 Here with Kubernetes by example insider. I'm Langdon White your host and I am a professor over at Boston University and for this episode This is actually our 24th episode so we decided that it was our two-year anniversary Even though it's actually been slightly more So we decided that we would do a little bit of a recap of You know some of the episodes we've done in the past as well as kind of what we're doing with the show and how It's evolved and that kind of stuff. If you haven't seen it We actually did an episode called episode zero Where we introduced the show and what we were about and why we were doing it at all Way way back and we'll drop the link in the chat so you can check it out if you want to But I'm sorry short. I'd like to welcome at least is this all the co-hosts? This might be all of our co-hosts that we've had Well, not anymore. So early early on we we also had Chris short as a co-host and I think there's anybody else but Most of the ones that we've been who've been on the show in the past year or so are all here today and so I will Round Robin to have them introduce themselves And because he's in the top left of my screen Josh, Josh would sorry. Why don't you go first? Fortunately, I'm in the top left of my screen arrangement to Lingon. So I was expecting that you I might get Hi everybody, welcome to the 24th edition of KB insiders. My name is Josh wood. I am a developer advocate at Red Hat I focus a lot on open shift, but especially Community efforts in the open source realm that surrounded extend it and add new features and kind of Focus on the future direction that we see in the industrial orchestration platforms like Kubernetes I've hosted. I think three or four episodes of the show over the last couple of years I have some good memories of that We've talked about some neat topics with some folks outside of the company and involved in projects that we're all involved in together That's a great opportunity to do that All right, so let's keep it in the Josh's and Josh Burkist, which I can see yourself Yeah, other Josh Josh Burkist, which is why we have not hosted a show I have not hosted a show with Josh wood because it would be too confusing the I I'm in the Red Hat open source program office Where I I help Red Hat steward its contributions to Kubernetes. I started with Red Hat seven years ago actually because I was already working on Kubernetes at the time which is like That was before 1.0 but early days and and I Join the KBE team to help them find I Guests to interview From the Kubernetes community Savita you want to go ahead? Oh sure. Thank you. Thank you Josh. Hi everyone. My name is Savita I am a senior software engineer at Red Hat focusing on application modernization and I'm also a maintainer of the conveyor project which ties into the application modernization efforts But my heart lies in the Kubernetes community. I am a contributor to Kubernetes community That's where I met many of the folks in coding Josh Burkist And it has provided me with the opportunity to be here and I am looking forward to meeting more new folks Nice Daniel, I think it last but not least. Yeah, sure. Hi. My name is Daniel. Oh, I'm developed out of kid like Josh and then really different business unit But pretty much because some like a normal Java developers and then I also like some hosts is some of the couple of Shows for the past years and which he allows me a lot of opportunity to meet people like a super awesome Like a Kubernetes ecosystem people, but also it's a really really good chance for me to learn something new because I've been using Kubernetes for many years and then I Spent a lot of time bring more business application into Kubernetes for Develop a standpoint how to make it easier how to make it fast and how to make it useful and then sometimes I really be Just tie into very small area, but whenever I joined this Dubai example and KBE insider lot of awesome guest speaker and then just let me know Yeah, this is a better way. This is a maybe something new. So that's I'm really thankful and then still be here Nice so Like I said, we're kind of doing a little bit of a recap show And so one of the things I kind of first want to talk about is to see kind of if we're all on the same page And so I was gonna kind of ask You know, what do you think this most salient kind of component of the show is like what's the most? You know kind of differentiator or whatever that you think is why people would want to come and and watch the live stream And I would say, you know raise your hand if you want to go first or just kind of jump in Rather than calling on you and putting you on the spot So yeah, all right Josh go ahead. So something that makes me think of immediately It actually calls back to something Daniel mentioned in introducing himself, which is So at a general level and for all topics, I think this show has helped me Stay on top of trends that are coming down the road And I think right now the key trend that I see in this community is that effort around making things simpler for Developers to actually deploy things on what we might refer to in a general way as modern infrastructure So OpenShift Kubernetes container oriented orchestration on cloud services and all the different places We see that running and the last say say DevOps has been a 10 or 15 year cultural initiative That's changed a lot of the ways that folks work inside of companies and has joined the concerns of development and Operational teams in new ways that there was kind of a wall between them in the past There's a lot of advantages that come out of that and then kind of Corollary to that is some of the kind of SRE or site reliability engineering stuff that comes out of Google's Practices and how we run this infrastructure. All of those things are really cool They make our apps more robust in production. They make them easier to upgrade and release but they also dramatically proliferated the concerns for the daily developer the regular lady or fellow writing Java and Living inside of that application adding features and bug fixes to it all the time where five years ago I might have been worried each day mostly about my IDE and my compiler I build something might kind of throw it over the wall and I know someone else will take care of deploying it Now I have all this YAML To describe how a deployment works How we route traffic into that cluster once I get it deployed how the automated pipeline that builds my thing works and all of those they're really tremendous features that have brought a lot of cool cool advantages to the organizations that employ them, but they've increased the load on on individual developers so Being in the community and keeping track of efforts around abstracting away some of that proliferation And immediately something that's been a hot topic around the community lately that comes to mind Like backstage out of Spotify. It's not directly Kubernetes related Kubernetes is a major Deployment target for that, but it's really about tying together all of these disparate services and maybe Lightning that load of YAML that doesn't have anything to do with the application You're trying to work on but as all of those the applications deployment concerns and so that's just one example of kind of that trend I see around projects in the wider community that are oriented not necessarily in the central engineering core of the infrastructure, but are in Are about improving that user experience and Deep proliferating some of that complexity That that these initiatives over the last decade or decade and a half have brought into the daily life of your average developer You know, I also wanted to start quickly clarify one thing there is like it's it's not just like there's there's certainly a little bit more Workload, but it's also the cognitive load, right? It's really the that like how many different things you kind of have to keep in your brain all at the same time I think that's really what it kind of adds to that complexity, right rather than that literal work, right, right? Sorry, Josh. Go ahead Bergus. Oh, I was going to say, you know One of the things that I've thought about with Kubernetes is that YAML may have been a mistake But but not for the reasons that people talk about it being a mistake. It may have actually been a mistake because it was too easy Right because because the YAML was never meant to be the UI Right, the assumption was that people were going to go ahead and build tools that would write the YAML for you But I think the problem is that the YAML was just easy enough to write that people were really slow to build those tools Right, if it had been if it had been XML or syntactical, you know JSON HTML something like that something that that really required a lot of sort of annoying punctuation People might have been a lot faster to actually build tools to write it for them See, I was gonna say that it's some some people here Definitely and some people listening to the show may remember that exact same assumption about XML Which also took forever to get any tools because it was like yes It's harder to write than YAML. Sure, but it's still pretty easy to write And so as a result you ended up with a lot of handwritten XML You know doing kind of service-oriented architecture type stuff or you know, portally stuff. Yeah, I Think I think you're exactly right. I think that's a little bit of a problem You know with right, you know kind of these markup language models for sure Configuration if you're able to write it by hand you will by definition drift farther and farther away from a conventionalization or from Standard ways of doing things like one of the great things about stamping this stuff out with tooling is That you sort of get an You know, you are creating and enforcing a standard or a convention around how you do things in the boilerplate parts And the parts that get repeated between multiple YAML files if they're written by hand that that convention is necessarily going to drift Farther away. I mean just as one out outcropping of kind of having These markup style configuration languages that are very easy to get your fingers inside of yeah Yeah, one of the things I've read Not that long ago within the last couple of years, but I thought it was a really interesting is that You know the that people who tend to use IDEs tend to be working in languages that are more established and people who are More on the text editor side like VI emacs, you know, whatever Attend to be in languages that are more, you know, they're more dynamic They're changing more right because an IDE when you have a brand new language can't actually bring very much to the table Right because it's not standardized enough or whatever and I kind of wonder to some extent why if Especially like YAML with Kubernetes the dynamism of how fast it's been changing Also makes the tooling hard. So I think, you know, to Berks this point, right? You have both the easy to write or Easy feeling at the first, right? It's it's like writing pearl code, you know It's really easy to write the first draft and then all of a sudden two months later. You're like, oh my god What did I do? You know and then but you also have the problem of the IDEs can't really are having a hard time keeping up Because you have, you know, a new custom resource definition every, you know, six hours And it and it's something we saw and and it's the funny thing is that like people like me who who are infrastructure people, right? Because all the really what happened was the Kubernetes got built by infrastructure people and and this is why We're only now sort of Really having folks focus on hey, let's make things easier for application developers Because the problem is well, even as infrastructure people you start with the idea of making things easier for application developers But since you are not an application developer, your idea of making things easy is not necessarily their idea of making things easy the and and As a result you also end up with a lot of the command line stuff, right? I worked on Postgres for years And the primary Postgres UI was always and still is you know the extended SQL command line There are still things you can do on that command line that you can't do with any other interface the and You know, yeah, and you know, some of it you looked at like I don't have you saw some of the early Kelsey high tower Demonstrations of Kubernetes and showing people how to use Kubernetes and he even went More direct where he just inserted JSON directly into at CD Which you can totally do it actually works as long as you don't mess it up, right? The mess up you can break down your whole cluster, right the but anyway One of the things we tried to bring people with this show is To bring in people who talk about these sorts of things, right as in where are we going with Kubernetes? You know, what's sort of the next step? What's the next step not just in Kubernetes, but also in cloud, right? You know, are there other things that sort of thing we've looked at some stuff like eBPF and wasm and other technologies that either might replace Kubernetes for some people or Might be the next layer on top of Kubernetes, right serverless that sort of thing And so to really sort of give people an idea That's why it's insider to give people an idea of Where we might be going Because we're not done sort of building this thing All right, I'm gonna I'm gonna call on Savita. What do you think? You know, and and we're focused. I think a little inwardly, which is I don't think it's a bad thing But like, you know, kind of what is it that you find interesting about kind of, you know Being on the show or interviewing people are you know watching the show? Yeah For all the reason that both Josh's mentioned and for me There is also another thing that there is at this show brings people together I see the way that I see is that we connect people together when we interview Insiders, I mean when we interview interview people who are Kubernetes insiders or like The people who are maintenance contributors of the new technologies like what and what not You get to understand how the community is and not gives an opportunity for people who are listening to the show on How to get started on board it Contribute and grow the community in ways, you know, like make it sustainable So for me the connection that Puts the issue apart, you know Cool. Yeah, no, I don't agree. I mean, you know, it's it's right there in the word Which you know, people kind of gloss over a lot, right, but community, right, you know Open-source projects say largely they work because there's communities and you know sub communities and all that other stuff and You know communities have a Require a lot of effort to maintain and grow and all that jazz, you know Kubernetes is one example, but we have lots of open-source projects like that But we like even here at BU right like we have a community of instructors who will help us teach classes like we You know get people from industry to come and help teach some of the classes But we want to build a community there so that they can kind of learn from each other because the experience of an industry pro Becoming a teacher is not the same experience as somebody who went to undergrad and then went and got their PhD And then became a teacher and so you know giving them, you know, kind of like-minded people helps You know with the key differences about the things that they already learned in industry that maybe a You know a more traditionally taught teacher didn't and vice versa So I think that's been really interesting, but it's a lot of work to maintain it You got to build relationships You got to you know try to convince them that there's that there's stuff to do when you're not in the meeting You know Things like that and and I think communities are really hard and I think that's really really important So I sorry, I really strongly agree with you speak that that the you know Trying to build those connections and recognizing that there are people in your community who are driving the bus And paying attention to what's going on with that really can make a big difference All right, so Daniel, I think you're at your last what we're turning to you What do you feel like is the kind of biggest driver for you for kibby insider? Yeah, so I just a little bit thinking about that why you guys actually talked about I Pretty much agree with the Josh and it's a beta and another Josh thought as well, but in my case So why people are actually wanted to come to the show again It's a pretty much he the guest speaker and then they actually share their own experience rather than just Presenting their own great Chinese feature They actually everybody actually can go to the some conference in personal or virtual YouTube channel They can find all this is a new cool feature, but they cannot find what is people they could some Like a failure or some use cases or the real experience. So how do we how do that new? Technology and community help their daily job just like Josh said so, for example, we already Showed like a new trend and technology for example was and then like a ebbf It's a really buzzword and even like a ML stuff and then everybody's Talking about same thing. Okay. This is a recall product or technology. You can you should use that But people really wondering if I adapt to that technology how to make my life easier and then better I don't want to spend my time in after 6 p.m. I don't want to be bothered or something else And then the already show actually showed that that real story, okay And then they did our show actually share, okay Why this matter and then why would this technology was born to figure out this kind of stuff So people actually have some desire to Understand that and figure it out that so that's why people coming to the show and they're really willing to revisit again Yeah, that's what I'm thinking Yeah, and I I want to just like second or add to that a little bit because that at one level of listener or audience member like Folks who are already in the daily practice or working in a tech job somewhere doing development There's that sort of information. That's a really practical kind of case-studied information from the folks who have been guests and guest speakers on the show as opposed like Daniel saying to a Flashy demo that you might do at a conference might actually talk about implementation failures and complications and ways where Project X and project Y didn't initially work really well together And then it's sort of a level Even more of a beginner than that one of my favorite parts of the segments that I've watched and that I've been involved in and co-hosting is We'll have somebody who is maybe a co-founder at a startup or a Group leader inside the Kubernetes community or a project leader on some other open source project And we get to hear their story about not just why that project started but how they began in open source And how like what were the first projects they worked on what kinds of things that they do in those projects? I mean, I remember some guests talking about like really early on I was always looking for documentation bugs to fix and that was one of the ways that I got PRs moving up a channel and got To know some folks in the community and that led to other Projects I worked on and more like kind of coding oriented things. So I think that kind of Instruction for folks who might be really early in the path on kind of exploring open sources Not just how do you participate? But even how do you find the communities beyond the show and get into those communities because that community is such an important Support structure if you're going to employ Technologies that are generated by and coming out of those communities So so I think that's almost a two-fold function that that Daniel alluded to and then hopefully I had usefully expanded on a little bit Business information how to use the stuff and also some some career and community insight Well, how do you find folks? How do you start working on an open source project of improving something that you yourself or are using? Yeah, that's exactly what I asked a lot when I go to some Conference area like a booth area when I met a lot of the university and the college's student They're actually looking for the so okay Where I cannot find some of the more valuable information to get better job after graduation So, yeah, yeah, you should go to there and then there are tons of the information even if you see in several landscape There are thousands of projects. There's too many projects which one I need to choose That's why people need to like a find some Inspiration and insightful information rather than just normal knowledge Yeah, so I'm gonna want to jump in there because Josh you you stole my comment Which was you know one of the things I really like about the show is we asked two questions of every guest that comes on right? What got you into open source and what got you into Kubernetes and exactly to your point? I think it's really important about the kind of showing an example of how you can more get involved in open source By looking at these people who are successful in open source and kind of how did they get to that point? That's I think a hugely useful thing to know the other thing is kind of to both your points a little bit that I think is really useful at the show is You know coming from the developer side of the house, right? One of the things that is always terrifying about open source is betting on the right horse And so if you you know you often will have multiple technology options for how to approach a particular problem And when it's open source, you're not always sure which one is going to continue, right? So what I think is really useful about the show is that you know talking about like ebpf and wasm in particular Given the number of people that we've interviewed about it that given the number of people we've talked to about it You can make a pretty good bet that wasm and ebpf are going to continue They're going to be important and so making an investment I think Daniel you said you know kind of after hours or whatever in kind of learning those technologies Is probably a good one Whereas you know some other tool chain and for me one of the big ones where I got burned really badly and still Still think it was a great piece of technology was spring batch Way back in the day And I I spent a lot of time and energy. I centered an application around it, etc And then it's just kind of died At least as far as I could tell I don't maybe just come back and since I was really playing a lot in spring But the you know it was just kind of like there was you know had all this promise But you know what I didn't have a connection to really was a strong connection to the spring community To find out if there was a lot of people behind it You know, I just kind of have the docs, you know or the website or whatever And so I think that's one of the things that's really valuable about the show and something I hope we're bringing to our audience Um, I feel like I just closed the show, but um, all right, so let's see should we Uh, we could try talking about Um, uh, actually I like I like one of the other questions. We kind of jend up beforehand Can you all comment on like, you know, one of the things we try very hard to do is to get guests from really disparate backgrounds as well as disparate Kind of projects that they work on Um, I'm kind of curious about your perspectives on why is that a good thing or do you think it doesn't much matter? And I know the question is a little biased and I'm pretty confident I know all of your answers that it's more on the good side, but you know, feel free to to kind of weigh in All right, who are we going to make go first? How about uh, Savita you're up? Um, so I'm going to summarize it one more time before I answer it so the question is that why, uh, we are Are bringing the guests from the diverse background Yeah, so like Right, so like I think we work really hard to go And and find guests who are coming from diverse backgrounds So they kind of got into the space from diverse places, right? But then also a diversity amongst the the kind of things they're working on You know, because we do, you know, we interview people who are, um, you know Release managers, we interview people who contribute to docs We, you know, we had somebody who built one of the, um Which we call it certification, you know exams You know as well as things that are a little more obvious, right like ebpf and loss of et cetera And so the question is just kind of like Why does that help the audience when we when we work on that problem? No, thank you for that. I think, um For me personally, right before I became a co-host In rotating course bench in this show I was a guest and before that there were like four episodes and I was the fifth one um, I was in the fifth one so When I looked at the show I The first thing I noticed was like the divorce if there was like diversity in the show and yes and One of the reasons is that when I started after my grad school I There was lack of diversity around me. So there was no, um, it was hard to figure out Like I couldn't find a mentor and I couldn't find my place, right? I couldn't find The ways in which I could grow it is um, I moved from India to here So new place new culture new community and everything Um until I found Kubernetes community. So that's that one of the other reasons it's like super close to my heart There I saw so many diverse folks and so many folks who are like me who are different or like pushing and making things better and they all have like one common goal and Some of the learnings that I've taken from that place and then I try to make that Happen and conveyor to like make it a warm and welcoming place for everyone So that's one of the reasons that I feel like the diversity in speakers is Really really important that it gives someone to look forward to like You can reach out to the speakers later if you want to if you want to learn from them Or they can be our mentor if they have time for that of course But you can still learn from their experiences and you can follow them along if you want and Grow from there. So that is one important thing for me And another thing is that we have already discussed in That's like how it is beneficial that we talk about other technologies which are related to kubernetes Which might replace kubernetes or which can add on top of kubernetes I think that it's really important to stay in this cloud native world because everything moves so fast and like I cannot keep up And I think like Daniel said there are like thousands of projects in the landscape So need to know like what is important on which the industry is Picking up next or like what is at the end of the day one thing is like about open source another thing is also like You work on a product that pays you or like the employer like you have to make money the employer has to make money to pay you and sometimes you need to pick the pick out the right thing and these I feel like these little projects and The Directions can provide insights into what would be the next hit and that can actually tie into the roadmap and vision of Your product that you're working on at your workplace Yeah, um I would definitely agree with all that. I think the um, you know, I think it's really important that we try to make sure that You know, people can kind of see themselves in the work, right? You can see yourself in the show Um, and so I think that's a really big aspect and I think you know, I think as we've been talking about right It's really also really important to get kind of the whole spectrum so that You know any individual doesn't have to try to keep track of the whole spectrum, right? So, uh, all right, who else wants to answer that question? What else do you think? Well, I I certainly think um, I mean, I agree that there's a definite important function and kind of publicizing the the diversity that does exist in the in the kubernetes community I mean, we've just heard from someone who said it helped them feel like they might have a place in that community When finding that place was not initially a really easy thing. So I think um, With the show helping to illustrate that diversity by talking to folks from across that community who are both socioeconomically and and sort of that dimension of diversity that we think about very commonly But also like experientially diverse. Um, what do they work on what's important to them? One of the things I'm always trying to do is get daily developer types Involved in these upstream communities for exactly like a reason we've already touched on In other ways from another angle, which is If you don't have folks who like their main concern is writing java or writing python every day Then the whole community and the projects at its spurs uh, are all kind of uh dominated by infrastructure systems thinkers and that gets you a distributed system of a certain kind of predictable pattern and picture That that doesn't necessarily handle the concerns of folks who are doing the daily application development work who are doing maybe A security and compliance type work So in addition to like what we would call maybe traditional diversity It's really important and I think we're good at showing that diversity across the community as we interview different guests We're also, uh, I think trying to Engender a real viewpoint diversity and like the experiences and kind of the engineering mindsets and practices that each of these different groups Bring into the upstream projects because fundamentally if kubernetes remains something that You know, like if you've been around distributed systems or hpc or or systems thinking in general you look at it Oh, wow, it's really cool. There's so many patterns. I recognize they're very exciting to me there While failing to serve the actual most important audience Which is the the folks who create the applications that get deployed there because the the large pretty distributed systems I've spent a lot of my career working on Don't really do anything that human beings care about until application developers run code on them, right? They're like an operating system kernel, right? Right. So speaking of people with an application development background Daniel Yep So, uh, yeah, so diversity It's always super important thing and not only like some find out speakers But also like some looking at there's some specific topic You know, also kubernetes is already super major stuff everybody's doing kubernetes Even if everybody don't know what kubernetes is but the kubernetes is like a like a defector everybody needs to do that If you do not it's there's something kind of you behind there's some history the the the thing is I'm more looking into like an ecosystem on the kubernetes, which is just some kind of make me like a more Some new diversity approach. It's not only speaker, but also relevant to technologies. So some speakers should be Come from a new user perspective And then some guest speaker should be come from like a more founder of their technologies Because some like just josh said like a daily life. Okay. I'm just 20 years java Developer but not super awesome developer. Just like a normal guy I just spent like a six or seven hours at a home and then just develop my physiologic and java And then my challenge is I need to make some new Observability of my application, but the application should be running on kubernetes And then I don't want to use all technology and how do I do that? And then that is my mission challenge for the next two weeks And then maybe go to like some kbe or there are some guys actually Talk about how to make it better your application have observability Using open telemetry, which is the second largest commitment in the cncf of projects And then but I don't want to listen this is an awesome project and this is a new feature and how to instrument What is the SDK? I pretty much more percussion. Okay. How to integrate this new great Community project into my actual application with some example sample and the practices and what is the Some kind of troubleshooting Practices something like that. That's why we also looking forward to Find some people from Many different perspective like end user and the developers of platform engineer Even someone stuff like some like a rock star speakers as well. Yeah, that's I'm really That's my thought actually for the diversity and then find them speaker and then our upcoming topics as well So, uh, I I couldn't agree with this more. Um, you know speaking as a Um, you know strong development background, right? And then I spent a bunch of years at red hat, which is a bunch of infrastructure people, right? You know, or a lot of them, right? Um, and you know shameless plug, right? This is actually why we have recently started this It's kind of like on the bubble of whether it's a working group or it's like part of working group or whatever But um about trying to do a curriculum or higher ed Around what we call cloud native development, right? So, you know, if you go and look at most of the Documentation the training the certifications all that stuff. It's around kubernetes Most of it is about how to make kubernetes go, right, which is good But it doesn't really give much to the application developer about like What does it mean to build cloud native applications? Like what is the venture of an architecture? Why Why do I want to build things such that? Or like what do I have to change about how I build things such that they live natively in places like kubernetes Um, and so like I said shameless plug We're hoping to uh, maybe do a meet-up actually at cubecon Um, but you know, we're starting to try to put together some materials and really not focus on You know how to make Whatever thing go and instead focus on what changes in your application architecture Do you need to make to be truly cloud cloud native applications? Because they're the only ones that are really going to be able to Tolerate the environment that kubernetes provides So yeah, like I said, I'm wholeheartedly in agreement But uh, and I think that it's kind of getting back to the original question I think that's the key right is that the diversity is how we Provide You know software probably lots of other things too, but I've only you know I've only really worked in software is how we provide software that actually meets most people's needs It'll never meet everyone's needs But unless we're going and looking at perspectives from you know, kind of different, you know socioeconomic backgrounds different experience levels, you know different goals, right? You know infrastructure people versus application engineers, you know, if we're not kind of Bringing everybody into the house We're not going to build as good software as we will if we do So I think that's a huge benefit for it. Yeah an example I can readily give that I know berkus could speak on for hours is we built kubernetes for years without ever talking to any database people Exactly I can say that's not true. I was involved very I was involved very early on I was involved very early on. Why do you think we have stateful set? Well, I mean to be honest though the part of the problem with things like kubernetes, right? I mean problem, but is that you know It's not solving a new problem, right? Like at all, right? It's it's trying to universally solve it for a kind of a cross a bunch of environments The thing is is that things like databases in particular, right? Have actually solved this problem Most of the problems that kubernetes is trying to solve for a long time and does it very very well, right? If they're very good at horizontal scaling, they're very good at being modified to do vertical scaling If that's your preference, you know, they do it well. And so kind of getting in the mix of it Hey, you know, wait, wait, what what databases were you using? I would say You know, I I mean, you know between so it's not trivial to set it up But you know, I would say like I think postgres skills much better Um, you know as kind of a single instance with um, you know or replicas or whatever Then it kind of like I think kubernetes kind of gets in the way until you do things like as you say like stateful sets Like like it's kind of core design Kind of doesn't let a database do what it does. Well, does that make sense? Explaining that because you have applications that have their own notions of clustering is the way I have always sort of talked about And databases are the kind of canonical example There's also kind of a distributed file systems that have done a lot of work down kind of their own paths That were outside of kubernetes techniques and that proved difficult to adapt to that kubernetes api interface That's a lot of what uh the original inspiration behind the operators pattern and concept was really about how do we map Postgres's notion of what a cluster is to kubernetes notion of what a cluster I think bergus is just really close to postgres. And so therefore it only sees it and and in database technology in general It just I I found that really amusing because the whole reason I got involved with kubernetes to begin with Was I was trying to automate postgres failover Hmm because there was no way to do that natively with postgres tools Right like you could actually make the failover happen and it worked really well, but a human being had to trigger it um and you know as you do started down the road of of multiple steps of automation until we ended up on kubernetes because The problem is you get okay. We're going to have you know A watchdog in a script somewhere, but where does that watchdog live? What happens if the watchdog fails? You know getting all the way down that and and and you end up with oh, hey, I'm now deploying postgres in the kubernetes cluster the And and I do think actually I mean weirdly I think that Um kubernetes ended up solving problems. Believe it or not for databases before it really solve them For things like java application development um for the simple reason that In order to make kubernetes work the community attracted a lot of infrastructure people um right and they um and Those infrastructure people built things, but you know, there's a huge overlap between database people and sort of general systems infrastructure people There's not as much a huge overlap between infrastructure people And people for example who build application to front ends Right. Well, so so but I think java is a great example there Where java also had a similar problem, right because java has this You know robust model for doing, you know, this this virtual machine thing, right? Where it's meant and designed to be kind of like clustered, right? It's a big iron component and then you run all your little applications in it So when you came along and said, oh, yeah, we're gonna kill your your big iron virtual machine You know every 10 seconds Then something like cork has had to come along right because it's just like fundamentally at odds with the design of how Like java works, right? You have these application service and arguably c sharp back in the days of like mts and that kind of stuff Has similar problems because they're just they're designed for a long runtime Right, and then you have the applications that kind of come and go Because they were trying to solve the same problem that kubernetes is trying to solve You know, I just I think it's interesting right is that, you know, basically if you have a technology That's actually solving a problem pretty well Sometimes when a new technology comes along that's trying to solve it for everything It actually is the least, you know, it's kind of like cities, right? You know the inner city might be the first to get like cable television But may take forever to get whatever the latest upgrade is, right? Because they went through the cost effort and everything else for that first run and and took the You know took the hit, right? But they can't imagine going for the incremental improvement because it's you know, that's a really cool example You just use there about cable television because I live in an area of the country where that's very true about electricity You really don't want to buy a house that is in a former fancy neighborhood That was one of the first places to get wired because it still has that wiring right In the time since the neighbor is the neighborhood grew around it That stuff has newer wiring and that is actually that same problem RIT in physical terms that that we see with Adapting a former paradigm to like a more generalized newer newer api kind of like we're talking about Clustered apps and now I want to run them near adjacent to or on kubernetes and have some of these same Kind of problems. Well, you know, we we wired the jvm and an application server Very successfully 20 years ago. So it still has all of those techniques And it can prove harder to adapt to The the the latest edge of those techniques Yeah, I mean not to go down that rat hole too much I mean so boston very much like my neighborhood has this problem right getting fiber laid When you already have cable lines is, you know, kind of a daunting prospect right and like, you know And even getting power lines put underground because you know originally they're just holding on towers, you know, so It's kind of a similar problem if you have kind of a very advanced area early on It actually sometimes has a hard time adopting the latest in the advancements All right. So there was a rabbit hole. Um, hopefully all y'all, uh, don't Have to worry about getting fiber or something. Um, so let's kind of maybe move on from there a little bit. Um, you know I I have a very hard time answering this but one of the other questions we came up with in advance was like What was your favorite episode? Um, you know, I think for me the only one I can kind of say as kind of sort of my favorite was zero Like I really liked that we went and did a show about why we're going to do the show I thought that was a really good idea. I thought it worked really well I thought it made it a lot clearer especially for our early guests about what the point was um And and so I really like that but like since then I don't know that I have like I I think they've been really interesting I found most of them interesting most of them. I didn't know a whole lot about the community I found really interesting to talk to people Um, so I don't know what about all y'all Do you have a do you have a bias towards an episode? Or or content I'm trying to bring up the list of old episodes Oh, gotcha So for me, I actually uh, one of the yeah, I loved all of them But for me, it's just one of them actually around the K native uh The the uh, I remember that the project name Uh, what was that? Yeah, which one was it? I'm looking right now. Uh the chain guard. Yeah, the Oh, yeah. Yeah. He's the guy actually we met uh, and they uh Cube call in los angeles at the just I just left the cube call and then we just bump up and then he just hey So you love to k native because I really Spend a lot of time k name a project and he turns out and oh he actually I'm the founder one of the founder of the k name projects and then after that we Keep in touch with and then so I asked hey, so can you just Come to our show and then really we talk about to chain guard and the k name stuff Why actually you invented this cool project? And then it was good. I mean so It's uh, I just wanted to say it's uh We can meet a lot of people in person and virtually And then once we meet some people and then we have some relationship and then we can actually Uh Ask them hey, so I really want you to share your own experience and knowledge Uh on our show So that's my really good experience. I'm really I thought I'm so grateful Uh to be part of just the kbe team and then I'm putting some people into the show and then Give some chance to share their own stuff to so I will Random like a watchers and then our subscribers Yeah, um, yeah, I I definitely do You know have a bias towards like the serverless stuff as well Um, and you know, but then again the venture of an architecture is kind of the name of my game. Um, so yeah, uh, that's that's an awesome story I'm really glad that connection kind of got made and things were kind of moved from there Uh any other episodes or or topic areas that any of you like in particular? I know it's always hard to uh single anything out Yeah Go ahead. Go ahead. Okay. I will for simplicity's sake. I'll go ahead. I think this one will be short. Um Uh, certainly for my personal hobby horses. I really enjoyed the episode on sieve Which is this uh kind of test harness framework for kubernetes operators Uh that michael got michael gas who's a pretty good friend of mine I'm certainly a long time acquaintance of mine from core os days and forward We've we haven't worked together at one company, but we've worked together on a number of different things And I think that tech is fascinating Uh, but in a immediate, um, uh, professional life kind of way the episode. Um, that's pretty recent Uh about podman desktop Um was really kind of a one of the first stones and what has been a really popular series of efforts So and like I work on the dev advocates team Some of this has been worked out of my team rather than by me directly But podman desktop has been an extremely important Topic and an extremely popular topic for us over the last four to six months Um, and and I really kind of got my first exposure and any details about it and kind of started thinking about it And it promoting it in importance in in my own brain Um as a result of the episode here, um, and it went, you know, quick aside Podman is a container runtime analogous to docker and all of the other uh oci container running engines And podman desktop is a graphical user interface that makes it a lot easier to engage with the features of that container engine And and those those things have been Really popular, uh, both for uh myself and some other folks on on my team that have begun working really closely with the The desktop project promoting things they're doing getting information about new features out there Yeah, plus one podman desktop have been using it for I don't know since it was created You know, I actually podman in general but the you know desktop less open somewhat. Um, but uh, yeah, so savita. What about you? Far be it was the single release episode with shasha grune Um, I hope I'm saying this last thing correct. It's not I'm not sorry But that because I uh was part of a single part of the single release single release is one of the Saves for communities upstream and that is responsible for the For three releases we moved to three releases per year. They do more stuff, but that is the high level of you and um I really like the episode because I worked with shasha for A while and I have interacted We are slack messages, but that was the first time That I got to talk to him in person, right? We work for the same company He works, uh for redact as well and I have interacted in for your slack messages but It felt like I got to do them in person and I really enjoyed that and it was a fun episode. I didn't know like I I was like Would we be able to get to like the 50 60 minute mark of that episode, but I think we were like On time we We talked and talked and it was really nice. So I enjoyed connecting with um, shasha on that I like other episodes all other episodes too, but that was my personal favorite because I worked with someone and I got to know them Really well All right berkis. What do you got? You think there's a particular episode? Yeah, my my my favorite both to watch and to participate in um was the special episode out of detroit With with driving around interviews Yeah, yeah, uh, those were a lot of fun. Um, yeah, so uh, I I enjoyed doing them quite a lot. Uh, and so that's cool Um, yeah, I don't know Yeah, both both, you know from perspective of being in there and being in the the ford special electric vehicle, etc and and also I kind of feel like the fact that You were driving around Actually resulted in the guests being a little bit more free ranging in what they talked about Right that it you know, even more than than in a studio where they started, you know Following, you know a thought to to something else Um and getting off and discussing sort of tangentially related things some of which were quite fascinating Right. Yeah. No, I I would agree. Um It it it was funny actually because one of the things we talked about in the amsterdam was actually doing like like a canal You know like on a boat or something Where but I think it would have changed the dynamic a lot where because I would then be looking at them And like interviewing them rather than kind of looking ahead at the road, right? Where they're just kind of talking about their subject area, right? And I'm kind of like poking a periodic question Not so much. It's really uh, you know, I don't want to say it's a monologue as much as to say it's like But it's a little as you were kind of saying it's more free ranging Um, I think because there's no or there's less pressure Of you know of questions and you know and somebody actually talking to you Yeah, I thought I thought they came out really well. I'm putting the link in the chat as we speak. Um So thanks thanks for not taking me to the canadian border. I didn't get my passport with me right, right. Um, yes, that was a constant fear of because there's a Excuse me like this forced on ramp where uh, if you get on the on ramp, you can't get out again until you're in canada And I didn't have my passport many of my guests didn't have their passports So we were mildly concerned Right insert kubernetes analogy here the um Yeah, that's uh, that's a validating webhook, right? Yeah, right Um, so I wanted to take it back. I'll just a little bit. Um, I realize that uh, There's a topic that has come up in a whole bunch of shows that I found really interesting that hasn't actually had a show about it Um, which is backstage. Um, we even it even came up in this conversation. Um, you know, but I think backstage showing up And uh, you know, I just thought you know If you're if you're want to watch a whole bunch of old episodes, uh, basically the course of this year Backstage has come up almost in every episode. So that's been super interesting Okay, so I think we are just about out of time. Um, and so I wanted to thank all of you For being on the show today and reminiscing with us But also for all the times that you co-hosted with me. I really appreciate the support And your expertise and you know kind of bring in a different perspective as we were talking about earlier You know, if it's always just me, um, it's just my my questions, right? It's just my opinions And so I really appreciate the help With giving some different perspectives on the kinds of things we want to know And then kind of related to that in the background helping us identify guests helping us identify You know, some of our seed questions to try to have a good idea of what we want to talk about during the show Really appreciate all that work and I can't thank you enough. Um, so thank you all I'm going to give a little round of applause Um, anybody else want to have any other kind of closing remarks or should we kind of call it there? Many more All right, we'll see Okay, we'll see you all next time. All right. Thanks. Thank you guys. Yeah, talk to you later. Bye. Bye