors like Red Hat and Rancher just to implement Kubernetes, whereas today if I get infrastructure like from Amazon, Google, from VMWare, from Dell, wherever, it comes with Kubernetes.” That’s exactly the same thing that’s happening. You see that, right?

Shimel: Yeah. History repeats itself. When you get old enough, Sheng, you realize that. Let’s talk something else. I wanted to talk a little bit about CNCF because that’s their mission to make sure that Kubernetes is part of the plumbing. Truthfully, it’s not just Kubernetes. When you look at CNCF’s mission, they truly are trying to develop a stack or at least help foster the development of a stack built on several open source projects.

Beyond Kubernetes, for Rancher now, what are some of the other CNCF projects that you see making their way into Rancher either now or in the future?

Liang: I would say pretty much the whole stack. We’re very deeply involved in CNCF. We’re very early member of CNCF, and my co-founder Shannon Williams is elected to the CNCF governing board. Just this week I was elected to the CNCF TLC.

Shimel: Really? Congratulations.

Liang: So, I have some power kind of deciding what projects are gonna go into CNCF. CNCF fundamentally is a governing body for projects, and really our goal is anything and everything that–we wanna provide basically the complete software stack that organizations need to complete their cloud native transformation. Kubernetes is really just the start. Beyond Kubernetes you see everything else that’s going on from monitoring to logging. I’m talking about projects like Prometheus and ____.

Those are some very mature projects now. Some of the more recent projects, things around security and policy management like OPA. I think those are very, very–

Shimel: Love OPA.

Liang: Yeah, and service match Linkerd. That’s also part of our CNCF. We last year contributed to a project called Longhorn. It’s cloud native storage, Kubernetes storage into CNCF. It’s a CNCF sandbox project now. So, it’s pretty much really the whole stack. If you think I was saying Kubernetes is TCPIP, then the rest of the CNFC, it’s the rest of the software stack, from the platforms to some of the key applications to some of the tools and equipment that more people have used.

They’re gonna build this cloud native ecosystem. So, I’m really excited about it. I’ll be serving this CNCF technical oversight committee for the next two years and play a role in really growing the scope, further grow the scope of CNCF.

Shimel: I just wanna say one other thing about CNCF, and that is we did a virtual event here called Predict last month where we took a look at a year in the future, and one of the points I made is that to me, CNCF is sort of the poster child for this third generation of open source that we’re in. I call it the first stage, sort of the cathedral in the bazaar back in the day. I know you know about it, Sheng.

And then we moved from the cathedral and the bazaar to what I call the big brother era where every open source project was really owned by one company, and everything that that project did was for the benefit of one company, and that discouraged others from joining in and helping do it. Now we’re in sort of the foundational era of open source where you have foundations managing these projects for the good of the whole community with multiple vendors involved, people like you on the board who are doing it.

Yes, it’s good for Rancher, but it’s good for all of the community, and that’s a very different–as I said, it’s the third generation of open source and it’s a great thing. Let’s focus in on Rancher a little bit though, Sheng.

Liang: Yeah, absolutely.

Shimel: So, we’ve got all these great projects. You’re literally building the stack or at least distributing the stack, your distribution of the stack. Fantastic revenue growth. What are people doing with Rancher that this demand is in such high demand?

Liang: Yeah, I know. It’s one thing that Rancher of course supports and distributes a lot of CNCF technologies including Kubernetes and Prometheus and Linkerd and Fluentd, and we’re gonna soon be doing OPA. That’s all valuable, but fundamentally, a company like us really only have a reason to exist if we innovate. We basically create things that people can uniquely get from us and not from anywhere else.

There are two pieces of content or two technologies that we created that actually help put Rancher on the map. One is K3S. It’s one of the most–if not the most–popular open source project in the Kubernetes ecosystem today outside of Kubernetes itself, and it’s just got tons of momentum. What K3S achieved was it took Kubernetes that was already proven in data center and cloud, and we brought it out to Raspberry Pi’s and surveillance cameras and set top boxes.

Shimel: This is K8 lite. I remember we spoke about this, the one in the KubeCons. I remember.

Liang: We launched GA at KubeCon, so it’s very big. We’re very excited about it. You remember, again I keep using TCPIP as an example because in the 90s when TCPIP just came, it was obviously a computer networking standard, but nobody probably could’ve imagined like a phone would also be talking TCPIP, right?

Shimel: They all do.

Liang: They all do now. I can see in ten years every phone, every camera will be running Kubernetes as well. Every car will be running maybe multiple Kubernetes. So, we’re very excited about that. Then the other big piece of innovation that really Rancher helped to make it happen is this thing called multi-cluster management. We believed Kubernetes again is a protocol. It runs everywhere, so you kinda need something to tie them all together.

It’s not really good enough to be able to operate individual clusters. You need to treat all the clusters as a resource pool. In fact, you actually kinda have to treat every single cluster as a cattle or as a pet. That’s where the Rancher analogy came in.

Shimel: I remember talking to you about that three, four years ago.

Liang: I know.

Shimel: It’s cattle, not pets.

Liang: Our namesake product Rancher, which is named after our company, is a Kubernetes management platform. It takes all of your Kubernetes clusters in the cloud. In the cloud, that could be GKE, EKS, AKS. Could be Kubernetes clusters on-premise. It could be RKE. It could be open shift. It could be Project Pacific, PKS. It could be K3S clusters on the edge in cameras.

We present it to your DevOps teams and ITOps teams as a unified computing platform, so you can manage access, you can deploy apps, you can run CICB, and that’s what Rancher is. It really started to resonate last year, even though we’ve been building it for three, four years, and that’s ultimately what propelled our tremendous business growth last year.

Shimel: Great. Sheng, we’re coming up on the end of our time. I’m gonna throw one more kind of topic question out at you we can discuss, and that is really what does the future hold for Rancher not necessarily from a product deployment point of view but you’ve started and had exits with multiple startups now in your career. Where do you see Rancher in two to three to five years? Do you go public with this? Is this something maybe some larger entity wants to acquire to kinda own the stack if you will? What do you think?

Liang: I think these days I probably think a little less about ultimately how Rancher will exit, just making sure we build a great business. I think we’ve kind of grown past the point that sort of an early stage easy equal hire or a technology would be viable.

Shimel: It’s not a tuck in.

Liang: We’re big enough to be sustainable by ourselves, and honestly, I absolutely believe it’s a big enough market. We solve a big enough problem that there’s a potential for tremendous business. Think about people actually all talk about hybrid cloud and multi-cloud, edge computing, but I think a lot of these terminologies are very attractive from a market demand or business case perspective but always lack some technical meat to actually bring it to reality. That’s actually been the problem.

Of course, we all know hybrid cloud is a real business need, but the problem is because there was no fundamental consistent experience that could actually implement that. So, as a result, no big businesses have been built. We kind of approach it a little bit differently. We’re very much centered around Kubernetes as a common protocol for compute.

I think by building on that, I really believe there’s a tremendous business opportunity here because fundamentally, we’re gonna solve these multi-cloud, hybrid cloud and edge computing problems. Those are business problems with Kubernetes as the common compute protocol. So, I think that will keep us busy for the next five, ten years, and I think this business has lots and lots of room to grow.

Shimel: Let the chips fall where they may. Sheng, we’re about out of time. I wanna thank you for taking time out of your busy day and joining and talking to our DevOps.com and Container Journal audiences. Always a pleasure to speak with you and get an update on what’s happening at Rancher Labs. I will see you in about a month and a half in Amsterdam for KubeCon.

Liang: Thank you. Thank you, Alan. It was great talking to you.

Shimel: Always a pleasure. Sheng Ling, CEO, co-founder Rancher Labs. This is Alan Shimel for DevOps.com and Container Journal, and you’ve just listened to another DevOps Chats.

— Alan Shimel