The Relicans

loading...
Cover image for o17e – Kubes, MUDs and ADHD with Rich Burroughs

o17e – Kubes, MUDs and ADHD with Rich Burroughs

Mandy Moore
Single Mom 👩‍👧 🐶😺😺😺😺 Owner/producer: Greater Than Code 💕 #DevRel 🥑 WiT/D&I 👩🏻‍💻 Podcast Production 🎙 #BlackLivesMatter #python 🐍 she/her
・28 min read

Jonan Scheffler interviews Senior Developer Advocate at Loft Labs and creator and host of the Kube Cuddle podcast, Rich Burroughs about being someone in tech who struggles with ADHD and how he is coping and getting help with that, doing Developer Relations in the DevOps space: DevDevOps, perhaps? And, how he thinks Kubernetes is a good choice if you're just starting out in tech. Because as it turns out, it appears that this Kubernetes thing is not slowing down anytime soon!

Should you find a burning need to share your thoughts or rants about the show, please spray them at devrel@newrelic.com. While you're going to all the trouble of shipping us some bytes, please consider taking a moment to let us know what you'd like to hear on the show in the future. Despite the all-caps flaming you will receive in response, please know that we are sincerely interested in your feedback; we aim to appease. Follow us on the Twitters: @ObservyMcObserv.

play pause Observy McObservface

Jonan Scheffler: Hello and welcome back to Observy McObservface, proudly brought to you by New Relic's Developer Relations team, The Relicans. Observy is about observability in something a bit more than the traditional sense. It's often about technology and tools that we use to gain visibility into our systems. But it is also about people because, fundamentally, software is about people. You can think of Observy as something of an observability variety show where we will apply systems thinking and think critically about challenges across our entire industry, and we very much look forward to having you join us. You can find the show notes for this episode along with all of The Relicans podcasts on developer.newrelic.com/podcasts. We're so pleased to have you here this week. Enjoy the show.

Welcome back to Observy McObservface. I am your host, Jonan Scheffler. And I am joined today by my friend, Rich Burroughs. How are you, Rich?

Rich Burroughs: I'm great. I'm really excited to be here.

Jonan: I'm really excited to have you. I was a little bit surprised to realize that we had not done this thing yet. And one of my favorite things about podcasting is I get to see all my friends. [laughs]

Rich: Can I tell the story about how we met? Because I think it's funny.

Jonan: That'd be great.

Rich: One day, we're on Twitter and Kyle Shevlin, who's really rad, posts a tweet, and it's in the middle of the lockdown, and he's like, "Hey, I'm just going to open up my Zoom to random people. Come and talk to me." And we both hopped on that call, and that was how we met.

Jonan: Yeah. I love Kyle for that. And I had a couple of other friends who did similar things. I kept thinking I should do this, and I never did. [laughs] But it was such a lonely time for everybody and being able to pop in and make friends…and many of us happened to be in the Portland area as we both are. So since then, since we met on there, I don't know, in the midst of the chaos, the dark times, which are slightly brighter now, you were making a role transition, and now you've landed at Loft Labs.

Rich: That's right, yeah.

Jonan: What does Loft do?

Rich: Really, the big focus on the company is solving the headaches that people have with Kubernetes multi-tenancy, which if you've used Kubernetes, you probably have had headaches with Kubernetes' multi-tenancy. [laughter] The real issue is that most people, if they want to have shared clusters, they use namespaces to segregate people. And it's not a great model because things like CRDs are cluster-wide, and you can't just deploy those inside of a namespace. And so, we've got several open-source projects. The biggest thing that we're excited right now is a thing called vcluster that lets you spin up virtual clusters. It's pretty cool. So basically, on your host cluster, there's a dedicated namespace for the virtual cluster. And inside of that, we put a K3s API server. And so, to the end-user, it looks like they're in a real cluster, but they're just in a namespace inside of your host cluster.

Jonan: So I'm communicating with...like, I'm typing away on my Kubectl, and I have to --

Rich: I believe it's called Kubectl.

Jonan: I believe it is called Kubectl. And I get a lot of arguments from people, but they are entirely wrong because it's so cute.

Rich: Since I host a podcast named Kube Cuddle, I'm going to have to go with that.

Jonan: Yes! It reminds me every time I hear it of Portal with the cute, little Companion Cube. It's all of our Companion Cube. So you're using Kubectl, and you're communicating with actually this K3s API that then communicates with the host cluster in which it runs. It's not that it's communicating with multiple virtual clusters.

Rich: Actually, what happens is there's a separate process. There's a sinker, and what it does is it sinks certain types of objects to the host cluster. So the virtual cluster doesn't have a scheduler inside of it. It's the host cluster scheduler that actually runs the Pods. And so the sinker basically tells the host cluster, "Hey, here's a new Pod that you've got to run." And there are a few different types of objects it does that for, but because K3s is a conformant API server, it's certified. It lets you do a lot of things, and it really does feel for a lot of uses like you're in a real cluster. Probably the biggest use cases are for things like development, giving developers really easy access to their own cluster that they're not having to run on their laptop, that sort of scenario.

Jonan: Right. And so we're able to get our own clusters and go and break things, and we're not going to affect anyone else's development environment.

Rich: Absolutely.

Jonan: Even if we go mucking about with the CRD, which you'll have to explain to me what that is.

Rich: A CRD is a Custom Resource Definition. It's...

Jonan: Oh, I know this one and then...okay.

Rich: Yeah. It's still early on. There are improvements being made all the time. We just launched this thing like a month ago during KubeCon, but I think it's really exciting. There's also another project from the folks in the SIG Multi-tenancy group. So they have a thing called cluster API nested, which is their version of virtual clusters. It's very different than ours. One of the things that I love about vcluster is you can literally just delete the namespace, and it's all gone.

Jonan: Oh yes. This is a frustrating part because I am clearly very much a Kubernetes newcomer. And I want to explain the Custom Resource Definition thing and why that might not be part of this real quick to the best of my ability. I'm very limited here. But my understanding is there are a certain number of resources that come built into Kubernetes, things that are just like, here are the Legos that you have to play with, and you can define your own Legos.

Rich: That's right.

Jonan: You can go in and say, "Here's a Custom Resource Definition. And now I want to deploy a flibberty widget to my Kubernetes cluster. And this is how it behaves." But naturally, then I guess, well, it makes sense that that wouldn't operate well in a namespace arrangement, right? Because you can't deploy those two namespaces. I find that kind of surprising, though.

Rich: They are a cluster-wide resource.

Jonan: They are a cluster, okay.

Rich: There are a number of other cluster-wide resources to...

Jonan: Even some that are built into the platform.

Rich: Right. But literally, you might go and grab a Helm Chart, and it might have a CRD in it. And if you only have access to a namespace on the host cluster, you're not going to be able to deploy that.

Jonan: I see. Okay. This is fascinating to me. I started playing around with Kubernetes because I wanted to run it on Raspberry Pis. I've been into Raspberry Pis for a very long time.

Rich: [laughs]

Jonan: And I was playing with Kubernetes, and then there's this thing Pixie. I don't know if you've heard of Pixie. It's like an instrumentation tool that uses eBPF, the Extended Berkeley Packet Filter stuff, where they're able to get in the kernel and run hooks on system calls. And I was really interested in that as a strategy because it basically, in large part, sunsets the entire observability instruments. [laughs] Like, oh hey, thanks so much for the decade of instrumentation. We're just going to go here to the source. And it doesn't entirely replace instrumentation, but I expect over the next five years, or so it's going to replace a large part of it.

Rich: It's already becoming extremely popular. My friend Duffie Cooley went over to the company that makes Cilium recently, Isovalent. And I think Liz Rice is there as well, which they're really heavy hitters. So it's…

Jonan: I think it's catching on this thing. And I'm interested to see how people overcome the traces piece because, if I understand correctly, there's no way right now from eBPF to know which thread a request initiated in. So you can't do the cross-application tracing.

Rich: Oh, interesting. Yeah, I don't know a lot about eBPF, to be honest, except for the fact that it's really cool.

Jonan: Yeah. So anyway, I wanted to play with it. And then, I got my Kubernetes cluster set up on my Raspberry Pis. Shout out to Alex Ellis, who blogged a lot about this and all that other stuff.

Rich: Oh, awesome.

Jonan: And I was using K3s, and I get to the end of this whole thing and find out that Pixie at the time didn't run on ARM architectures. And I was like, noooooo.

Rich: Oh, no.

Jonan: I don't get to play with the thing I wanted. But I had a lot of fun, and I have a little side project wherein I may end up running Kubernetes on a fleet of Yoda dolls that I've installed Raspberry Pis in.

Rich: [laughs]

Jonan: You know, very important business work over here.

Rich: As one does.

Jonan: As one does. So you've been over at Loft, and aside from our work, we're whole people. So tell me about you, the whole person outside of this.

Rich: Imagine being more than a job.

Jonan: I know, it's weird.

Rich: That's just bizarre. I think that it's an interesting time for me. I went into DevRel I think it's about four years ago now. And I was in Ops for many, many years before that. So I spent a lot of time in roles that I don't really think exist anymore, but they were called application administrator. So I was doing manual deployments of code and configuring applications before things like Chef and Puppet existed and troubleshooting problems with the apps, and so it was very technical work. And then, I transitioned into being an SRE later on. And now I think it's been like four years since I've visited those practitioner kind of roles. And it's been really interesting. When I was young, I always was very interested in the arts. I always had good communication skills, especially writing, I think. And I was a theater major in college, and I used to do improv and sketch comedy. And communications was very much my background. And then I sort of stumbled into being a sysadmin in like the mid-nineties. It was interesting to me. Linux was a hobby of mine. And I just had these roles for a long, long time that were very technical. And now I feel like I've come full circle because I'm talking about these very technical open-source projects, but I'm using the skills that I have in terms of writing and talking to people. And I'm learning how to make YouTube videos and all of these things. So it feels like the whole person, as you put it like everything's coming together. I guess outside of work, the biggest thing going on in my life right now is that I was diagnosed with ADHD a few months ago, and so that has been a journey. Already, it's made a huge difference in my life. Aaron Aldrich, who's at Red Hat now, did a talk at our DevOpsDays in 2017 where I knew it was going to be a talk about mental health, and I was actually on the program committee, right? [chuckles]

Jonan: [laughs]

Rich: So I had even seen the abstract and things. But I get there, and I'm sitting down watching this talk. And it was a lot about his ADHD and the struggles he's had in tech. And it was like he was talking about me, and I was literally almost in tears at the end of the talk because all of these things that I'd struggled with someone else had these problems too, and there was a reason for them. And so, being someone with ADHD, I then, of course, didn't do anything about it, right? [laughs]

Jonan: Of course, yeah.

Rich: I even recently saw that I had photos in my phone of me holding one of the classic ADHD books trying to tell myself, "Hey, go buy this book and read it," But then at my last job, I think a lot of it was the pandemic, but I was really struggling. I just figured I needed to find out if this thing was real for me, you know? And so I went and got diagnosed. And yeah, like I said, it's been pretty amazing. I've been learning a lot. I have an ADHD coach now who is awesome.

Jonan: Who is that? Is it someone who does this professionally like we can share, and they could reach out?

Rich: Yes. Her name is Dusty Chipura, and she is fantastic. A number of my friends like Ian Coldwater and Chloe Condon and people had been to these bootcamps that Dusty does before and just raved about how great they were. And so I went to one of her bootcamps, and now I'm doing one-on-one coaching with her on a regular basis. And it's been so good for me. And then a lot of learning on my own, reading books, watching YouTube videos, which is the kind of research I like to do. Basically, watching TV is sort of what it's like. [laughs]

Jonan: It's a very engaging medium.

Rich: It is.

Jonan: And it's becoming more engaging by the day. It's like the internet expands into this...Being in DevRel, I think we have been maybe more familiar with this transition that we're starting to see across the industry what people are describing as polywork, which is now a company as well where you are working for yourself in various capacities and for other people in various capacities, and maybe also have a full-time job. And if you are in DevRel, for example, and you get a little audience rolling on YouTube, you've got an income stream there. It's unreasonable for your new employer to think, oh, well,; therefore, all work you produce is ours. It just doesn't work for the world anymore. But I want to give a shout out real quick to Dusty's consultancy; it's called Vancouver ADHD Coaching. But you can also just type Dusty ADHD into Google, and it'll come right up. I just tested that.

Rich: It does not surprise me. She's pretty well-known. I honestly feel pretty fortunate to be one of her clients. Like, she had a really big waitlist when I first talked to her, and she managed to get me in for sessions. The other exciting thing that I'm doing is next week I'm starting to take classes. The first class in a series offered by this...it's ADDCA. And that's actually the organization that Dusty got her certification from, and that's what they do, is that they teach people how to be ADHD coaches. So it's something I'm considering doing myself. So I'm taking this first class. I figure worst case; I’m going to learn some things about myself, some things that I can apply to my own life. And then, if I want to do more study and pursue it and maybe become a coach someday, I can do that.

Jonan: Polywork, you could do all of these things.

Rich: You know, it's funny because the CEO where I'm at, Lukas Gentele, is really awesome. And I like to be pretty transparent about these things. I talk a lot about my ADHD on Twitter, and I knew I was going to want to talk or blog about this journey. So I just talked to Lucas the other day, and I was like, "You're going to hear me talk about potentially becoming an ADHD coach. This isn't happening tomorrow. I'm not quitting my job next week. Here's the context." And he was super supportive.

Jonan: That's awesome.

Rich: And he was like, "Yeah, maybe you start it off as a side hustle while you're working here." So he very much, I think, appreciates the fact that I am a human being with interests and desires outside of tech as well as inside of it.

Jonan: I might start selling Kubernetes Yoda dolls as a side hustle. Nobody knows where the future could take me.

Rich: Wow.

Jonan: I know. So you've had a number of different roles over the year. Were you ever called a webmaster? Was that ever a job title of yours?

Rich: Oh, wow. I don't think I ever had webmaster.

Jonan: I took webmaster for myself when I was consulting. I was all about it. I thought that was the pinnacle of software careers when I was first starting.

Rich: I may have but not as a job-job. So I did a little bit of freelance web design when I was in college,, and this was in the '90s. So it was like HTML. “Can you write some HTML?”

Jonan: [laughs] Exactly.

Rich: That was freelance web design back then. But no, I think I first started off in a call center and then I became a sysadmin, and that was my title for many, many years. And then infrastructure engineer, SRE, all these new titles that were coming up basically for the same kind of work I was already doing.

Jonan: Now we call this DevOps, and now you do developer relations in the DevOps space, which is DevDevOps, I think.

Rich: [chuckles]

Jonan: We just keep backing Devs till we get to the end.

Rich: I just tweeted last night that I think it's DevSecKubernetesOps.

Jonan: Oh, wait, isn't that... it spells...it spells something too, K17...

Rich: Oh no, no. My friend KF tweeted this. It was d17s was DevSecKubernetesOps.

Jonan: Which if you're using leet-speak spells dits, and I thought it was very funny.

Rich: [laughs]

Jonan: Yeah, d17s. You always have to cut out the numbers in the middle. There's a VC firm, a very popular one too, that does this a16z, Andreessen Horowitz. I'm curious to know...I think internationalization was first. And then probably the VC people got in on it to get some street cred, and then it just kind of spiraled from there, and now everything has to be...

Rich: And of course there's o11y.

Jonan: o11y, yeah. Observability, o-1-1-y.

Rich: Yeah, but that doesn't stand for 11 letters.

Jonan: Oh, it doesn't?

Rich: Does it? [laughs] It probably does, actually.

Jonan: It must. No one would have made that off-by-one error. Actually, it would be kind of funny if someone built an off-by-one error into one of these.

Rich: [laughs]

Jonan: So I wanted to ask one more thing actually about your Linux...you said Linux was a hobby of yours. What was your very first Linux?

Rich: Slackware installed from floppies.

Jonan: Oh, you were the cool ones. Slackware is the cool...I feel like I was Red Hat, and by the time Red Hat came out, people were like pssh, it was all corporate by then. Slackware was the cool kid.

Rich: So I want to say I probably started using...I don't remember the exact timeline, but it was early '90s. It was before '95. And literally, I would go to the Linux users meetup here in Portland, and it was like 10 of us sitting around a conference table. You'd use RawWrite, and you'd burn the images to the floppy disks, and then you'd install it. And then, a while later, I picked up the Running Linux book from O'Reilly, and that had an actual CD in it that you could install.

Jonan: Nice.

Rich: And it was Red Hat. And so that was my first time using Red Hat. And immediately, I was like, holy crap, there's like this thing to manage packages, and they're not tarballs. And you can do things like look at what's in the package.

Jonan: What?

Rich: Yeah. So it's a weird story. But the reason that I got into this stuff is that I was playing MUDs. I don't know if you remember MUDs.

Jonan: Yeah.

Rich: So Multi-User Dungeons. So they're text-based games that were like the kind of RPGs that people play nowadays, like MMOs and stuff, but it was just all text-based. And so you'd be like, “Go north.”

Jonan: We had to have all the graphics in our heads, and I thought that was an awesome piece of it. For me, it felt like a very creative pursuit. MUDs were fantastic in that way. And now I love World of Warcraft and these kinds of games. They're fully immersive open-world-style games like Breath of the Wild. I am all about it. But there's still something to that. It may just be nostalgic. I'm sure if I were to play MUD today, I'd be like, okay, I'm done now. But it was pretty amazing. So you got into MUDs and the BBS scene.

Rich: Not BBSs so much, so I actually didn't own a computer at the time. So I would go over to my friend's house, who is a college friend, and he had an old Apple computer. I'm not even sure what model it was, but it was early Macintosh, I think. And I would go over there, and I would play on his computer. And then I finally was like, well, crap, I should get my own computer, so I don't have to go to my friend's house [laughs] for whenever I want to play this game. So I got a computer, and I was dialing in to the ISP and running the MUD client on the big sun box that they had. And then I heard about this thing called Linux, and I was like, wow. So I could use this Linux thing, and I could run the client on that. So I wouldn't have to deal with all the lag that was there from all the people that were doing random things on the big sun box with the ISP. And that was literally how I got into Linux, was so I could run a client to play this game.

Jonan: [laughs]

Rich: And it ended up just changing my life, you know? I got super interested in it. I got really engaged. I learned a lot on my own. Like I said that Linux users group was a really big resource for me. I met a lot of people, and that was the point where I started doing this stuff for a living. You know, I got a job at an ISP. The internet was taking off, and suddenly the bar just dropped really low for what you needed to know.

Jonan: [laughs] Yeah.

Rich: You didn't have to have a master's in computer science anymore to be a sysadmin. And I ended up doing that for many years.

Jonan: I was in Southern Oregon at the time, a very different experience from growing up in the Portland area, I imagine. As far as user groups went, there were none. I was one of maybe four people I knew in the greater Ashland area who knew what Linux was. And also, I didn't know very much about Linux. I was on the tail end of the knowledge in that group. And so I would install it, and then I couldn't get GUI out of it because graphics drivers were fallacious. And then I didn't really use it for a long time after that, actually, maybe until I got off to college, I started playing with it again. In the meantime, I learned to code on my TI-85 calculator, where we exchanged games with each other via cable that had to be connected. And we would write a Blackjack game, but then someone at a competing high school...you had to physically walk up to another human being and connect your cable calculator to calculator to transfer things. And so there was this kind of very interesting viral social phenomenon where the neighboring high school comes up with another Blackjack game with better graphics, and then we got to answer that. We can't stand for this kind of nonsense, you know?

Rich: [laughs] So these are basically diss tracks, your Blackjack games.

Jonan: Right. And we had a small group of friends who would exchange stacks of floppies with DOOM. I remember feeling like the...I'm trying to think of a way to describe like the elite-est hacker ever walking into a TacoTime somewhere and sliding someone a stack of DOOM floppy disks to trade for whatever it was. So this era, though, you make a point that the bar was significantly low. I once spoke to someone who was in 1997, 1998 making $125 an hour working in Silicon Valley just writing HTML. They were an HTML expert. And at the time, HTML was actually pretty straightforward. As broken as the internet was, there wasn't a whole lot of depth to HTML when we started to go into DHTML and XHTML and all of these other things. DHTML is Dynamic HTML wherein you just use JavaScript, but it had its own moniker and on and on. And I think then the bubble came around. And this is a thing I should know just off-hand, but it's like right around 2000, 2001.

Rich: That's right.

Jonan: And suddenly, even the people with the master's degrees, who got a Ph.D. in computer science, and you're out on a corner in Silicon Valley with a sign that's like, "Will code for food."

Rich: After the ISP, my next job was working for WebMD. And I was again in an application administrator kind of role administrator. So I was the person who would deploy the webmd.com site. And at that point, the company was superhot. It was taking out Super Bowl ads. I just felt like a rock star. I just felt bulletproof. Like, if I lost my job, I would get another one in five minutes. And there was a merger. They ended up closing the office that I was in. We got a retention bonus to stick around for a few months. So I was basically getting double pay for like three months. And I was like, all right, I'm going to take a break. And I took a month off before I started looking for my next thing. And in that month, the market had just completely dried up.

Jonan: Wow.

Rich: I spent ten months unemployed, and I had to take a role that was not nearly as interesting or challenging and paid 40% less just so I could pay my rent.

Jonan: Wow.

Rich: So, yeah, that was a really tough time. And the perspective has been interesting because the pandemic has been brutal. It's been hard on so many people. But in my head, in terms of my own employment, this is likely not going to get as bad as when the Bubble Burst. [chuckles] And so I always was able to calm myself down a little bit because of the fact that there were still jobs out there, at least for me personally.

Jonan: I think tech is actually going to be...there is an incoming recession, I am certain of it and a big one. Like, in many ways, we're in there now, but the tech industry, I suspect, may be pretty well-insulated just because we saw during the pandemic a huge increase in demand for internet services and many more jobs now. In DevRel, I've seen maybe a doubling of the people doing this work around the world over the past two years, and I don't see that slowing down. I don't think the internet's going to get less popular as we add another few billion to the high-speed internet world over the next couple of years like Starlink or whatever…the Google Crane project, I think, was the one where...I don't remember what they called it. It was some bird, Heron, where they had blimps with WiFi things on them. Like, someone is going to get to a place where we have global high-speed internet in the next several years, maybe five on the outside, where we almost double the global audience then for internet services. I get that there was a bubble that happened. But at the time, the global addressable market with access to high-speed internet and those modern internet services that existed at the time, I mean, high-speed for the time, that was basically the United States. And now we have grown all over the planet, and we are going to expand even more quickly. I think now is an excellent time to be in software.

Rich: When I was at WebMD, the company was just burning money, right?

Jonan: Yeah.

Rich: It wasn't all about the economy, right?

Jonan: Right. It might have been about champagne.

Rich: People were literally in this bubble. Yeah, we used to hear stories about the enormous bar tabs that the sales guys rang up. Salespeople, I think, were mainly guys at that point.

Jonan: At that time, yeah.

Rich: I do feel a little more comfortable right now. It is interesting what you talked about in terms of the pandemic itself. Because when I was at my previous employer, we actually saw that we had customers who were in certain verticals that were doing extremely well, right?

Jonan: Yeah.

Rich: Like, their business was going through the roof because of the fact that they were selling some kind of remote service or some kind of service that was oriented very much to people who were staying at home. And those people, their business was skyrocketing. But if you had a physical retail location, [chuckles] forget it.

Jonan: I know. I actually dodged a bullet. I've always wanted to own a game store, kind of a side hustle. I had a couple of friends; actually, we were going to go and buy a game store. I mean, I was ready. I walked in the week that things started closing down, still kind of thinking, all right, we've had bird flus, and we've had this, and we'll get this thing under control. And I almost bought a game store right before.

Rich: Wow.

Jonan: That said game store would have immediately gone out of business. There are few businesses you can acquire right before humans can't sit next to each other that would be a worse choice than a game store.

Rich: Yeah, absolutely.

Jonan: I got lucky, but maybe someday that dream will come around for me. So, in the next couple of years, if you were to make a guess at where things are headed for the industry or some segment of the industry, really...My goal here, my priority, is to be able to have you back on in a year and be like, "Rich, you were so wrong. Can you believe what you said?"

Rich: Oh my gosh.

Jonan: I'll play the tape back over and over again just like --

Rich: So I listened to your episode with Wesley, and I know that you did this to him.

Jonan: [laughs] It is my favorite question.

Rich: I really thought you weren't going to do it.

Jonan: I have to do it every time.

Rich: Two episodes in a row. I think that to me, probably the thing that's going on that I think is the most interesting is this idea of using the Kubernetes control plane for other things. And so the Red Hat folks, I think, are the ones who...is it K9C [laughs] I can't...

Jonan: K9CR917S

Rich: I can't remember all the letters anymore. But there's a project that's literally like a Kubernetes API that doesn't know about things like Pods. And I've heard Kelsey Hightower talk about this a lot, about this idea that we're standardizing on this idea of the Kubernetes control plane, or a lot of people are. And so it makes sense to extend that and use it for other things. And that's probably the thing that I'm expecting will become pretty real in the next few years.

Jonan: Yeah. If you're out there looking for something to hitch your career to, I think Kubernetes would be a good choice if you're just starting out in tech. [laughs] This Kubernetes thing is not slowing down, as it turns out.

Rich: Speaking as someone who just got a job at a Kubernetes vendor a couple of months ago, I'm going to agree with you. It's a fascinating project and for a lot of reasons. I first heard about it back in 2015. I think it was when Kelsey Hightower did a talk at a little local conference here at Portland that nobody knew about unless you were in the right circle of people. The thing that struck me right away was the idea that it was encoding a lot of the things that we were already doing. So in the shop that I was working in at the time, we had health checks, the equivalent of the Kubernetes liveliness and readiness probes. But we had to build that ourselves. Somebody looked at the API for our Cisco switch and had to figure out how to use it to determine when things were up and down and work that into the control mechanism we had for our internal platform. And so that was a lot of work. And every time somebody new showed up...like, we hired a guy who was great, but he loved Erlang. And he wrote a service in Erlang, and all the other services were in Java. And they just pulled in the library that had the health check logic in it. But he had to do it again from scratch. And, of course, they didn't have a spec because it was an Ops thing. It wasn't a real customer-facing application. And so, it was like reinventing the wheel. And now, these things are built into the platform. They are known quantities. There's no arguing about how they should work. There's no negotiation or building consensus about how a health check should work, right?

Jonan: Right.

Rich: It's just there in the platform, and you get it for free. And there's a lot of things like that, the scheduling and all of that. So that's the thing that I find the most interesting about it. I do have to say, though, I'm not a person who's dogmatic about tools.

Jonan: I definitely am. I'm all in.

Rich: Are you?

Jonan: No. [chuckles] I hope there aren't that many people in software generally who are like...I mean, I know that they may be speaking one way and exhibiting a different type of behavior. But yeah, I agree with you that you should not be. So go on. I'm sorry.

Rich: No, it's okay. So there are use cases that I think Kubernetes is really good for, and there are use cases that serverless or something else is maybe better for. I'm a big fan of the HashiCorp tools, and Nomad seems super cool. And I could definitely see being in a shop where I use Nomad instead of Kubernetes depending on the context, depending on the team, and the company, and the workloads, and all of these things. There's not a one-size-fits-all platform that's going to be the best for everybody.

Jonan: I 1,000% agree with you. I think my recommendation stands, for anyone starting out today, Kubernetes may be a good way to go, but don't discount Nomad or serverless or really anything HashiCorp does for the near future. They're pretty clever over there. So if you are giving advice to someone who's just starting out, someone who's, I guess, doing the equivalent thing to what you were doing back in the day with your MUDing and Linux floppy disks...I cannot even imagine what that equivalent is. But they're out there, and they aspire to work their way into this career. What advice would you give to them?

Rich: I think it depends a little bit on the kind of role, but especially on the Ops side, I'm probably stealing this all from Kelsey Hightower.

Jonan: As everyone does, yeah.

Rich: Yes. So I do a podcast called Kube Cuddle that's about Kubernetes; specifically, boy, you should follow Brendan Gregg on Twitter and read everything he's ever written.

Jonan: Oh yeah.

Rich: Speaking of eBPF, [chuckles] he is so smart and a lot of other folks too; Jess Frazelle is awesome, and a lot of other people talk about these things. But I really do think that for anyone going into any kind of DevOps or SRE role, that's the place to start, or with Windows, if that's the world that you're in. If that's what you want to work with, Windows has come a long way in the last few years with the ability to automate things and the Nano Server and all the stuff that Jeff Snover and the folks over there have been doing. So I think you probably have to pick. I think it's probably hard to go deep on either one. But that's the place to start is with the operating system and the basic primitives that are there. And like I said, I feel like I'm parroting Kelsey from our interview. But it's like the other tools that you work with, and the other things are going to build on those same basics, right?

Jonan: Yeah. Start with the basics. Kelsey gives really good advice. I think the thing that comes to mind is like maybe you won't ever SSH into the server. But when the time comes that that becomes necessary, the world is probably going to be on fire around you.

Rich: [laughs]

Jonan: There's going to be a very serious crisis, something going on where it's handy to be the person who can SSH into the server and strace the process. There are useful things in that level of basics, and everything builds on top of it. Learning just the layers of abstraction is not something I would recommend to people early on in their careers. Well, I think with that, maybe we have reached a natural conclusion here. It has been an absolute pleasure having you on the show, Rich. I'm so glad that you came to talk to me this morning.

Rich: I had a blast. I'm really glad I was able to come on. And I look forward to hearing you berate me in a year about how stupid my prediction was.

Jonan: [laughs] I can't wait for it. If I get my Yoda dolls to run Kubernetes, will you let me come on Kube Cuddle to talk about it?

Rich: Absolutely.

Jonan: I have a long learning curve between here and then. I think I've got a lot of CRDs to learn about. But I have faith a year from now,; I will be prepared, and I'll come in and talk about my Yoda dolls with you.

Rich: We'll almost have to do a video or something.

Jonan: Yes.

Rich: Because I feel like a podcast is maybe not going to capture all the glory.

Jonan: It may not do it justice. All right. Well, thank you again. I hope you have a wonderful day.

Rich: Yeah. You too.

Jonan: Thank you so much for joining us. We really appreciate it. You can find the show notes for this episode along with all of the rest of The Relicans podcasts on therelicans.com. In fact, most anything The Relicans get up to online will be on that site. We'll see you next week. Take care.

Discussion (0)