[MUSIC PLAYING] ABDEL SGHIOUAR: All well, this is a little bit of an unusual episode. We are doing a 2024 recap, and this will be our last episode of the year. And we have Mofi on the show. Hi, Mofi.

MOFI RAHMAN: Hey, everyone. Um, I don't know if it's back, but I'm here.

KASLIN FIELDS: Welcome back, Mofi. We're glad to have you on for this, our first year-end recap episode. We haven't done this before, so we hope you enjoy it.

ABDEL SGHIOUAR: Yeah, we just figured let's just jump on a call, the three of us, and then talk a little bit how 2024 has been, not only for the podcast, but I think for everything Kubernetes, Cloud Native, KubeCon, life.

KASLIN FIELDS: The universe and everything.

ABDEL SGHIOUAR: The universe and everything. The universe and beyond, I guess.

[LAUGHING]

KASLIN FIELDS: Sure.

ABDEL SGHIOUAR: So I don't how do you want to get started, Kaslin. Should we go through how the year has been for you?

KASLIN FIELDS: Yeah, we could start with general overviews, or we could start from the beginning.

ABDEL SGHIOUAR: Beginning of times? Epoch, 1990.

[LAUGHING]

That's such a nerdy thing to say.

KASLIN FIELDS: Well played.

ABDEL SGHIOUAR: We can start with some stats if you want.

KASLIN FIELDS: Oh, yeah, that's a good idea.

ABDEL SGHIOUAR: Yeah, we typically publish a report. We did that last year and we're planning to do one this year. And usually we share a little bit of how much followers and how much kind of listeners we had through the year. And I have some stats in front of me. Maybe we can start there. What do you think?

KASLIN FIELDS: Yeah, let's do it.

ABDEL SGHIOUAR: All right. So starting with 2024, we are as of now at around 300k total downloads, right? For the entire year. Bear in mind, we're recording this kind of end of November, so we still have almost a month to go. I think that the first thing to bring up is we have been much more consistent this year compared to last year.

[LAUGHING]

KASLIN FIELDS: So much more consistent.

ABDEL SGHIOUAR: Yeah, we published every two weeks, except the month of June, when we had four special episodes for the 10th year anniversary of Kubernetes. And so based on the stats, it actually shows we don't really have that much spikes in downloads. It's more consistent.

KASLIN FIELDS: Yeah, I'm excited to publish a report like we did last time. Last time we had like a graphic that we put together of how the podcast was doing, and we posted that on all of our social media. So look out for that soon. I'm very interested in how much more consistent the graph is this time. [LAUGHS]

ABDEL SGHIOUAR: I mean, I'm looking at it. It is very consistent. We had a little bit of a dip in listening in July, but that's because we suddenly went from four episodes in June to two episodes in July. So that's to be expected, I guess.

KASLIN FIELDS: Yeah, it was an interesting time this summer with the 10 year anniversary of Kubernetes.

ABDEL SGHIOUAR: Yeah, that was a lot of work. We should have probably brought Mofi to help us a bit.

KASLIN FIELDS: [LAUGHS] Yeah.

MOFI RAHMAN: So I think we should just start off by talking about the tenure of Kubernetes. Kubernetes is ten-years-old.

ABDEL SGHIOUAR: Sure.

MOFI RAHMAN: One decade old. So what has been your experience in talking to the community? I have been talking to a lot of people about how they feel about Kubernetes in its second decade, and I have had a chance to chat with some of the original maintainers, as well as some of the newer maintainers, as well as end users. So what has that been like for you all?

KASLIN FIELDS: It has been certainly an adventure. [LAUGHS] The interesting thing about it for me is that it's gone the entire year, really. From the very beginning of the year, there was planning for KubeCon EU in Paris, and so the community was all excited. It's the first KubeCon within the year of the 10-year year anniversary.

So of course they're going to do a whole bunch of 10-year anniversary stuff, to some extent. Except also the time between KubeCon North America and KubeCon EU was particularly short last year, and so there wasn't a whole lot of time to plan stuff for KubeCon EU.

So people were kind of scrambling to get things together. But we did get some really nice stuff done. I worked with Tim Hockin and Josh Berkus to design a t-shirt for contributors for the Contributor Summit.

I also designed a new Kubernetes contributor pin. If you are a regular contributor to Kubernetes and you attend one of the contributor summits or now maintainer summits, you used to get a patch to say you're a Kubernetes Contributor, but we switched it recently to pins that say that you're a Kubernetes Contributor. So fun fact and another reason to contribute.

And then in July or June, of course, there was the 10-year year anniversary event which planning. That was a whole big thing. I got to be in a documentary about the 10-year anniversary of Kubernetes, which was super cool. Got to go to the event where they premiered part of it, and they had all of those wonderful talks from folks who were involved in the early days. That was a really fun event, just to talk with people. And then KubeCon North America. So the 10-year anniversary spanned the entirety of 2024, let's say.

ABDEL SGHIOUAR: Yeah, I think for me, it was certainly interesting. So the first thing I'm going to say off the top of my head, I can finally say I am a Kubernetes SME because I have seven years experience on Kubernetes, right? I can put that on my resume. Subject matter expert. All this, from time to time, like a job posting on LinkedIn ask for people to have 15 years experience on Kubernetes. Sure.

MOFI RAHMAN: Yeah.

ABDEL SGHIOUAR: Obviously. [LAUGHING]

MOFI RAHMAN: I think like only person in the world that would probably have 10 plus years are the people originally at Google that was started working on Seven. That's the history for-- right? Seven of Borg.

ABDEL SGHIOUAR: Yes.

MOFI RAHMAN: So they are just over 10 years of experience. Seven of Nine, yeah. Right

KASLIN FIELDS: Of the Borg.

MOFI RAHMAN: Oh, boy. That's history. We can talk about it all day, but--

KASLIN FIELDS: Pop culture.

MOFI RAHMAN: Pop culture references. But like, yeah, that listing that we saw for, like, requiring 10 years of experience, like about two years ago, there was a job listing that actually came out--

KASLIN FIELDS: Longer than that. I feel like from just a couple of years into Kubernetes, those started popping up.

MOFI RAHMAN: Yeah.

ABDEL SGHIOUAR: I've seen 15, so.

MOFI RAHMAN: Oof, yeah. I think 10 is like such a solid number, 10 feels like it just established to a certain degree. So I think that's what the recruiters were coming from. Like if you have 10 years of experience on something, you probably know what you're talking about for the most part.

ABDEL SGHIOUAR: Yes. Yes, yes, definitely. Yeah, so for the 10-year anniversary for me it was two meetups. So I'm going to say the names here, but they're like two small cities in the Nordics, one in a city called Aarhus, three hours away from Copenhagen, and another one in a city called Bergen, six hours away from Oslo. The funny part was that there was actually a 10-year anniversary party in my city, in Stockholm, but I couldn't make that one, so.

[LAUGHING]

But like, yeah, as Kaslin said, the year fell short. There was a lot of things going on following each other. And here we are at the end of the year.

KASLIN FIELDS: I'm not sure if it felt short or long or both, but-- [LAUGHS] One other thing I want to call out is Mofi, you mentioned that if you have 10 years of experience now, which a number of people do, you're probably working on Borg, but also folks who are at Red Hat and working on the very early bits of Kubernetes, like Clayton Coleman.

MOFI RAHMAN: Yep, definitely.

KASLIN FIELDS: It was really nice to get to interview Clayton and Dawn and Kelsey Hightower and Tim Hockin, such a variety of folks who are involved with the early days, and then some of the newer leaders in the community, like Nabarun and Nikhita.

So it's amazing how the community has grown. And one thing that I really loved that I learned from Dawn was about the origins of Borg. I've read the Borg papers before, but hearing her talk about the original ideas behind Borg and how that inspired Kubernetes.

And to me, the most interesting-- I don't know, very interesting thing at least, was that Borg was really originally designed for AI style workloads, and then that kind of got left behind when it moved to an open source project, because that wasn't so much needed at the time. But now that it is, it's like going back to those roots. I find that fascinating, and I keep thinking about that whenever I see stuff about AI on Kubernetes.

ABDEL SGHIOUAR: Yeah, definitely. Yeah, it's kind of interesting how we're coming full circle, how it was designed and then how it was used. And now we're back to almost like square one.

KASLIN FIELDS: Except with a lot more technical debt.

ABDEL SGHIOUAR: Oh, yes. Yes, yes, of course, of course. Morfi, what about you, how was the 10-year anniversary for you?

MOFI RAHMAN: I did not get to do as much stuff, but there were a few events that I got to go to. There was a meetup in New York that I was participating in. I think for me, it was mostly about talking to people and how they're feeling about. One fun conversation I have had where someone was like oh, yeah, Kubernetes is now legacy, we need to start a new thing because it's too old.

Which is kind of like the cycle of tech, right? Like you do something for some time and then the initial ideas probably gets overloaded with newer ideas. And at some point, you feel like, OK, this thing is too complex. You got to start over. Hopefully that doesn't happen with Kubernetes, but it is the cycle of tech. It will happen eventually where newer and newer people are going to come, and a lot of the original ideas that initially started Kubernetes are going to get layered under a bunch of newer ideas.

So it's going to be kind of complex and convoluted to understand exactly what the purpose of this whole thing is. And people are going to look for something simpler. We already see this with tools running on top of Kubernetes for application delivery, or CI/CD that writes their own DSL instead of directly targeting Kubernetes YAML, which again, even in like 2015, Kelsey Hightower was talking about how Kubernetes is not the end goal, it's the platform to build platform on top of.

So ideally, in this world where I have spent so much effort and time learning about the system, we build on top and not just build anew, but we will see what the next decade takes us. I think this is partly our responsibility to make sure the new people that are coming into the system have enough of a resource for them to feel like they belong, both basically technically and as a community, as well as they have rooms to grow and flourish and learn and build new things.

KASLIN FIELDS: I think those are the most important things for the community going forward. I am amused at how we kind of all framed the 10-year anniversary in terms of events, because it's really a community moment. The technology will keep developing. There's no--

We had three releases this year, which were awesome, and there was a lot of really good stuff in them. But there's nothing that's like, this is the 10-year anniversary. And so technically, the project has changed dramatically. It's a moment for the community to celebrate how far it's come, and to think very seriously about how we're going to continue to maintain it and improve it going forward.

MOFI RAHMAN: And speaking of those releases, right? So those releases are like a kind of time locked, where we actually decide when the releases are going to happen even before the year starts. So it's more about features making into the release rather than features making a release happen.

KASLIN FIELDS: That is a good point.

ABDEL SGHIOUAR: Yes.

MOFI RAHMAN: Right. It's not like some new code got written. So we need to make a new release, is that the release is going to happen in August, whether or not your code makes it is up to you. For each project maintainer. So for that reason, each individual release is I guess, less-- I'm not going to say less important, but I'm going to say it's less dependent on the code that gets written for it, because release just gets time locked three times a year this time is going to happen.

They fluctuate by about a one week time at max, but usually it just happens on the day it was said and scheduled for months at this point. So whether or not a lot of new cool code comes out or not really is just keeps happening.

And Kubernetes, I don't think, will ever get to 2.0 because it just like it comes with breaking changes, which is, I don't think at this point, feasible for the amount of things that are running on top of Kubernetes. So we're going to just keep having newer and newer stuff in the same brand, like same 1.0 Kubernetes.

ABDEL SGHIOUAR: Unless there is a major change in the architecture, which I don't see it happening myself. But like unless there is some major changes.

MOFI RAHMAN: Yeah, I hope that does not happen. We don't want another Python 2.2.3 shenanigans.

ABDEL SGHIOUAR: Yeah, no. I agree.

MOFI RAHMAN: That took-- 2022 is when finally they deprecated Python 2 right? That took 14 years. So no, thank you. No, thank you.

[LAUGHING]

I don't want it.

ABDEL SGHIOUAR: It's definitely an interesting observation what you're talking about that. It's like each release is about whichever feature makes it to that specific date. And if you don't make it, you don't make it. You have to wait. But I'm willing to bet that there is probably people that are not willing to wait and they just like, build their own release because it's open source. Like nothing can prevent you from just pulling some code and running it yourself, right?

KASLIN FIELDS: Kubernetes is such a big project though, and it's used in such important production environments. I'm sure the majority of folks just let the community do the release, and probably don't try to mess with it. It's such a large code base, I would be nervous trying to build it myself with code that is not in the last release. You could technically, I suppose, but--

ABDEL SGHIOUAR: Speaking of that, I have one thing to bring up here. I think 2024 was the year of LTS, long-term support.

KASLIN FIELDS: Yeah, that's an interesting take. All right.

ABDEL SGHIOUAR: I mean, finally, all the cloud providers now have lined up behind LTS. It started with Microsoft last year, then I don't remember when AWS really like announced it. It was like early this year or late last year and then Google joined in. So now the three major cloud providers at least are all providing some form of long-term support.

KASLIN FIELDS: That's a fair point from a cloud provider perspective. I feel like from the open source perspective, a lot of those conversations happened at the end of last year, beginning of this year. So I don't know that I would call it 2024 the year of LTS from an open source perspective. But you got a good point on the cloud providers.

ABDEL SGHIOUAR: On the cloud providers, yeah. I mean, definitely the working group LTS still exists and they're still in discussions, right?

MOFI RAHMAN: Yeah.

KASLIN FIELDS: Their last communities discussion was very interesting.

ABDEL SGHIOUAR: Oh, yeah.

MOFI RAHMAN: I think 2024 is, I would say the year of implementation of LTS.

ABDEL SGHIOUAR: Sure.

MOFI RAHMAN: So you start doing something. This stuff takes time. You are trying to move a lot of-- And LTS is such an interesting concept. Because even without LTS, a lot of enterprise did not have the velocity to just hop versions that way anyway, so Kubernetes was always LT. Now with the S, it's the support.

[LAUGHING]

ABDEL SGHIOUAR: Definitely.

KASLIN FIELDS: The community isn't going to do that much on that's S itself.

MOFI RAHMAN: Exactly. Yeah, for most people, Kubernetes always was LT because they could not just version hop in every three months.

ABDEL SGHIOUAR: I guess if I get a dollar for every time I talk to customers and they say, yes, we are still running 1.26, I'll be probably rich by now. [LAUGHS]

KASLIN FIELDS: They're that far already? 1.26.

ABDEL SGHIOUAR: Yeah. Yeah, I can assure you I had many, many conversations. Yes, I had many conversations with people who are still running 1.26.

KASLIN FIELDS: Also that is the reason that I say that the last long-term support community discussion was very interesting. It was basically that the open source community can't do much about the S because we don't have funding to fund support people for you to call or anything. So what does it look like for the community to do long-term support? And I think the cloud providers implementing it is a good way to frame it.

MOFI RAHMAN: One other thing I was going to actually bring up from an earlier point that you mentioned about Kubernetes being open source and anyone can pull any changes they need to make. This is probably a call out to the listeners. If you know anyone successfully doing that for longer than three months, or let's say six months, I would like to hear about it.

KASLIN FIELDS: That would be fascinating.

MOFI RAHMAN: Yeah, because you could yes, fork and make that one change yourself. But Kubernetes release is going to go forward. The new version is going to come in. How do you then go back and reconcile back, or do you just keep for every new release, just keep forking and just re-adding the change that you just made? Because that is-- That sounds like a nightmare scenario to me, forking.

KASLIN FIELDS: Yes.

MOFI RAHMAN: Maintaining that.

KASLIN FIELDS: I think the S is going to have some problems on that one.

MOFI RAHMAN: Yeah, so if you or any of your friend that does this tell me, I would like to hear your stories. And, yes.

ABDEL SGHIOUAR: Or your nightmares, really.

MOFI RAHMAN: Or your nightmares.

KASLIN FIELDS: Another example of-- I don't know if we can call LTS standard that was implemented by the community, and then the implementations are created by other things. But in that vein is another trend in the community and in the technology, which is Gateway API. I was excited to see Gateway API continue to develop this year.

So the idea behind Gateway API, if you haven't heard our episodes on that or checked out any of the sessions, which are always packed at KubeCon, is that the community created this custom resource definition, they did not implement it as a feature for release for a few different reasons, but it's a standard that is then implemented by the providers in how Kubernetes handles ingress.

ABDEL SGHIOUAR: Yeah, you can call it Next-Gen Ingress, basically. You can call it next-gen Ingress. So it started in 2019, but it has matured quite a lot over time. And as you said, like every KubeCon that talks about Gateway are always packed. On that same topic, I think what I'm excited about next is the LLM Gateway, which is based on the Gateway API. So that's the next thing that is coming out.

KASLIN FIELDS: That was something that I found very confusing at KubeCon North America, honestly. Because there was this one set of keynotes where everybody was talking about LLM Gateways. And it seems like there's several different implementations going on, but there is one that's coming from the Kubernetes open source community based on Gateway API.

ABDEL SGHIOUAR: Yeah, there is a reference architecture, I guess, or a reference implementation that will be coming out. I guess we'll have to figure out a way to cover that in the podcast for sure which would maintainers.

KASLIN FIELDS: Yeah, we should have an episode.

ABDEL SGHIOUAR: Yeah. But yeah, it's definitely something that I'm excited about. I don't want to go to-- Let's not use time now for the details, but it will be very interesting to look at for next year for sure.

KASLIN FIELDS: Well, I do kind of want to dive into the details a little bit since we have our AI correspondent here, Mofi.

ABDEL SGHIOUAR: Oh, Mofi.

KASLIN FIELDS: Do much about LLM Gatewas, and why are they so exciting? I don't quite get it, honestly.

MOFI RAHMAN: I think the main part that is going to be exciting with the LLM Gateway is that ideally, you could treat your large language model serving as a regular web application, but the things you are looking for on a regular web application that-- like, your HPM metrics and other things, usually depend on your resource like CPU and RAM. LLM Gateway is going to give you a bit more signal of what kind of queue you have, and what kind of requests are coming in, and how do you scale your accelerated workload in a different way.

So it is very similar, but I think the reason they are a little bit special because they depend on accelerated hardware, which you need a little bit extra care. So the gateway is still the Gateway API, but because it's large language models, because it's like GPUs and TPUs, you just want to have this additional metrics to be able to scale properly and do a lot of fun stuff.

And security is a big one. Managing queues, managing being able to route traffic to the right place within the least amount of time and the least amount of cost, that is important. So there is some additional care that is required.

KASLIN FIELDS: I feel like hardware is the name of the game for why AI is so exciting in the world of infrastructure is that it's just putting so much more focus on what kinds of hardware you have available to you, and how you're using a variety of different types of hardware in your distributed systems.

ABDEL SGHIOUAR: Yeah, definitely.

MOFI RAHMAN: One other thing I was going to bring up, I think 2024 also, it's not as commonly thought about, but it seems to me like a year big win for security, as well. Falco graduated this year and this KubeCon Cert-Manager graduated.

ABDEL SGHIOUAR: Yes.

MOFI RAHMAN: So we have two big like a security focused CNCF project that graduated, so that is a pretty big news.

KASLIN FIELDS: And there was also a whole day of keynotes at KubeCon North America, just dedicated to security. And I believe in Paris, they had that like unconference security, unconference thing that they did at one point. So it has definitely been a big theme this year as well.

MOFI RAHMAN: Yeah.

ABDEL SGHIOUAR: Yeah. And for 2025, they are consolidating OSS security and cloud native security. So there will be one conference called Open Source Security Con.

KASLIN FIELDS: Oh, interesting.

ABDEL SGHIOUAR: Which will be a flagship event, actually. So there will be five KubeCon. And then they don't call it cloud, they call it an Open Source Security Con. So yeah, that's interesting.

KASLIN FIELDS: One of the interesting things that one of the folks I interviewed at KubeCon mentioned as one of their favorite things that they learned was in the security keynotes, Nikhita gave one that talked about the risks of quantum computing into computer security in the future. And so that was something that I had not had on my radar or on my bingo card.

And I don't know that that's around the corner, but it will be interesting if someday that is the big hot trend that we're talking about on here is the introduction of quantum computing into our infrastructure realm.

MOFI RAHMAN: I mean, the devices working group probably will be interested in figuring out how to add a quantum-- I don't know, compute interior Kubernetes nodes. That could be fun.

KASLIN FIELDS: That'd be very interesting. She was, of course, talking about, the potential for quantum computing to break encryption.

MOFI RAHMAN: Yeah.

ABDEL SGHIOUAR: Oh, yeah.

MOFI RAHMAN: Yeah of course.

KASLIN FIELDS: Which is a very different area than ever entering the infrastructure realm. But interesting to think about.

ABDEL SGHIOUAR: I want LLMs running on GPUs, running on quantum computing managed by Kubernetes.

KASLIN FIELDS: Such future.

ABDEL SGHIOUAR: That's what I want. Yes. When I see that, I will happily die. It's fine.

[LAUGHING]

Do you want to add to the buzzword, Mofi, anything?

MOFI RAHMAN: I feel like the security risk for quantum-- I don't know if there's a recording of this that I could go back and watch, but I think a few years ago I saw this video of why the quantum computing would be able to just break all encryption is because most of our encryption is based on multiplying really big prime numbers together. And finding the factors of that number is very difficult. But in quantum computing, there is this algorithm called Shor's algorithm that can just find the prime factor of any two numbers very quickly.

So that's why quantum computing would be the risk against encryption, because it can just find the two prime factors. And once you find it, you could basically brute force your solution to any encryption that you could see. When a regular computer figuring out the prime factors takes billions of years. And a big enough quantum computer could do that in minutes.

So I think right now, the biggest quantum computers are still not as powerful enough to be able to handle the situation. But some people are working on it and they will. But I think that the silver lining there is, once we have quantum computers strong enough to break encryption, we're probably going to move on to quantum encryption to begin with. So it's going to take even more powerful computer to break those encryption.

So it's not really all doom and gloom that oh, yeah, the moment quantum computer becomes a thing all passwords are obsolete. But there is also the movement of passwordless. A lot of things are instead of typing in a password, you are doing some other way to verify that you are you. So, yeah. So don't be alarmed. Don't start putting everything under the ground in a bunker. It probably will be OK. It's not a guarantee, but it probably will be OK.

ABDEL SGHIOUAR: Actually, you know what? It reminded me of an interview we did last year with Emily Fox. And I remember very well in the interview, one of the things that was brought up was Emily telling me that in a conference, somebody told Emily, "I cannot believe I'm going to say that, but I think that putting your password on a sticky note under your keyboard is probably more secure than having it online."

[LAUGHING]

So just like following the "don't put everything under the bunker." Yes, but just be prepared just in case. I'm just saying.

MOFI RAHMAN: Yeah, no, you should have-- Everybody should have two factor. Your two factor should not be your phone number. So I'm turning this podcast into a security podcast at this point, but I'll take my speakerphone on the town square and tell people to be more secure, yes. Your messages are not as secure. So use other ways to two-factor. Use authenticator if you can.

ABDEL SGHIOUAR: Yes.

KASLIN FIELDS: And everything we're talking about here with quantum security and quantum computing is, I think, more farther future than anything else we've talked about, at least.

MOFI RAHMAN: Don't use your birthdays as your password. Don't use your children or pet's name as your password. Those are public knowledge. People can figure this out just by talking to you. So don't do those.

ABDEL SGHIOUAR: Definitely.

KASLIN FIELDS: That is still valid.

ABDEL SGHIOUAR: Yeah.

MOFI RAHMAN: Yeah.

KASLIN FIELDS: Social engineering, still a vector for security attacks.

ABDEL SGHIOUAR: Yes. I want you to just pull us a little bit back to the podcast itself, just insert some plugs here. First of all, we are on Bluesky. So please go follow us there.

KASLIN FIELDS: That's been an interesting thing we could talk about.

ABDEL SGHIOUAR: Yes. I think it's been interesting that everybody is moving to Bluesky My number of followers are blowing up. It's quite interesting.

KASLIN FIELDS: Yeah, if you haven't checked out Bluesky yet, I have probably mentioned it before on the podcast, but I am one of the leads who helps to run all of the social media for the open source Kubernetes project, and we created Bluesky accounts. There's one for contributors, there's one for the project at large that's more focused toward end users in the general public. And we also have another one for newsletter, the last week in Kubernetes development newsletter.

And just we created it maybe like 10 or 11 days before KubeCon North America. And the numbers just went way up. And the engagement there is really high. So it's an interesting platform to keep an eye on.

ABDEL SGHIOUAR: Yes. And the other one is, I guess, you have figured that out by now if you're listening, but we do not have the podcast on Google Podcasts anymore because it was deprecated. So we had to move to YouTube. So if you are using YouTube or YouTube Music, you can still listen to the podcast there. And we are going to be doing some more video content going forward, which I'm looking forward to.

KASLIN FIELDS: I'm also very excited.

ABDEL SGHIOUAR: Yeah. Mofi and I have been bothering people at KubeCon North America shooting some content. Well, mostly you've been bothering. I was just recording, so-- Yes. [LAUGHS] So we'll be publishing some video interviews in there.

KASLIN FIELDS: I always find it so intimidating to walk up to attendees at KubeCon and be like, "Hello, would you like to be on the Kubernetes Podcast from Google?" [LAUGHS] So thank you very much to everyone who was willing to be a part of it for KubeCon North America and for any of the events where we do these interviews, and now videos. So if you do see us at events, I will always remind everyone, please do come up to us, find us. We would love to have you on.

ABDEL SGHIOUAR: Yes. And speaking of that, one of the feedback we heard quite a lot, and we are listening, is that people told us want to hear more end users.

KASLIN FIELDS: Yes.

ABDEL SGHIOUAR: In the show.

KASLIN FIELDS: I'm excited about this.

ABDEL SGHIOUAR: And we already actually have a couple of people lined up from few end user customer companies, so we're going to be doing some more end users. I will be very excited to talk to people who are actually using Kubernetes day to day. That will be exciting. So yeah, so we are listening to the feedback. And please keep this feedback coming if you see us. Or you can use our email address, <kubernetespod@google.com>. Or is it kubernetespod or kubernetespodcast? I always forget. I think it'<kubernetespodcast@google.com>om.

KASLIN FIELDS: Yeah, I think so.

ABDEL SGHIOUAR: Yes, yes. It's in the show notes, anyway, you'll find it there.

KASLIN FIELDS: You'll hear it at the end if you stay till the end of the episode. So yeah. All right folks, we are at about 30 minutes. So I think we I don't want to drag this for too long. Maybe we can go quickly around table and you tell me, what have been your highlights of 2024? Let's do it this way. I am putting you on the spot. Mofi.

MOFI RAHMAN: My highlight for 2024. Well, among many things, but I'm going to talk about the most recent ones. Abdel and I got to give a talk at KubeCon North America, and as of yesterday, we got the feedback back from the CNCF from after the talk.

ABDEL SGHIOUAR: 10 out of 10.

MOFI RAHMAN: 10 out of 10. Again, that's not that important. I'm OK with 10 out of 10. But the part that is like that is the best for me is one of the comment basically said both speakers, these guys were helpful and hilarious.

[LAUGHING]

That's highlight moment. I actually made a post on Bluesky saying that needs to go on my epitaph. Like "Mofi, he was helpful and hilarious." That's it.

ABDEL SGHIOUAR: That's how I want to be remembered, right?

MOFI RAHMAN: That's how I want to be remembered. Yeah, helpful and hilarious. That's just a double H, that's me.

ABDEL SGHIOUAR: Yeah. I would add to this just very quickly. Mofi has actually-- you hosted two episodes this year, right? Or more?

MOFI RAHMAN: Two episodes and a couple of episodes I came in as like, a reading out the news. The two episodes I did was on Ray and one of the episode, on the Ray one. We actually talked to Spotify, who is end user company, too. Right? Spotify is an end user?

ABDEL SGHIOUAR: Yeah, they are an end user company. Correct.

MOFI RAHMAN: Yeah. So we talked to them about not directly about Kubernetes but their ML platform that uses Ray on Kubernetes. So that was interesting stuff to see how companies are leveraging Kubernetes.

And part that was really interesting for me is that a lot of folks talk about how complex Kubernetes is and how expensive it is to run and all that, but when you reach a certain level of complexity, then Kubernetes starts becoming easier than any alternative you could have or cheaper than anything managed both for engineering hours spent or the time actually have to spend on building the stuff. So that was an interesting conversation there for me.

KASLIN FIELDS: And just to call out that Mofi and Abdel's talk at KubeCon North America was about Ray. So if you enjoyed the Ray episode or you haven't heard it yet, maybe check out their talk as well, which will give you an introduction to Ray and how to use Ray to build AI platforms on top of Kubernetes.

ABDEL SGHIOUAR: Definitely. What about you, Kaslin, 2024?

KASLIN FIELDS: I was going to go with the boring answer of the 10-year anniversary, but I actually thought of something that I am more excited about. It's not a moment by itself, but I am really excited that I got the chance to mentor folks in the open source Kubernetes community this year. That's something that I always care about and have done for years and years, but I really went hard on it this year, trying to help some folks level up and take on new responsibilities and leadership in the open source community.

And I've trained up a couple of new leads for communications about contributor events. So we used to do the Kubernetes Contributor Summit and we are switching to Maintainer Summit. That's also an exciting thing that's happening, and I hope to send more comms out about that soon. I have some guidelines for contributors on what's happening with that. Feel free to reach out to me on Slack anytime about that.

But I trained up a couple of new leads there and I had five shadows for my comms leadership position with the event this time. And they all did really, really well. So I think folks got something out of that experience. And in the comms team and in SIG ContribEx, we've had a couple of new leads come up. We've got a new person helping with our mentoring subproject. We have one person who we need to get the PR in to make them a SIG ContribEx co-chair.

ABDEL SGHIOUAR: Nice.

KASLIN FIELDS: We also have a number of folks in the comms subproject who have been shadowing for lead roles within that subproject. So there's been a lot of mentoring and shadowing going on. And I'm really excited at the progress folks have made.

ABDEL SGHIOUAR: Awesome, great.

MOFI RAHMAN: So what has been your highlight, Abdel?

ABDEL SGHIOUAR: Oh, where do I get started? Well, I think obviously the talk we did at KubeCon have been one of them. That's because it was my first talk at KubeCon, and my first time being in KubeCon in North America. I've only been in the European version before.

I was very, very happy and humbled that we managed to get some very prominent members of the cloud native community to be on the show. Matt Klein from Envoy. Well, the director of Envoy. Solomon Hykes, the creator of Dagger, ex-Docker.

A bunch of the episodes that Kaslin did with Kelsey with Tim Hockin, Dims. The list of names is too long to mention in one episode, but I think it was very humbling to be able, in a position where we can reach out to these people and they say, yes, we're happy to come on the show. I think that that's a privilege we have with the podcast, for sure.

Besides that, just being able to do like keynoted a couple of KCDs been to five KCDs this year in Europe. That was exciting. Attending quite a lot of conferences, talking to the community, talking to people in cloud native space. I think all in all, it was a pretty good year. And I'm excited about 2025.

KASLIN FIELDS: I think that's a good answer. [LAUGHS]

ABDEL SGHIOUAR: Thank you.

KASLIN FIELDS: I am judging your answers.

ABDEL SGHIOUAR: Yes, yes. I know. I know. I skipped it. All right. I think we should probably just wrap it up here. Thank you for being on the show, folks.

KASLIN FIELDS: And thank you all out there for listening in 2024.

ABDEL SGHIOUAR: And I guess if you are listening to this, happy New Year's and happy holidays and we'll see you in 2025.

KASLIN FIELDS: We'll see you then.

MOFI RAHMAN: Bye.

[LAUGHING]

KASLIN FIELDS: That brings us to the end of another episode. If you enjoyed the show, please help us spread the word and tell a friend. If you have any feedback for us, you can find us on social media @kubernetespod. Or reach us by email at <kubernetespodcast@google.com>

You can also check out the website at kubernetespodcast.com, where you'll find transcripts, show notes, and links to subscribe. Please consider rating us in your podcast player so we can help more people find and enjoy the show. Thanks for listening, and we'll see you next time.

[MUSIC PLAYING]