The Vibe at Cloud Foundry Day

Episode 246 · Mon, 14 Nov 2022 16:56:35 +0100 · 39:32 · SRT · archive.org · all Tanzu Talk

Well, we mentioned, you know, KubeCon, VMware Explorer, which is last week, maybe we'll talk about that as well in a episode soon. And, you know, also our friend, the Tanzu application service based on Cloud Foundry, but I don't think we've actually dug into a little bit of checking in on the Cloud Foundry community in a while. And I was digging through the exciting stuff from KubeCon over there in Detroit, or as some like to say, Detroit, which is fun. And then I realized there was a Cloud Foundry day there, which you had commented on a little bit here and there, Nick. And so I thought we should have you on, back on for the second or third, I don't know, I've lost track time, to just like, it's very open ended. What I'm going to call, what's the vibe at Cloud Foundry day? What's going on over there? What are people working on? Like, just what's happening? What's up with that community? Gotcha. Yeah. Cool. So this is like, obviously the, I guess, first in-person event since, you know, the COVID happened. I was actually at the last in-person event. I think that was in Philadelphia in 2019 as well. So it was good to kind of reconnect with everyone again. But this is the first time I was doing this as working for VMware before I was a customer of, you know, using Cloud Foundry. So that was pretty cool to kind of, you know, see that transition. But it was definitely kind of a tight knit vibe where it was a single track, like single room. I guess, what do we call them day? It was a day zero event at KubeCon, but it was on the second day. So I guess, you know, the math is a little confusing there, but another day zero event of many where, you know, we can kind of get a very kind of nice focus and just really talk Cloud Foundry all day. There's a lot, a lot of talks that we can kind of go into, you know, giving the different projects that the community is working on. Kind of the two big themes are Cloud Foundry Karifi and the Piquetto Build Packs as well.

Those are kind of the two, you know, let's say next evolution or evolving evolution of Cloud Foundry and that community. And it was just really nice to see everyone in person. And then kind of the single room, single track view was pretty nice because you kind of get all the content and you didn't have to worry about shuffling to different rooms or, oh, am I going to make it to that type of thing? So I really enjoyed that. You know, you kind of, you know, got set up and were able to kind of take it all in and not have to worry about, you know, the logistics of traveling to different rooms and that nature. I always like a single track conference. I think that's the best. It's, you know, it's, you know, everyone's got the same breaks. And I think also like, especially if you have a lot of content, it really, uh, make sure most all of the content is good and you don't have, you know, I don't know about y'all, but I've been on a lot of talk selection things. And sometimes you're just like, I don't know, there's a slot over here. So let's put in a talk about like how to like make better slides or, you know, something like that. I mean, you know, they, they, they can be helpful if you have a big conference to have sort of like career and, uh, sort of like, uh, self-development, uh, talks, but it's sometimes also nice just to do the nerd stuff straight through. Yeah. Well, well, let's, let's, let's start with, uh, well, first of all, you know, I, I think, I think, but what, uh, what's always most interesting is as you were saying, is kind of like what the future plans are and, uh, like, uh, you know, getting, getting cloud foundry into the, uh, the Kubernetes world. But like, what was the, uh, like, like give us a sense of like the usage of cloud foundry, like who, who's using it nowadays? Like, uh, if, if, as you were saying people like your former self, you're still your same self, but in your former role, uh, showing up as a user. Or a customer, like what, uh, what, what, what, what are, what are people doing with it nowadays? Yeah. I mean, so I think the use cases are still the same as before. So, you know, cloud foundry, so it's, it's like, you know, kind of niche or like really, uh, awesome use cases for kind of large enterprises or large companies that have, um, really large development shops to really kind of give, uh, give, um,

our customers and people in the community, this really kind of very curated and opinionated stack to just focus on delivery business value. Um, so, you know, a lot of the, you know, the, uh, the companies and customers that have been working on this for the past, uh, almost, I guess, I think it's decade now have, you know, still been kind of, you know, I guess I used to say in my, uh, previous job here, uh, you know, we're just over here running the business and, you know, making all the money for everyone. So, um, that is, you know, very still true today. Um, there was a talk with some of the SAP folks where they were going into some of the scale where they run, uh, cloud foundry, you know, worldwide. And, uh, some of the scale of those foundations they were talking about was, uh, it was just kind of absurd how large they were. Um, let me, I took some screenshots from one of my tweets. So I'm going to try to see if I can find those really quick, but. So, so what is, what does SAP do with it? I, I remember, uh, you know, they were, they were part of the initial, like, uh, big sprout, sprout cloud foundry community. What, what have they, uh, been up to these years? Um, so, I mean, they're still very, uh, active in the community and, you know, they, um, like, you know, uh, a partner or a, uh, they have a lot of developers committing code to cloud foundry itself and developing that. And, uh, they use it as, uh, I think a part of their services, like service offerings that they run, you know, they use cloud foundry to run, uh, business applications for their customers. I don't know the exact, you know, application names and things of that nature, but they, they use cloud foundry to run, you know, their custom applications that then they serve to their, their customers. Well, it, it could be anything from, uh, annual benefit selection to, uh, you know, reordering, uh, ball bearings that were shipped to, uh, to Rotterdam instead of like, I don't know, Sweden. Yeah. And then some of these numbers too, for something like that, you know, they were measuring some of their app run time.

So a lot of the, the cloud foundry run time is measured in, in, in memory, uh, that is allocated to the, the container schedule, which is Diego. So they, they, they had, uh, I think recorded 317 terabytes of Ram allocated, which is, that's a lot of Ram. So that's a lot of, it's a lot of containers running. So we talk about, usually they talk about terabytes in terms of, you know, spinning disc or, you know, cold storage, but, uh, DRAM that's, you know, usually pretty expensive. Uh, I think at one point it was, you know, $10,000 a terabyte or something of that nature. So, uh, to have that many, you have, have that much memory allocated to some of these, uh, deployments is, is pretty significant. Yeah. That'd be a lot of Minecraft servers. Yeah. A lot. Maybe, maybe all the Minecraft servers in the world. I don't know. That might even allow Mac users to run Slack and other things at the same time. Oh, yes. That's now, now, now. Wishful thinking. And Nick, is this the first time that there'd been a conference of this type since the whole, uh, sort of merger with the CNCF and the, and, and that, uh, that style of, of managing the project? Um, yeah, I mean, so this is the first kind of event since COVID happened. I think there might've been a virtual event last year as well. Uh, you know, I wasn't quite sure, but I guess one thing too, to key out is, so it was in person, but it was also virtual so that the single track made it really, um, I think easy for the virtual folks online as well. Cause it was like a, a zoom room where, yeah, the presenters would come up and kind of attach, they had some kind of like server type thing that I think was doing some kind of AV streaming back to zoom or something. But, um, you know, they'd hook up to the, that device. And then, so then they'd present in the room and then be presenting online. And then, um, we had Chris and Ram from the, uh, cloud founding foundation kind of monitoring that. And, you know, if there are virtual questions coming in, kind of, you know, like, uh, interject that to the speaker or like in-person questions, obviously you've raised your hand and that type of thing.

So that seemed to work out pretty well too. Like kind of a good balance of, uh, you know, physical and virtual presence. Uh, I guess, I guess we call it hybrid these days. I don't know. Uh, but yeah. Yeah. I, I gave, I gave a talk at a, uh, a hybrid conference on Friday and I was the hybrid. I don't know. I w I was, I was presenting remotely and, and, and. And, uh, the, the, the feed that I got was from the back of the room so you could see people, but then you could also see my screen and they had positioned my head as like a third of the screen with my slides behind me. And so it was kind of odd to, uh, see my big head there, but it did, uh, it did make it seem a little more real as a speaker. That, that was enjoyable. Uh, well, so, so what is the, uh, well, let, let's start with, you know, the thing I think over recent years and, and, uh, at the moment that it's of most interest. And that is, so what's, what's up, uh, what's the state of, of getting cloud foundry in, in, uh, to have a foot in the Kubernetes world? What's, what's going on there? Yeah. So the, the, the, the, the latest, uh, iteration there is, is called cloud foundry Karifi. Uh, and that's the, I guess the rebranded or like the latest name of the previous, uh, iteration of that work, which was called cloud foundry on Kubernetes. And then, you know, uh, previous efforts before that were, um, uh, attempted as well. And, you know, with different kinds of mechanisms, but the, the latest work, which is Karifi, uh, is really targeting, um, uh, recreating, you know, those kinds of cloud native interface or sorry. Cloud foundry interfaces natively within Kubernetes using things like custom resource definitions, instead of trying to shim some of the cloud foundry components, like in a container and just run it in Kubernetes. That was kind of not necessarily the best architectural decision or, you know, it was going to have some significant scale limitations. So, uh, I get the team has, uh, has been working out for at least the past year, probably longer as well to kind of revamp all that and, um, kind of give you, give you that, that CF push, uh, on Kubernetes, but, you know, natively without trying to, you know, uh, you know, shim things in or shove things into containers that really shouldn't be, uh, that type of thing.

So that was, uh, you know, a big thread and a big theme for the day. There is, uh, a demo by, uh, Dr. Dave Walter and Andrew Ritchrock, uh, who, uh, kind of give it, gave it an overall introduction to Karifi and they actually showed, um, installing Karifi, Karifi locally on a Kubernetes cluster and then, um, pushing an application and using that, you know, uh, cloud foundry, uh, user credit, user credit service, service binding, binding it to a database. So that, that kind of high level, uh, basic, um, cloud foundry experience that we all know and love, and it's super simple to get started, you know, recreated within, uh, you know, on Kubernetes and in, in that landscape. So, you know, you know, something I I've been wondering, uh, recently, uh, you know, when, when I'm not trying to figure out why the kids can't like get their shoes on time to get out the door in the morning or, or other real life things like that is. So, so like when people talk about like having cloud foundry run on Kubernetes and you were kind of touching on this today, like, like, it seems to me there's two possible things that could mean one would be like, let's take the cloud foundry code and run it on Kubernetes. Right. And you know, obviously it's not that simple, but you would have the whole system of doing things. And I guess you would swap out the container management. I mean, it gets a little complicated here because basically cloud foundry has its own Kubernetes built into it, but I guess you would swap it out in the same way that like, if you had a, uh, a car and you were like, I want a different type of engine. Uh, I'm tired of these like measuring bolts and nuts and inches I'd like to measure them in centimeters. So you want to swap your engine out or something, uh, which is totally an equivalency between those two other systems as far as I understand it. Uh, but then it also seems like also, you know, something you're saying is like, you could do the, uh, I don't think there's a word for this in systems, but you could do like duck typing. Where it's like, well, I guess this would be the, the cloud foundry Turing test. You could be like, as far as I know, how will I use it?

It's exactly like using cloud foundry. I type in the stuff, whether I'm an operations person or a developer, and it does all the things I would expect. And like, so like, like what w w what's up with the cloud foundry community as far as like those two or a third or a fourth option that, that they're looking at to, to kind of like bring these things together. Right. So the, from the, you know, if we look at the different personas of the developer using cloud foundry or, and, or like the operator managing cloud foundry, right. So from a developer's point of view, Karifi aims to really kind of, you know, make it, I guess, you know, the same or really hard to tell the difference. So their existing, um, applications, their existing pipelines and workflows don't really have to change that whole lot. You know, they'll have to target a new API, uh, endpoint cause it's, you know, running in a different location, but that's really all they'll potentially really have to change. So that's kind of the whole, you know, having that CF, you know, API compatibility from the CFCLI and the API layer. So having that not really change or really having, um, it's really having Karifi, uh, you know, adhere to and emulate all those existing interfaces. Uh, that's, so that's, that's the, from the developer point of view, not really having to change a whole lot now from the operator's point of view, um, you're going to have this concept really classic cloud foundry, which is based on Bosch. And that, you know, Diego and all these, you know, all these different, you know, kind of, um, components that, you know, were created and maintained outside of the Kubernetes ecosystem. And then with Karifi, you really land on an existing, uh, Kubernetes deployment, and then that you can make use of all the different kinds of components within the CNCF. Right. Uh, without, and obviously too, from an operator perspective, uh, you know, not necessarily learning how to, you know, learning how to use Bosch and all of that. You could just, you know, if you already know Kubernetes, you're kind of, kind of already in that zone type, uh, so to speak. Um, so, you know, they're, uh, uh, using helm to install Karifi and kind of wire that up into existing, um, Kubernetes clusters with some, some other prerequisites, uh, from the kind of CNCF type landscape.

Um, so that's kind of like the two different, um, personas and delineation where the developer experience is trying to not really be, um, changed whatsoever or very little changed. But then from the operator persona, there's obviously going to be some significant change. Yeah. Okay. That, that, that makes a lot of sense. Cause from, from, for a developer, well, I guess for a developer, you could just have some translation layer. That's like, you know, I, I don't remember my cloud foundry very well, but like in cloud foundry, the way that, you know, you define routes, which are like the way two parts talk with each other or your ingress and egress, which is always a fun word, just words to say. And so I forget what the, the, the word you use for that in Kubernetes is, but you would just have some translation thing that when a developer like does, you know, CF make me a route that has this name that connects to my application. You would just translate that to whatever the Kubernetes words are and the structure, and it would do it for you. And then I guess, I guess the only like dissonance you would have from that is if the developer then went to some like management or monitoring or whatever you want to call it, observability tool. And like the vocabulary was kind of different and they would have to translate in their mind, like, like what was going on there, but you know, whatever you would assume that would be part of what, what was delivered as well, or it would be a nice thing to deliver. So the like developer thing is pretty easy because it's just like translating stuff around. But then I guess where it gets difficult is from the operation side, if you wanted to customize anything or introduce anything in, in Cloud Foundry, you do that with Bosch and you don't in Kubernetes. So you would have to, if you wanted to customize things on your Kubernetes side, you would, you know, whether you're bringing new types of, you know, something to handle terabytes of whatever, or your, you know, data stores or various types of, to use the Cloud Foundry word backing services for it. Like a lot of that stuff comes down to like, if, if it's not out of the box, you've got to use, you know, Bosch and other stuff to add that in there, which as the operations person, you wouldn't do that.

You do it as a, as a Kubernetes thing. And then, and then again, they would also have the thing of whatever vocabulary they, they use, you'd have to map between the two different domains. I don't know. Is that a fair summary? Yeah. So the, so the, yeah, so Karifi is doing that mapping for, for the developers and somewhat for the operators as well, where these, these custom resource definitions that are getting defined or, you know, they're called CF app or CF route or CF domain. So these, these kind of CF command and, and API constructs are getting represented that as objects in Kubernetes that you can interact with, like with, like, so you can do like kubectl get CF app. And that shows you your Cloud Foundry apps have, have been deployed, but then you can, you know, take that a step further and interact with them as like, you know, as a pod. So you can do like kubectl get pods, you can, you can take it even further if you want to, but, and so it's kind of, you know, multiple layers that you want to interact with, but the developer will still probably, you know, you know, use the CF CLI for the most part. Yeah. And, you know, maybe break out to the Kubernetes, uh, kubectl or kubectl, however you want to say it. Um, if they need to, that type of thing. Uh, and from that, that route perspective that you were, uh, that mentioning. So like, you know, you, you see a map route, you get a route from the, from classic Cloud Foundry, you get that go router, you know, setting up that route for you and, uh, TLS termination and all that fun stuff. Well, that's taken care of, um, in on w with a project called contour, which is an ingress controller in Kubernetes and that, and using, um, some other things like cert manager and things to manage certificates. So that's kind of like mirroring the kind of the classic, uh, Cloud Foundry go router experience to the Kubernetes world using ingress controllers and Kubernetes services and that type of thing. And that, you know, that, um, that CF route is also a custom resource definition that you can interact with as well on the, uh, kubectl side of the house.

It sounds, it sounds better than sort of some of the past attempts then, because some of those kind of fell down in between the cracks, didn't they? They sort of, you know, developer will get so far and then have to rely on kubectl. Same for, um, you know, if you were used to certain constructs, like, uh, you were saying there, such as roots or whatever, you would have to figure out which components in the Kubernetes world you needed to use in order to do that. So the workflow would be more difficult. So it sounds like what's happening here is that the, the workload, the workflows that CF users are already familiar with are becoming, um, sort of first tier supported features in Karifi. Whereas before it was, uh, you know, a little bit more messy. Is that fair? Oh yeah. I think it's a pretty fair statement for sure. And, and definitely too, from an operational perspective as well, that's, that's very much true as well. So what was the sort of reaction to it? How were people, you know, were folks being introduced to it for the first time or is it, is it something that they're already comfortable with and that they're looking for, you know, maturity on what, what's the sort of feel around it? So I think there's a lot of excitement around it. I think, you know, obviously the engineers who've been working on it, we're very familiar with it, but there's a lot of folks who, um, haven't, hadn't seen it ever. So this is their first time being exposed to it. So that, you know, folks are pretty excited about it. And, you know, there's a lot of questions and, um, you know, doing, uh, that, you know, they did a really, uh, in depth live demo. So like, you know, those, like people were excited about that, seeing everything that was going on. Um, so that was, you know, that's very, you know, I think it was, uh, you know, probably, um, one of the most exciting in terms of net new technology sessions that were, you know, talked about. Um, so people really, um, I think reacting pretty well to it. Um, I'm obviously really, you know, excited about it as well, because I think it gives a lot of, um, options and, and, uh, you know, uh, it really expands the, the, uh, existing, um, ecosystem more, um, going forward.

Um, so, and I guess one thing too, I want to also note is that, so all the, all the work, the Cloud Foundry, Karifi work, um, from the, the VMware Tanzu side of the house. Uh, we just released the application service adapter 1.0, which is based on Cloud Foundry, Karifi, uh, and ties into our, uh, VMware Tanzu application platform. Uh, so that, that released, uh, last week. Um, so that's kind of like our first commercialization of Cloud Foundry, Karifi as well. So definitely keep your, uh, your eyes peeled for that as well, or, uh, keep your eyes out for that. Uh, but there's a blog and, and some, uh, I did a demo on, you know, kind of the whole, the kind of whole life cycle as well. That you can go check out and see, um, you know, the application service adapter or really, you know, Cloud Foundry Karifi in action as well from, from the VMware Tanzu side of the house. Hey, Michael, I think, uh, you might be on mute there. I noticed that you had some, uh, some videos there and, uh, I, you know, I wanted to say, I really liked the aesthetic you were choosing. They're kind of like maybe early nineties, uh, mall video arcade. Yeah. I think, I think it was looking pretty good. You know, where you can go to the mall. Did you all have this over there? Uh, you know, in, in your neck of the woods, uh, Ben, where you could go to the mall and they had a recording studio and you could, you and your friends could record yourself singing the hits of the day. I don't know if I noticed that in my hometown, but, uh, you know, uh, maybe my hometown wasn't bang up to date in the eighties nineties, but, uh, I mean, I feel, I feel at the time you could play video games and you could also go record yourself. You go, go to the bookstore, maybe the skate rink food court. Yeah. Real stranger things stuff. Yeah. Video games for sure. Yeah. Our version of, uh, Las Vegas is black pool, which was very close to, uh, to where I grew up. So I would in the summertime, go down to black pool and play all the arcades all the way around, along the seafront, all the way down to the bottom and then back again, uh, to where the train station was.

So many misspent hour in, uh, in various arcades as a youngster. Yeah, for sure. And I, yeah, I liked the video, Nick. I liked the, uh, the aesthetic you've gone for there. Yeah. I'm surprised you didn't do yourself as some sort of max headroom though, just to finish it all off. Oh, well, I'm still learning. Uh, my, my, my first time. Yeah. Yeah. Just need some papilla mash. Sure. Yeah. Yeah. Uh, well, so, so another thing you mentioned is, is, uh, is, is build pack stuff, which like, you know, I, I think, you know, to ask my general high level question in this area, it seems like, um, you know, like, it seems like there's a lot that build packs and cloud native build packs have to offer everyone, regardless of what you do with cloud foundry. And, uh, I don't, I, I mean, I feel like in, in the Kubernetes world, like they have, I mean, they, they, I always forget. Cause K native keeps like a bundling and unbundling, but at some point, and maybe still like build packs were part of K native or not. But anyways, it see, it seems like a good general way to like package up and deploy your stuff. Uh, you know, kind of in like a. DevOps sort of way. That's a bunch of gobbledygook words, but what's, what's, what's up with the, uh, with the world of build packs at the moment in cloud foundry. Yes, sure. So from a build pack perspective, I, a new thing that was, uh, we, you know, was discussed was the, uh, the Piquetto build pack project, uh, which is an implementation of cloud native build packs. So when, when we talk about build packs, we have like build packs V two, which is essentially the, uh, you know, the cloud foundry classic, uh, you know, what we know today in cloud foundry build packs. And then the cloud, uh, the, you know, the V V three build packs are really the cloud native build packs. And, um, the Piquetto is an implementation of those build packs. So Carifi was using, uh, the Piquetto build packs as well. Um, we use some of those build packs and what we call Tanzu build service, which is bundled in Tanzu application platform.

Uh, but yes, the, the concept of build packs could be used. I mean, really like everyone, in my opinion, should be using build packs. Um, because it's this really this way to, to, you know, get, uh, you know, best practice industry standard for your runtimes and having people fiddle around with containers. Uh, build them kind of on the fly, you know, like why would, why would you do that when something could just do it for you? Right. I know it's crazy, isn't it? Yeah. And then also having, yeah, this, all the security leaders taken for you as well, you know, just out of the gate, like not having to worry about any of that. I mean, I don't know. Like if, you know, if I was, you know, still writing code, I would not want to be like worrying about if my Docker container secure, that type of thing. If you, if you went up to a team and said, would you like a thousand hours of DevOps for free? They would probably say yes. Yeah. And yet surprisingly, when you say you can use a build pack, you know, folks kind of, you know, need, need convincing. And it's like, well, there's thousands of hours of really good research and development gone into these things. Just use them for goodness sake. They're built into spring. Now use spring or Maven, for example, you can just do build dash image. It will build you an image. It will build you an image. It's going to use the same build packs that you're talking about there. Right. So, uh, it's a, it is a source of irritation. You know, I, I think, I, I think it's too late now. I think they might've closed this, but that maybe next year at, at Fosdim, we should propose one of those day long tracks. It's just about like builds and you could have, you could have build packs as one thing and, you know, building stuff by hand as another thing. I think, uh, you know, Maven's still around, right? Like you could like all these things that are basically like, how do we build stuff? Not, not like CICD pipelines and not like runtime environments, just strictly like we're just building stuff. And that there must be a lot going on in the, uh, the open source world, uh, of doing that. We should have a race. We should have like two teams or something. Maybe five different workloads in different languages and just say, okay, you've got a day to build these into.

Oh, and yeah, that, that's, that would be fun. And maybe we could also, the other competition could be. So if you have five or six build systems, the, the, the competition is, can you build them all? Like how, how can you build the build systems and see what happens? And that's just funny, I guess. And recursive. And recursive because it would be, it would be funny if, if building the builder was difficult, which it often is. It seems, uh, well, that was a distraction from, uh, from, from build tech stuff. But yeah, if you, if you just use build packs people, come on, it's 2022. Uh, stop reinventing the wheel. Just use, just use the build packs. Yeah. Use, use the build packs in your life. We'll, we'll be happy. And you can sleep more and not have to worry about security and that type of thing. Or worry, or worry less. Let's just say about security. Um, yeah. Yeah. So what else do we want to talk about here? So we've talked about Karifi and Paquetto. And those are kind of the two, um, really new things. Like, you know, the new kind of exciting projects going on behind the scenes, I guess. Um, well, what was like the, uh, the future looking thing? Was there some roadmap talk? You know, you get the community people together. What, uh, what, what, what's kind of the plans for the next six and then like 12 months? What, uh, what's people rumbling on about? Um, I don't know if there was like a full on roadmap talk per se, but I mean, obviously continuing development within Karifi and Paquetto are, are, uh, you know, kind of going forward for the future. And then also, you know, just continuing to ship, um, you know, the classic cloud foundry experience as well as, as a big theme. Uh, there's a lot of different talks about different sub components of that ecosystem. Um, going forward and, you know, there's talk about, you know, kind of the future of logging and metrics, how we're gonna, you know, you know, what's the current state and, you know, how, how we'll progress during, you know, going forward.

Um, there's some, um, app runtime talk as well. So, I mean, I can't cover every video and every topic in this, but all of these videos now are out. If you go to the cloud foundry, uh, YouTube, uh, channel and you can see like the fall talk. So definitely check those out as well. Um, let's see, let's see if we can key up. There's a very interesting talk, um, by David Stevenson, one of our kind of tech leads here at VMware, but he was, um, he did a lot of work about, you know, kind of measuring, uh, the cost to run, uh, cloud foundry across the, you know, the three different hyperscalers and, uh, and kind of revamping some work, um, to look at all the different, you know, VM instant sizing, because, you know, if you've ever used these, you know, the three main cloud providers, the VM sizing is, it's kind of strange and obscure. And, uh, the naming doesn't really make a whole lot of sense. And some of the pricing is also very strange. And as new generations of VM types are allowed, they're typically cheaper than the old types of VM generations. And, you know, you get locked in and it's like, it's kind of, uh, you know, uh, uh, almost a thing you need to decode to figure out, uh, what's really going on behind the scenes. But he did a lot of work in investigating those. And, um, you know, his talk was entitled how we brought down the cost of a new deployment down 50%. So I think that's definitely a key for folks running, um, cloud foundry on a, you know, big hyperscaler of choice to go check out. Um, but, uh, you know, it was all about measuring the different kind of VM types within each hyperscaler and optimizing, you know, each cloud foundry component for potentially a new resource type and, uh, something to, you know, definitely check out there. If you're, uh, you know, have cloud foundry deployments running, uh, out on a hyperscaler. Well, 50% is a lot more than 0% or negative five 50%. Yes. I guess, I guess in this case going, well, that would be positive 50% cause you're really getting negative 50%.

But you're, you're, you're all at saving every month for sure. That's right. Well, you know, I will, I'll, I'll, I'll have to put a link in the show notes at tanzutalk.com to the, to video list. As you're saying, you know, the, the, the CNCF was very quick to upload everything, which was just peachy and, and fine. Is there, is there any, uh, any other, like, uh, you know, as the last thing, how are the snacks? Was, was, was the food good? How would you rate that? Did they have, did they have the mid morning snack? And then usually in the afternoon you have two snacks and you know, there's always a debate. Are we going to put like sweet stuff out there or are we going to put healthy stuff? And then you end up having both. And, uh, you know, you got some fruit, apples, things like that. Was there any savory things, maybe a hummus spread on some sort of like stiff brown cracker? Yeah. Oh yeah. There's definitely a wide variety of snacks and they were, uh, they were a plenty. Um, I think for breakfast, I got kind of a, uh, strawberry. I tried, I tried to eat healthy for breakfast. It was a. Good job. Good job. A strawberry oatmeal type mixture. Oh yeah. It was, it was actually pretty good. Um, you know. Was it served in a, in a little Mason jar that, that you popped open? Uh, it was, I mean, it was all plastic, but yeah. Uh, sure. I don't know. Not quite a Mason jar, but yeah, you had a, had a little bit of a granola that you kind of like popped in and mixed it up, you know, type situation. Uh, so yeah, that was, that was pretty good. And then, um, they, you know, the lunch was in the, the, the main hall, so to speak. Uh, and I would think that definitely about, um, uh, cube con in Detroit, the, the, the conference food was, was, was pretty on point. They had a lot of different offerings and a lot of, uh, I guess maybe comfort food type things as well. But, um, yeah, there's, uh, no shortage of, of, uh, food there for sure. And obviously, uh, Detroit style pizza was definitely a must. And I think, uh, buddy's pizza, uh, I ate and it was, it was, it was very good. So wait, what, wait, what, what is Detroit style pizza? Is it different from regular pizza?

I don't get it. They're good. Yes. Yeah. So like, well, you know, you start talking about Detroit style pizza versus Chicago style pizza, but Detroit style pizza is a very kind of a, a square deep dish with, um, kind of the corners, uh, you know, kind of with crusted cheese type thing. So it's like a soft deep dish, but yeah, it's not like a Chicago style, which is literally like a pie and you know, like very heavy type thing. But I see. So you make it square. Yes. It's square. And so, and I guess maybe the, the, if you've ever had little Caesars, it's not quite little Caesars, but little Caesars, I didn't know this is headquartered in Detroit and they have a square, square style pizza as well. It's not like their main style, but, um, I see. That's what they had when I was a kid in those, uh, you know, wishing I could had money to go record myself singing the hits of today. Yeah. Yeah. The arcade. Yeah. Lots of little Caesars. That was some cheap pizza, like three bucks for a gigantic sheet of pizza. Yeah. And it tasted like that too. Yeah. Although I have to say that I, I have fond memories of, of that flavor. It's kind of like, you know, uh, every, every expat loves Kraft macaroni and cheese. And, and what I, what I've come to recently with this is, is it's kind of like also stovetop stuffing. You're like the, the flaw in not liking Kraft and macaroni cheese or stovetop stuffing is there is no comparison. It's its own food item. So you have to appreciate it on its own. There's no like, you're not like, Oh, but what if you made a bechamel with your own macaroni and cheese fresh from the farmer's market? And you're like, no, that's not what I'm making. That's not what, that's not what this is. Yeah. This is a different thing. Yeah. I think, I think, yes, I think, I think little Caesar's pizza and you know, Domino's, all those pizzas are also their own thing. They're definitely not the kind of stuff you get in Rome, which is its own thing, much separate than anything we would have. And can we revisit very briefly this 1980s arcade game playing persona? What was the beard like back then?

And the hair? Question to all. So I'll start off mine. I wouldn't have had a beard. Absolutely not. But I did have like really short, spiky hair. I reckon back then. Oh, yeah. So please go ahead. I think, I think my hair was always a mess. I was not really into, into taking care of myself, I think. And no beard, I guess. No, no, that, that didn't happen until I was like, I don't know, 13 or 14. So it took a while. Or like this, like it is now. Something like that. I must've been off putting for so many people. Yeah. Yeah. 14. Full beard. Hair. I don't know. Hair back then. That was a little, you know, I, I was born in the 80s, but it was late 80s. So I don't, I was not. It was a mullet, wasn't it? Yeah. It was a mullet and you don't want to tell us. No. I don't know. I mean, I, I haven't had hair in a while. So, you know, you could tell about that. The shine off my lights. Right. Business at the front, party at the back. Was that it? Was that it? I don't think I've ever had a mullet, but no. I think I was, I was born, I guess, with somewhat, somewhat of a mohawk, I guess. A little bit of the hair on the top, but. Oh, that's strange. I don't know if I go digging through a baby picture to find it or whatever, but yeah. All right. Well, there's an image, dear listeners. Yeah. You'll never be able to erase. Yeah. What we might've all looked like. Maybe, maybe, maybe we can each send a picture of ourselves at the arcade and I'll use that for the, the album art, the cover art for this episode. I don't, I mean, I'm old enough that they're not digital, right? Oh. Yeah. Well, me too. You'd have to scan them in. Yeah. Well, okay. So, so that, that was a good, that's good. I feel nice and informed. That was, I appreciate you coming on here to, to, to give us a little overview of what's going on there and a little overview of the food. I always love a yogurt with a granola mezzanine. That's, that's always nice. Make something on your own there.

And as always, if someone wanted to, if, if they wanted to tell you about their hairstyles when they were a kid, Nick, where, where, where would you point them at? Where could they go? You can always hit me, hit me up on Twitter. I guess we'll, we'll see how long. I don't know. I don't know. It's kind of like, yeah, hit me up on Twitter. We'll see if it's still running in a week. I don't know. Yeah. Yeah. Yeah. You can also find me on LinkedIn as well. I guess we could put some links in there. I am trying this whole Mastodon thing out as well. I guess we could try that too, but. As is Michael. You've got, you've, you're trying a bit of Mastodon too, aren't you? Well, I learned my lesson back before I had a beard that when there's a new social site to go squat on it and try hard just, just in case things happen. You know, maybe that's what someone out in the spring community written a Mastodon server yet. I don't even know what that means, but that seems like something that, that, that should happen. If it has been, you can bet it's running on cloud foundry. Yeah. Yeah. Oh, very good. Well, and with that, if you want to get the show notes for this episode, I'll, I'll try to, uh, put in links to the, the cloud foundry day stuff. And, uh, if I, we can find any pictures of ourselves at, at arcades or not. Uh, but you can go to tonsutalk.com and just find this episode, probably unless you're listening to the entire back catalog sometime in the future, it'll be right there at the top. So you don't have to worry about it when you go to tonsutalk.com and find it. And it's always a great if you enjoy this, or even if you don't enjoy it to subscribe to it. And what I always recommend is anytime you encounter some device that is able to download podcasts, whether it's your children's, your enemies, uh, whoever delivered your mail, maybe you sent the butcher off to get you a good cut of steak, or you want some change for an Xbox that you're about to wrap some towels around, uh, grab their device and pointed at their face as they walk off and, uh, subscribed into this podcast two to five times and, uh, give them a little treat that they'll enjoy. And with that, we'll see everyone next time. Goodbye.

Bye. All right. I'll end our broadcast here. That was fun.

Tanzu Talk · 🍕 Get more pizza.