Tanzu Application Service 3.0 Overview, with Nick Kuhn

Episode 240 · Fri, 07 Oct 2022 08:40:09 +0200 · 49:18 · SRT · archive.org · all Tanzu Talk

This, this episode, we have a topical episode about the Tanzu application service 3.0 recently released. And we have, we've got a, we've got, we've got a nice guest lined up who actually knows what they're talking about in detail. They've organized their thoughts. They figured out the best way to present them for maximum knowledge transfer. Why don't you introduce yourself guest? Uh, hi, I'm Nick Kuhn. I'm a technical marketing architect, uh, working on Tanzu application service and, uh, this new product as well, the application service adapter for Tanzu application platform. Ah, yes. That that's that now is that, is that in the 3.0 release or is that still in, in sort of beta or whatever release? I forget. Uh, the, uh, the adapter is still in beta, but, uh, keep your, keep your eyes peeled for, uh, when it goes generally available here soon. Yeah. Yeah. Okay. That sounds good. Well, so before we get into, I mean, this is the, the, the 3.0, which would imply it's the third major release of the application service here, but you know, for, for not everyone, unfortunately, you know, for, for their own, uh, excitement in entertainment is familiar with the Tanzu application service, previously known as pivotal cloud foundry. Why don't, why don't you give us, uh, like, like, like what is, as we say here inside Taz, like what's, what's going on with this thing? Sure. So like at a very high level, you can think of, uh, Taz application service as a, uh, what we call a platform as a service or kind of like a application runtime. Uh, so that's kind of a, a higher level abstraction than what you see with maybe like a Kubernetes, which is most people call it a container container as a service container runtime. Uh, so it's kind of higher up the stack. Um, as you, uh, kind of mentioned before this, this is the artist formerly known as pivotal cloud foundry or pivotal application service. Uh, a lot of people still call it PCF just cause that's kind of what stuck with them. Um, but it is a distribution or release of cloud foundry and, uh, that it, you know,

there is a, something called the cloud foundry foundation and there are multiple kind of certified distributions produced by various companies, uh, of this, uh, you know, software suite, um, or, you know, kind of foundation of software. Uh, but, you know, it's all kind of geared towards the, uh, you know, elevating the developer experience, uh, and really kind of, um, abstracting developers away from, uh, infrastructure toil and, and, and, and fiddling with all the different bits that they can kind of get lost in, uh, you know, this day and age, especially with, you know, if you look at kind of like building your own Kubernetes platform and, you know, exploring the, uh, the CNCF and all the, you know, all the different, you know, ways you can get lost for, uh, just figuring out how to get ingress or certificates or things of that nature. Tell me, uh, how, what you think of this, Ben, but I think, I think that, that, uh, uh, that the Tanzu application services, I don't know what you, what you call like a classic paths that, that is like, it, it essentially tries to, and, and does pretty well removing from application developers, the need to really do very much infrastructure thinking, right? Like all of the sort of infrastructure is, there are still things there to do, like, um, I don't know, explain how to connect together different components of your applications sort of command that it gets done, but you don't really have to like do it. You just deem it so like, like you would sort of expect from a, a, a self-service platform. Yeah, it's really, really powerful. I mean, um, first of all, it's very mature and very trustworthy. So, you know, exactly what's going to happen when you're issuing your commands against, uh, Tanzu application service. And, um, Nick and I, we, we've both been on the other side of the fence, if you like. So, so I, I was in the field as a consultant, helping folks to learn how to use, um, Tanzu

application service and, um, in many large organizations, you know, folks like, uh, like banks and automotive companies and that sort of thing. Um, and they really, really get it. You know, they really dig the fact that, yeah, a developer can have a lot of control over exactly where their application is going to run. And I know for, for Nick, you came from a, uh, a client of, um, Pivotal Cloud Foundry, didn't you? And you, you came to us from there. So I'm presuming you had a similar experience with the folks that you're working with. Uh, yeah. So, um, I, I, I worked at a large retailer previously, um, uh, before coming to VMware. But, um, you know, with this kind of, first of all, from the developer experience that we talk about that, that command and, you know, that's really this, this command called CF push where it focuses on going from source code to production or source code to URL, uh, you know, within minutes. And kind of before that, what folks saw, if we'd go back in time, kind of before, um, the, the Kubernetes landscape, you had to go like fight with all these legacy kind of infrastructure application service servers and, you know, fight to get, you know, your code pushed to a server and, uh, you know, work with networking teams to get load balancing and certificates and things of all this nature. And with CF push that enabled developers just to take that code and the platform would upload it. Uh, it would, you know, put it in a runtime, a container, you know, um, back in the day when cloud foundry came out, that was kind of a newer, newer thing for most people. Uh, but it automatically built a container for you based on best practices called, uh, build packs. Um, and we kind of see that, that build pack concept carried through today through multiple technologies, uh, and that's kind of was all handled on, on behalf of the developer. They didn't have to worry about doing any of that. And then it gets promoted up the chain and, you know, gets into the routing tier where,

um, load balancing gets set up, um, auto scaling gets set up, uh, certificates and all of that is just handled on the, by the, by the platform for the developer. And they don't have to think about any of that. And that's a very common, um, it's a great story. You know, if we look back a few years and we look currently too, that's, you know, it's still kind of hard to do even with, um, some of the newer technologies as well out of the box. So it's very powerful from a developer perspective, but also from an operation perspective, uh, there's this little, uh, handy infrastructure, uh, as code technology called Bosch, uh, which really enables, uh, you know, customers or clients of a team's application service to upskill their operations game because then, uh, Bosch will, uh, automatically deploy Cloud Foundry, uh, on pretty much any infrastructure as a service as choice, whether that's vSphere, AWS, Google, or Azure. And so the, the operations team don't really have to, um, worry about configuring that or building a bunch of, you know, bespoke Terraform to get their platform built. Uh, so that's all kind of handled, uh, by default by, by Bosch. And once, you know, that platform gets built, it will also take care of it from like a day one to day two to day 99 or, uh, day a thousand where it will monitor all these components that it builds. And if there's an upgrade or a CV, like a CVE remediation, you can go and do a full like OS rebuild or a rebuild of that specific item. And Bosch will do that whole rebuild and repaid process, um, kind of out of the box. And also in a, in a manner that does not take the platform down and keeps it up and running the whole time. So, uh, at my previous employer, we, we rebuilt our platforms at least four times a year, if not more, um, from like, you know, that golden, uh, image that we had with Bosch called the stem cell. So it's, uh, it's very powerful from both the developer and operator perspective of really

kind of up, up leveling your developers and operators to, to just reduce a lot of complexity and, uh, really get folks, uh, focusing on what's really matters is, you know, how can we make our business logic process, you know, how, how do we produce more business logic faster to, to get features out to our, our actual end customers? Yeah. You know, I always think like, if always like, like I'm, I'm walking around while I'm cutting up garlic and things and always thinking here, but when it does come into my mind, uh, like I think, I think, uh, you know, the point of the tons of application services that if, if you don't need to, or want to build your own platform, here is one you can use that, that is all integrated together. And of course, uh, as, as, as one of our coworkers, uh, former coworkers used to say, you know, you've got to install the cloud. So there's some work that has to be done to get the infrastructure in place. But with that, with that baseline of work, as you were going through, you know, like you, you kind of set up the way networking is done, the way storage is done, the way middleware is managed and scaled up and down and, uh, and provisioned out to people, so to speak, how quotas are, uh, done, like all, all the kinds of controls you would expect. And, uh, it's just sort of ready there with defaults, uh, to use if, if you don't feel the need or, uh, or desire to customize things and build it on your own or, or use, you know, one of the parallel, uh, the, the other sort of platformy thing we have is the Tanzu application platform, which is more of a, uh, uh, an integrated toolkit for building that platform on your own. If, uh, if you're into that rather than having this, you know, black box is always a negative term, but this, uh, this, this functional uni unit, is that, can we see if that one works, Ben? What do you think uni unit? Is that, is that going to fly? It sounds like part of an educational establishment. Ah, yes. Yeah, there you go. Uni unit. I'm going to stop saying that word because I'm, I'm not sure that's a little too science fiction for me. I feel like I've stumbled in some sort of

like Tumblr aesthetic vlog that's, uh, going over things. I'm not quite sure there. Well, uh, so there's, there's, you know, uh, a pretty, uh, large install for, uh, for Taz out there. So I'm sure people are, you know, they, they kind of watched the, the versions coming out and, and the new features there. But so it's always interesting to see what's in a new release. And, uh, so we have, we have, uh, the tons of application 3.0 out now. Now, you know, tell me how you want to start. This is all sorts of ways to go over the release notes. As it were, we got, you can, you can go item by item. Maybe there's a big theme. There's also like the, uh, the kind of commercial side that that's always interesting. What, what people are doing with it. What's what, when, when you, I'm sure you have spent a fair amount of time figuring out the perfect way to represent this material. What's the part where you're like, I got to go edit this and I'm kind of looking forward to it. What, what part are you excited about? Let's start with that. Um, let's see here. So I think every theme so far, uh, we focus on rolling out new developer features and operator features. Um, so those are two, uh, kind of like two different, um, areas that I kind of focus on, you know, uh, obviously trying to get all the developer features and kind of read about them and, and, and, you know, trying to figure out what I'm going to, you know, maybe potentially demo or that type of thing. And then the operator features too, uh, they don't quite demo as well because, you know, it's hard to fit them in a short video, but they have a lot of impact as well. So if we want to kind of go through the developer features first and maybe then like kind of go, we can dive to the operator features. Yeah. Uh, as well as it would work out. So I think, you know, having actually done, you know, worked with TAS in the past, you're of course, most excited about the working with TAS part, the actual features, which is good. And it's true. You know, it can be difficult to show operation stuff because it's kind of like, it's kind of like maybe showing off like a robot, like vacuum cleaner. You're just sort of like, that's what it does. And it vacuums. And, and like, you know, what, what you don't see is someone not having to

spend all their time vacuuming on their own. It just automates stuff for you. And, uh, just, you just get a clean floor all the time. Exactly. So unless you, unless you had like a five-year time-lapse video, you wouldn't really appreciate, uh, what was going on there. But anyhow, how about the developer stuff? What's, what's in 3.0 for that? The developers will get excited about. Uh, let's see here. Uh, so the first feature we can talk about, um, so there's been some, uh, improvements in apps manager and that's kind of the, uh, GUI interface for the platform. And one kind of big thing that's kind of, it's kind of dev and operator-y, uh, I guess, but, uh, apps manager now lets you search for like specific users or really has a user search within the platform. Uh, so that's kind of big when you have these big companies that have, you know, potentially thousands of developers in, in, in the platform. And if you're trying to, you know, uh, like kind of space administrators or org administrators are trying to, to kind of add or remove certain permissions or, or, you know, kind of, kind of do some permissions management, uh, within the UI, uh, that's kind of a big deal to, to be able to find, uh, you know, to be able to search for a user and find it. So that's a, that's a big one in terms of just, uh, helping everyone keep everything managed, uh, type thing. Um, another really interesting feature that's coming out is the, um, the ability to kind of share routes between spaces. So cloud foundry as this, um, uh, kind of like hierarchy built in. So you have the organization, cloud foundry organizations and cloud foundry spaces. So, uh, an organization is usually tied to like a line of business and a space is usually a team within that. So, um, within our customers and really, and any company you see, uh, you know, over time, there's kind of like people change and reorganization where, you know, team A moves to team B or, you know, the, the application that team A supported is now moving to team B. Uh, so this feature really allows applications to be moved around

within the people structure, uh, without taking downtime. So before this, they had to, you know, do like a full, like redeploy that, you know, like for this, say, we're going to move application A to team B, we're going to have to redeploy application A into team, into team B space, and then take down the old app route, move it over. And like, kind of like, there's like unnecessary, uh, outages, I guess, really just to, to deal with people changes. So this allows us to share spaces and do that, that kind of application migration ownership change, essentially without taking things down. Uh, so that's, you know, that's good in terms of like, you know, people don't have to schedule change for tickets or anything like that, just to change a ownership type thing. I mean, I mean, to use, to use the word, I think is a little ironic because it's trying to specify literalness there, there used to be, you know, where your app run was like in a, in, was physically defined in, in, in a space or an org. And now it's more virtually like, like it's not, it's not hard coded into that, that space, for example. So you can just easily move it over without having to sort of recon, you know, redo everything and, and reinstall it, which, which seems delightful. I mean, you know, that's, that's one interesting thing to, to, to note about, uh, the, the like cloud foundry way of thinking about things is the way with, you know, the way as, as you were going over it, it divides up its cloud, so to speak is based more on people. Like, like the organization structure you have and the teams of people doing things rather than the, the infrastructure that it's running on, which has that, uh, interesting effect there in comparison to kind of like a Kubernetes or another way of, of thinking about dividing up your, uh, platform. Yeah. Yeah. So it definitely has hard. Uh, so Tain's application service in cloud foundry has a like very hard multi-tenant scheduler. So you can, you can do these kind of organization, like these people, like, um, I guess, delineation, uh, without really having the works have to worry

about like the blast radius of the technology versus if you look at Kubernetes, you have this debate of, are we going to give every team like their own cluster? Does their own application get their own cluster? So you kind of have like this weird, you know, debate of trying to organize that, uh, you know, you have the container scheduler sprawl of trying to meet the people requirements versus, uh, you know, that's something I think that cloud foundry, uh, that got really, got right really well. It doesn't really get a whole lot of credit for really to press about, but the, uh, orgs and spaces really help, help that problem. It's not, not be a problem, I guess. They help address it. That's exactly. So, so what else is going on in the developer side there? Uh, so we have some improvements to the app autoscaler. Um, so within the, uh, within the apps manager, you can now, um, toggle the app autoscaler, uh, to scale up, uh, instances just, you know, by a, you know, set number. So before the app autoscaler would watch an application and say, oh, it's busy, let's scale it up one instance at like a set interval. But now developers can say, hey, my application, when it gets busy, I need to scale up by more than one. So I'm going to scale up by two or five or 10, 10 instances of the application. And that's really just to address, uh, like if you're going to get a big burst of workload, like some, you know, some kind of event coming on some kind of sale, something of that nature that those applications can scale up and meet their SLA without, you know, without too much issue. So that really helps make the autoscaler, um, to handle, you know, big, large, you know, burst of traffic, whether they be expected or unexpected. Um, and that was in our previous release 2.13, that was in the, in, in the CLI configuration only, but in 2.13 now it's in the nice, pretty UI that people can just click buttons and, you know, have, have feel great about type thing. Um, now you can feel great two ways. Yes, exactly. And you can see it too. It's the, the power of seeing something in the GUI is,

uh, is always nice. Uh, another, another interesting feature or something to call out is the, uh, uh, integrated windows authentication support, uh, for the.net core build pack. So, uh, one thing to do mention is that we, we do actually support windows applications a lot on the platform, which is, uh, I guess atypical from other, other services. Uh, so we do run.net core and.net legacy or.net framework, uh, sorry.net framework, which is kind of the legacy, uh, like windows type, uh, uh, framework, uh, in the platform. Uh, so we've rolled out the integrated windows authentication support for.net core, uh, build pack only. Um, so this will allow those applications that are running the platform or that couldn't run on the platform, uh, that because they required IWA, um, to now come onto the platform. So I'm kind of thinking these, these applications are still maybe on this, maybe a legacy application server. Cause they're stuck with this, you know, config. Uh, we can now accept those onto the platform and anyone with, uh, these, these type of applications to probably celebrate. Cause then they're probably, you don't have to worry about maintaining a legacy application server anymore. They can go onto a modern app run time. Uh, so that's, uh, to me, pretty exciting, uh, for those developers. And as you say, for, for several years now, I mean, uh, like.net and windows have been a pretty good first class citizen and, uh, and Taz, which, which does like, I mean, I, I come across people frequently who, uh, they really like it for their windows workloads versus, versus running, you know, on, on newer things like Kubernetes or other container stuff. So that's, that's nice. I guess that would, you know, with that, you, as you were saying, you can widen it up to, uh, older applications or applications that have to use older services. Yeah. So, so to speak, which that sounds good. Cool. Uh, and then I think the last one we, uh, we'll cover on the developer space is the, um, fair CPU entitlement. Um, so that seems kind of, uh, I guess, you know, bizarre without explaining it,

but, uh, it talks about really about the, the underlying container scheduler, uh, of the platform, which is called Diego. And this feature now allows, um, basically better use of the, the CPU scheduler within the platform. So when you, when you enable this, this feature, it, it basically prevents, um, noisy applications in the platform from taking over, like taking all shares of, um, of, of the CPU from, from the cell that they're running on. So it basically, uh, allows the CPU entitlements to be, um, uh, kind of honored. And if you have a bad actor, they're going to kind of be throttled and not allowed to take over everything. Um, so that's kind of, uh, one of the kind of, I guess, things to note as the, the, that the container scheduler in cloud foundry, Diego was always very like memory centric from scheduling. So this is kind of some of these bringing in some new CPU scheduling, uh, efforts as well. So, uh, that should help out some folks that really, you know, when they're at scale, helping to ensure that CPU is kind of, uh, fair across the board and you don't have, you don't have certain applications, just, you know, hogging, hogging a cell or multiple cells at a time. So that should, and, and, and from a developer that should basically result in predictable or more predictable CPU usage on the platform. Well, developers are probably a lot like my son. He's, you know, 11 and very interested in fairness. I don't know if you've ever experienced this, but they, they, he is both a fantastic lawyer. He can tell you how you have the minutia of how you've worded something incorrectly and that we should now word it to be extremely precise. And then he is a, he's currently getting his third doctorate in fairness, I think is what he's pursuing. So even with chips, is this where the specialism in screen time or any particular Absolutely. I think, I think that might be his, uh, his minor if they have that in postgraduate work.

Or, or, or is it a doctorate of it, right? Like, yeah, he's, he's pursuing multiple degrees. Yeah. I think this, uh, I think this list, Nick is great. I, there's some stuff in there that I, I think is really cool. I mean, clearly doubling down on that sort of ability to share things, to scale things, making the administration easier. These are, and these are all in the, in the apps manager side of the house. So these are all things that developers will see and be able to, these are knobs and whistles that they can tweak and tweet. Right. Yeah. Tweak and tweet. I, I, I guess. Well, yes. Brilliant. Yeah. Okay. Um, let's, let's cover a few operator features real quick. Uh, well, as we kind of go down the list. So, uh, one really interesting thing, uh, and we see this with, um, when you get to a certain scale and a lot of our customers have, but when you think about like, you know, resources within the cloud, you know, you think about CPU, disk and memory, um, you know, that seems pretty, pretty straightforward. But when you talk about a container runtime or application runtime, you also have this thing called logs, right? And most of these platforms don't really have any kind of quota or anything of that nature. And when you start to have, you know, thousands or tens of thousands of applications, um, running on a platform, the log output could be, um, you know, substantial. So, you know, the cloud foundry, uh, platform has a very extensive, like logging backend system that takes logs, you know, collects and aggregates all these logs. And then you can, you can, it can send them to various sources or multiple sources, you know, third-party sources or, or whatever for, uh, you know, metrics and log aggregation. And, you know, to, to, to, to, to really help, like when you get to this in this stage of microservices and 12 factor

applications, the, you need SRE teams to basically debug the Death Stars and microservices that you built. And logs and metrics are kind of critical to that. So we see like the log, uh, egress coming out of our platforms, uh, being, you know, you can easily see like hundreds of thousands of logs per second in some, some of these larger foundations. Uh, so we definitely over the years implemented things to, you know, kind of level that and, and have that be more predictable. But this, this, this next feature is, is, is really giving operators the ability to apply, um, log quotas now as well, uh, at the space level. So you, you basically have to have this contract with your app teams to say, you're going to get a quota on, you know, uh, disk and memory, and now you can get a quote on logs because if you, if you, if the app teams kind of write something that that's logging poorly, and it's a very busy app, it can, you know, it could be very bad in terms of consuming that on a third party service that you pay per log. Right. Uh, so this is something, uh, that we're, you know, this, uh, the log quote is being introduced, um, that you can do at the space level to basically give, um, allow operators to give their dev teams a quota and, you know, kind of give them, uh, force a practice, uh, of, um, better coding and, and making sure they log what's needed and, you know, don't, don't, you know, don't fire something up and debug mode and leave it and that type of thing. So, uh, you know, on the blog, I entitled it, uh, mute noisy application logs, but it's really kind of, um, getting logging quotas and ensuring that, um, you don't have a few apps just, you know, spamming the logging system essentially. And, and this, it also gives a mechanism to, to, to find, uh, these teams that have maybe poor logging practices. Cause they'll blow out the blog quota and be like, Hey, uh, team a, you know,

why are you logging, you know, 20,000 logs a second or something of that nature? Can you, can we look at, you know, your log statements and maybe, maybe, you know, do a little bit cleanup per se. So that's something that you don't really see until you get to a certain scale. And then, um, that's a problem that needs to be addressed pretty quickly once you get to that scale. So that's kind of like one of my, I guess, what's called a flagship operator feature. Uh, and I, I think it will end up saving, you know, hopefully a lot of money for our customers as they are able to, uh, kind of clean up some of their logging practices and, and, uh, um, reduce just the backend logging, uh, constraint. And, you know, especially if they're sending it to a third party, that type of thing. Yeah. As you were noting, there's, there's several features that are around, uh, when you're operating at high scale, small things add up if there are many of them, which, which, uh, uh, that's a good truism for you, but that is nice. Maybe also it would encourage a briefer logs instead of logging everything. As you were saying, that's, that's always, uh, uh, could be a bit of an annoyance. Yeah, definitely for those who have to read them. Uh, but I think it's good as well because it's that sort of feedback round that, that wasn't really there before around sort of, yeah, this, this one is a bit too chatty, you know, it's a chatty Cathy that we've got on the, uh, on the platform here. And, um, you, you know, when you translate that into a dollar value, it can add up quite quickly, can't it? So, uh, so I'm, I'm in no doubt that it's been a popular request and it will be a popular feature that. Yes. I, and, um, I, I, there are a few other features too in 3.0 that, that target logging as well, but they're more on the, um, backend logging system. So I'm not going to get down into the, those features as well. We have a, uh, a full kind of GA blog on Tanzu blog for, um, for the release. So we have a couple other features too, that, that will help, you know,

that are all targeting that, that logging problem to, to just reduce that overhead and help our operators out there as well. So I think I, I don't want to go through every operator feature just because there's a lot of them, uh, but maybe we can go down into, um, I want to touch a little bit on the tains of application service for windows updates, um, just because I want to call that out again. Uh, so we've got two, two features coming out for that. Um, so our Bosch agents are now going to restart automatically on our windows cell. So on a side note, Diego runs full in, in a full windows server mode to, to support the, the dot net framework, uh, code basis. Cause you have to, you have to run that on windows servers per, you know, micro Microsoft licensing, uh, agreements. So, you know, that whole ecosystem of Bosch and that infrastructure's code process we've built and runs kind of as the first class citizen for windows. And that not many people have that, uh, but we're making some improvements around how Bosch, the Bosch agent works on the cell. So it will automatically reboot and restart now. So when a cell, if a cell died or restarted, uh, an operator have to kind of go in there and manually wake it up, so to speak. So that that's been fixed now. So that's, that's a big win for, uh, the windows, uh, windows operators of the house. Um, and then I think another thing to let's to talk about, uh, I hinted at it before of really kind of the concept of platform extensions. And, um, um, that is this, this application service adapter, uh, pertains application platform. Uh, it's really, uh, a new project within the cloud foundry foundation called Karifi. Um, so that is kind of the next evolution of efforts for cloud foundry running on top of Kubernetes and this,

this, this, this, this, this, um, this product or this adapter, it's been out in beta for, I think since since January of this year, and the team's been releasing a release every, every month and it should go to GA here soon, but it's definitely, I think something that we're targeting for some of our, our, our operators and our, our core customers who want to kind of, uh, dip their toes in the Kubernetes water, but don't really want to lose everything they have already today. And, and that's, you know, another really exciting thing going on. Um, it, it allows your, your developers that you have today to maintain the, the, the CF CLI or the CF push experience, uh, on top of Kubernetes and Tanzu application platform. Um, so that's definitely, uh, something to keep, uh, your eyes peeled about as well as we continue to make pro progress there. So, so tell us, tell us a little bit about that. Like what, what is, I mean, I'm not too familiar with it, but what, what is, how does it work? What's the layering there? Sure. So, um, if you, if we, if we talk about tap, right, or Tanzu application platform, sorry for using acronyms, but it's, it's, it's, uh, my nature, I guess, uh, you take a Kubernetes cluster, right. And then you layer, uh, Tanzu application platform on top of it. Uh, there's some, some tooling to, uh, kind of that's wrapped in the Tanzu CLI to make that experience easier to, to install those packages. And once you, once you have a Kubernetes cluster with Tanzu application platform running on it, it is just another Tanzu CLI package to install that installs the, the application service adapter. So it's making use of some of the, some of the bits of Tanzu application platform, like Tanzu build service or contour and cert manager. Uh, so, you know, you know, part of that new platform stack as you will. And it's also has additional APIs that it's running to,

um, to represent the cloud foundry APIs and, and CLI interface within native Kubernetes constructs, like the, uh, the CRDs or the custom resource definitions. Right. So that's at a high level what it is, but from a developer perspective, they still have that, you know, they can still use their existing pipelines and their kind of, you know, developer flow where they push to a, uh, cloud foundry API, uh, you know, that's compliant. So that's, that's where, um, that, that adapter is going. Right. That makes sense. So you have the same experience as a developer, whether you're using like a ton, a cloud foundry based thing or, or a case platform. It's not fully compatible with the, the existing set. So it's, you know, it's, I, I kind of viewed it as like a cloud foundry light or a teams of application service light, uh, cause it doesn't have that full feature stack yet, but a lot of the core features have been, have been, you know, have been re, uh, I guess, uh, re-implemented and reproduced on the, uh, teams of application platform side with that, with that adapter. Makes sense. So what else is going on in the release? Uh, so two, two things to call out that are like, I guess we'll call them program related items. Um, and I, one of those might be, uh, uh, prevalent with, with the version, but we've adopted a new version format from like the, the TAS program. Uh, so we're going to roll out or we'll, we'll, we'll be rolling out and we'll be, uh, you know, targeting the, this concept of a major minor patch plus metadata version. So before, um, we were rolling out like two dot X, like two dot one, two dot two, two dot three. And we've been on that two dot series for quite some time. And we were basically, you know, releasing all this new major feature content on, on the minor patch stream. Uh,

so now as of Tansu application servers, three dot zero, we're adopting this new release and, uh, uh, approximately every six, every six months, you know, we're going to target a new major release. So we'll read that to three dot zero, four dot zero, five dot zero, that type of thing. Um, and, um, that frees up that minor, um, branch to kind of be like an ad hoc, um, as needed, um, um, patch for breaking changes, that type of thing. Like there's a, some kind of vulnerability that needs to come out where we need to introduce something to fix something, but it might break something. Um, that's kind of what the, the minor patch is for, uh, sorry, the minor track. And then the patch track is, is just for non-breaking patch changes. Uh, and we've also, the, the metadata flag has been added to, um, track what releases are going to be long-term support or not. So, uh, from a program level, uh, uh, we introduced this concept of long-term support releases for folks who didn't want to upgrade every so often. So we've, you know, previously we had this, this long-term support program, which targeted, um, a release about every two years, and then this customer should upgrade about every two years. And then in the meantime, we had a release target every six months. And that was kind of like the fast track. So from, from that perspective, um, we've kind of rebranded the long-term support program to the long-term support track. And then we had the, the, we were keeping the, you know, the fast feature track or the fast release track that, so, uh, that long-term support track now is going to be released approximately once a year. And then, uh, we'll still have that fast feature track every six months. So, uh, those, uh, those long-term support track releases will be, uh, designated them with the metadata LTST. So we can easily identify those releases when we're looking back and, you know, looking through the release, not a release, um, documents. And, and, and, you know,

if you're looking at Tanzanet to find, to find an actual, uh, you know, thing to download, it'll be very clearly documented. And those two, those, those changes that the new version format change and the updates to long-term support really just trying to, uh, help customers get features a little bit faster, but still be on that more, um, kind of, uh, structured, uh, cadence where, um, you know, very methodical, you know, the, the, most of these customers are very, uh, mature in their, uh, platform practices and, um, you know, there's a lot at stake and they don't want to be, you know, rolling out things too frequently. So, um, especially with the, you know, new feature functionality, very, uh, they're very risk averse customers and they, they like to make sure their platforms are, you know, um, Stable. Stable. Stable. Yes. Mission critical and stable. Yeah. And then, and then we were joking about this last week, but you don't fall into the, uh, the Java versioning problem where you're, you're a 25 year old technology still on the, uh, the first version 1.X that they're operating on. Yeah. You can actually advance beyond that, which is nice. And then, I mean, to point out the obvious implicit in that is that there is a long-term support program for, for people who want to guarantee that, that stability there. Uh, which, which is, uh, nice. Well, well, what else is, is there anything else, uh, in the release that's notable? Um, let's see, kind of glanced over some of the operator features, but I guess from a, from a security perspective, well, let's, I'm gonna, I'm gonna call this out just cause it's, I like saying it, but, uh, jammy jellyfish. Uh, so the jammy jellyfish stem cell is, is being rolled out. So if, for those of you who don't know what I'm talking about, um, it's really about the new version of the Ubuntu operating system. So, um, Ubuntu has very, uh, awesome names for their, for their operating system releases. So jammy jellyfish is Ubuntu 22.04, uh, also their long-term support release. Uh, so that is being rolled out for, for Tansu application

service and, and operations manager as well. So, uh, kind of by default, um, there'll be a full operating system level upgrade going from Ubuntu 16.04 to Ubuntu 22.04. And that's all kind of magically handled by Bosch, but that's kind of a big deal. Cause you know, typically when you do a full, uh, operating system, you know, replacement that may or may not, you know, you know, require a lot of work, uh, depending on what you're doing. So that's a, that's a big feature too, for the, for the operator side of the house to just really ensure that, uh, we're on that latest, uh, OS and, uh, everything's ready to go there. Plus I like saying jammy jellyfish. So it's kind of hard not to like, great. Exactly. Yeah. Yeah. No. And, and, and as you're saying, right, like one of the, one of the, uh, uh, benefits or promises or whatever, uh, of cloud foundry is like by the nature of how stem cells, stem cells are like, you know, layered and then, and then your actual, like, uh, deployments are layered on, uh, on top of that. Like there's a nice separation between the operating system and the higher layers that makes it less painful to deploy a new operating system than it may be otherwise. Definitely. If you're doing all of that on your own, but that's, that sounds good. Well, it seems like the theme is, uh, you know, there's, there's, there's both like the, uh, quieting down all the noisy things and having more control over higher scale stuff, uh, that are used in, in bigger installs. Uh, and, and then there's the, the sort of like stability that comes with like, now there's a big install of things and we need to have like a multi-year sort of plan for how we're going to handle, uh, handle support for that. But also like, you know, what, what's the, when you want stability in a platform, it's good to know like the expected release schedule for things, kind of like the philosophy of adding in new features and, and releasing them and supporting them, which is kind of implicit in that, that version numbering there or, or, or behind it. Right. Yes. Yeah. Do you have anything you want to add Ben?

I do. I think there's one thing that, that is beautiful about, um, Tanzu application service, and that hasn't changed from a developer's point of view. And that is how it makes you feel. The, the very first time that you get behind that keyboard and you do CF push your app and a few seconds later, you get your URL and you can hit it. And you know that behind the scenes, all, all sorts of mountains have been moved to make that happen in such a smooth and predictable way that it's brilliant. And when I was showing clients it, you know, it regularly blew people's minds. It was like, okay, this, this has now suddenly become really easy. I like that. I like really easy. Anyone who knows me knows I like that. So I like the fact that that hasn't changed, you know, that is still there, still baked into the, to the platform to this day. And it's still putting smiles on people's faces as they are working through their everyday, putting business logic into production problems, you know, so I, so I think that is really, really commendable. And I'm very, very pleased that there's version three is coming along and that, and that we're still investing so much into it as well. I really like that. Yeah. That reminds me of my push, my, my first CF push experience. So sitting in, in a lab with a coworker, and this is quite some time ago now, but we, we had the teams of application service installed, you know, at the time it was called a pivotal cloud foundry. Uh, we got it installed. It took about, you know, four or five hours to, to get installed on top of our vSphere environment. And we were kind of fiddling with the CLI. And so we like, you know, we pushed an app and I don't even know what the app was. It was probably, you know, hello world or something. Right. And we got it to work and we're like, wow, that's awesome. And they're like, all right. What's this thing called CF scale, like CF, CF scale dash I a hundred, that scaled up to a hundred instances. And then a hundred instances started and we're just like

that, you know, like not only did you, you get your application running and then you just scaled it up to a hundred instances in a few, you know, minutes, it was just like completely blew our mind. And that was quite some time ago, but yet it's still that, that experience is still, still a hundred percent relevant today. Uh, you know, and the art, you know, the customers of this platform still, you know, get a great, that, that great experience still lives on and continues to, you know, has a bright future. Uh, I'm real excited about. There you go. If you haven't tried it, try it, it will blow your mind. I'm glad you shared the same experience I did. I got to say, yeah, it's really good. I love it. It's really good. Well, speaking of relevancy, I'm sure we'll be talking about the Tanzu application service a lot at the spring conference. We have coming up SpringOne conference, you know, despite its name, one has actually been around for a while. It's, it's, it's more than a 3.0 product. As I recall, I don't know the spring conference has been around for quite some time. I remember going to a very small one way back in the late two thousands, but it's much larger than that now. And, uh, it's going to be December 6th to 8th in San Francisco. Uh, and there'll be lots to learn about the Tanzu application service and application platform and Java and just, uh, spring doing.net things, programming and, uh, doing operations and even a little bit of, uh, I know it's your favorite topic, Ben digital transformation. You know, you and I, when we hang out, we're always like, let's go talk about digital transformation. Yeah. And, and we like talking to other people who like talking about that as well, which is always nice. Just get a, get a big cup of digital transformation and, and, uh, start drinking it. I, I joke, but what, what, you know, what's meant by that is, uh, hearing from people who are actually using software and, and, uh, building and running their own software to like run their business and their organization and, and new and better ways and, uh, changing over to it. Anyhow, that's going to be, uh, December 6th to 8th. And if you go to SpringOne.io,

we have a fair amount of the sessions posted. So you can kind of browse around and look at it. But when you, uh, when you're there at SpringOne.io and you register for it, if you use the code Coté 200, you can get $200 off, which is better than $0 off. Uh, last I checked how math works. So you might as well use that code and also being relevant. Uh, if you went to tanzutalk.com and, uh, looked at the show notes for this episode, we'll link to the blogs that, uh, that, that we mentioned and went over and, uh, all the other relevant stuff, uh, to take a look at. We'll, we'll have to get a special collection for people who go see the show notes at tanzutalk.com. And, uh, with that, uh, do y'all have any other final shots, uh, thoughts you want to share, whether it's about, uh, digital transformation or platforms. That seems like the ideal segue. Cause I think Nick, you've got a talk at SpringOne as well, haven't you? Oh yeah. I mentioned that. Yeah. So I, if you want to learn more about tanzu application service, uh, I'll be at SpringOne, giving you a talk about, uh, kind of a deep dive. And, uh, so yeah, come check it out for sure. Uh, you'll love to talk about tanzu application service there. And then I got one more thing to hopefully close the show out, but I'll, I'll wait till everyone else is finished. If we're any, is it? Yeah. Are you going to sing? No, I'm not going to sing. I don't, but that's no, nobody wants to ever hear me sing. Maybe for, maybe for next episode. I mean, I think all you, you, we just put a vocutter on it or whatever that an auto tune and then it'll be perfect. Right. There'd have to be a lot of auto tune to make that palatable, I guess. I'm sure it'd be great. All right. Well, what's, what's, uh, take us out with your one last thing. It's, it's this, you know, to sum up, uh, tanzu application service, it's best done with this, uh, classic haiku of here's my source code, run it on the cloud for me. I do not care how. And, uh, that's, that's, that's the whole kind of premise behind the platform. And I think everything we talked about that today kind of help drives that, but I always like saying that. And Nick, I'm right in saying that you have that embroidered on your pillow.

Not yet, but I think I will, or I'll get some kind of swag or something like that. Maybe a hat or a shirt. I don't know. Pillow embroidery might not be the easiest thing to get lined up, but well, everyone always needs a new hobby. Well, maybe, uh, maybe get Ryan to, uh, etch something with his laser etcher or something. So one of our teammates, I think just recently got a, some kind of laser device that does etching. So maybe, maybe that might work. He very nearly etched his kitchen countertop in the unboxing excitement. You know, that, that might be handy to have like, uh, you know, um, like imperial to metric charts, and then also just your imperial thing. So you can sort out what a tablespoon is relative to a pint, relative to a court and all of that. And you could just etch that on your kitchen counter and really never have to look that up again. It would just be right there. It would be really important to get it correct though. Yeah. That's what you want to shoot for. Yeah. Yeah. You can't really roll back that if you go, it goes wrong. Right. Well, the utensils, you know, it would be like CSI when there's someone lying on the floor and there's like this thing around it, you could have that, but for it, for your, for your knives and just baboons and your folks. Right. Right. See it all, it all goes back to, uh, having kids help you out. If you, if you etch that, they would know where to put things. You can just say it's like the, the outlines that you have. Yeah. Well, speaking of knowing where to put things and etching, it's probably etched in your mind, dear listener, that if you wanted to get this, uh, this, the show notes for this episode, you could go to tanzutalk.com. And what I always recommend is that if you haven't subscribed to it, you get, you get your, your podcast phone or computer out. And if you just search for tanzutalk, you can download it and subscribe to it. Also, if you have children who are putting silverware where it needs to be, because you've etched it, find all of their devices and subscribe to the podcast to download it. Any love of loved ones that you have, uh, you know, partners, spouses, grandparents, friends, uh, your kids, parents as well, just grab their phone and subscribe to the podcast episode. And I'm sure they'll, they'll

thank you. They'll, they'll send you cards this holiday about how wonderful you've made their year by subscribing to this podcast. And with that, we'll see everyone next time. Bye bye. Bye, folks. Bye.

Tanzu Talk · 🍕 Get more pizza.