Cloud Foundry Myths Dispelled (Ep. 63)

Episode 49 · 51:09 · SRT · archive.org · all Tanzu Talk

Now, I know that you're up in the land of Costco there, so you might actually have a good... Do you go to Costco a lot, Richard? I try to avoid it. My wife does with our giant school bus-sized car to fill things up, but I try to avoid it. You are a Costco family. However, you may not participate in it as much as possible. That's right. You don't find it delightful to go in there and find what Easter eggs they have in the clothing section? Maybe a good bottle of wine or beer? You know, it's the parking lot stresses me out, and then do I need a keg of relish? It's just the volumes. Do I need 62 razors? Like, yeah, so it'll last me until I die, but that's just such an aggressive purchase. And you've got to be careful. As listeners to my Software Defined Talk episode, available at softwaredefinetalk.com, sometime last year, I think, I decided I would go all in on deodorant, and I bought the Costco variant, which frequent Costco people know that they only carry the Gillette Cool Mint, or whatever they call it. It's got something like Cool Breeze or something. And, you know, you have to buy, like, eight of them, which is a mistake, because it turns out that scented deodorant is not a good idea. So now I've got, like, I've got this Gillette, like, you know, Irish Cool Breeze stuff sitting under my sink. You can get the odorless, arid extra odorless, right? And that's what you can shoot for. For Mr. Small Batches over there, you would think Costco would be the antithesis of your entire life. Yeah. Well, you know, it's part of my mode one approach. There are certain things that I know we'll just be doing a lot of, and relish is not one of them. But I'll tell you what, you got the bread, all the things the kids eat, whatever. Steaks, you know. Or a jug of mayonnaise. But this brings me to my original question, which is, Richard, what is your position on rotisserie chicken? I'm a fan. You know, it makes a good spontaneous meal sometimes. Got to get some rotisserie chicken, bring it home. Uh-huh. Uh-huh. I feel like, I feel like, so first of all, I love rotisserie chicken. I grew up eating rotisserie chicken and Popeye's chicken.

And not like, you know, not unlike, I guess, that Elliot guy over on that podcast. But I've been thinking recently, you know, every now and then there's one of these articles that comes out that's basically like, this food that you always thought was great is basically injected with death, right? Like, with butter or, you know, it's full of sugar, like the people who make it do that. And I'm wondering, I need to look this up, if there's something about rotisserie chicken that makes it so delicious that makes it really unhealthy. Because I feel like, I feel like if I eat a rotisserie chicken, I'm doing all right, you know, healthy food-wise. But I don't know, I'm so suspicious. I wonder, you know, I wonder what they'd like. I wonder if they sort of like, if they marinate it in a vat of like ghee or like clarified butter or something. I mean, there's just something's got to be going on. Right. It's like massage with lard or something like that, which you don't want to hear about. Yeah. Yeah. So I don't know. If anyone knows what's up with rotisserie chicken, I would like, I'd like to know if it has a clean bill of health or if it's something under advisory that I should, I should eat in moderation. You know, like, like peanut butter is one of these things. Peanut butter is delicious. And if you get, if also available at Costco, if you get sort of like the organic, whatever peanut butter, it's not just like half sugar, like some of the other brands, but it's probably not a good idea to just like eat a whole Costco tub of peanut butter in one sitting. Right. Right. Like you gotta, you gotta, you gotta control that. I think nuts on their own are okay. But once you grind the nuts up, kind of like a fruit juice, you gotta be a little careful, right? No, if you're, if you're eating peanut butter with a spoon, you have to question your life choices. That's right. You got peanut butter on one side and maybe a whole rotisserie chicken on the other. And you got to figure out what's going on there. And, and, and apparently, apparently a tub of relish. Right. That's just a cry for help is what I'm hearing. That's good. Well, we don't have a, a guest this week. You may, listeners may, may figure out for me not introducing them. So I thought we, we actually have a, a bevy if that's still a word of news to go over. So, so we'll cover some news. And then I was thinking, so Richard wrote a post recently about debunking cloud foundry myths.

And I thought we could, we could go over those. It's a, it's a delightful listicle that's sort of on meme overdrive as well, which is, which is really fun. It's always good to see a topical memes. But, you know, but before we get to the news, did you, uh, did you put together all those memes who, who, who did it? Yeah. I did. I figured being passive aggressive, it's better to do with humor than, uh, necessarily just jerk. So that'd be more fun to have some fun with it. Yeah. You've got the, uh, you've got the biting sort of, uh, uh, fisticuffs of marketing. And then you've got the, uh, you've got the, uh, humor, I guess. I can't get mad at Ron Burgundy. Like it just completely diffuses you. So that's right. That's right. I wish we had a Jaguar sound to put in here now, but we don't. Well, uh, first of all, uh, you know, um, I thought it'd be good to, you know, we bring this up every now and then, but there's actually a whole bunch of events coming up. Uh, some of them, uh, pivotal events. Uh, but this week I was just talking with, uh, with, with Richard about this before we're recording. We have DevOps stays Austin, as you might expect in Austin. Um, and there's actually, I, I don't, you know, Austin's one of the bigger ones. I bet they'll have like four or 500 attendees, uh, this year. They have it down there at the, uh, the, the university of Texas stadium, you know, that's a American football for people don't know that very large stadium. I think, I think we need to recall it the stadium of DevOps. It's the second year they've had their, you know, I have mixed feelings about having it at the stadium, but whatever it's cool. Right. And, uh, there's a huge roster of like all the famous DevOps people there, including our very own Andrew Clay Schaefer, as people like to say, you've got, uh, you've got Nicole there. I think Jean Kim, uh, of, of DevOps report fame. And, uh, I forget if John Willis is showing up, but pretty much all the, uh, all the DevOps people, I think maybe except for the Belgian guy, uh, will be there. So, so that'll be fun. I was at DevOps day Seattle last week. Oh yeah. How'd that go? Good. I'm actually wearing the shirt today just to show the pride, but just for folks who maybe wonder if it's worth attending, it definitely has a great vibe as a group. It seems like everyone's really amped up different backgrounds.

This was not all just ops people. It was a mix of folks. So, uh, definitely a ringing endorsement, hopefully for folks who are on the fence of whether to go or not. We sponsor a lot of those at pivotal and I go to a, uh, to a lot of them and they're always, they're always pretty good. There's, I think, I think there's about maybe like a 30% repetition of topics, but most people who go to DevOps days don't really go to a whole lot of them. So I don't think that's a, that's a problem at all. It's, it's a good show. And, um, I was just checking to see if the talks for DevOps days, Atlanta are up yet, but they're not, but once those are up, maybe we'll put a little link in here. There was myself and Josh Long talking there. And then also the week after that, which I guess is next week, uh, technically, uh, there's, there's OSCON in its new home in Austin. I think this is the second, maybe third year and, uh, I'll, I'll just be there hanging out at the booth, uh, just, just doing things. I am interested, you know, I should talk to you as a source for this, uh, Richard and other people, but I was talking with my editors over at the register and saying it might be a good opportunity to, to do a, uh, a WTF open source thing, but I don't know if there's, there's much, much of a story there beyond sort of like a, uh, classic Matt essay write up, but it'd be interesting to see what, uh, what happens in, at OSCON. And then also, uh, uh, we've got here at pivotal, we've got a bunch of cloud native road shows coming up. I think, uh, you know, we've already blown through like three or four of them since we started talking about them, but I think later this week, uh, there's one in Charlotte and then you got Detroit, Toronto, St. Louis, Paris, London, Munich, Stuttgart. Is that right? Uh, you got Dallas, Denver, Los Angeles, Seattle, San Francisco, Amsterdam, Seoul, Hong Kong, Sydney, and Singapore. Uh, so that, you know, that's all over the, uh, the spring and summer. And if you go to pivotal.io slash roadshow, you can see a list for all of them. And those are, uh, those are free to go to. And I think most all of them are done by pivotal and the people from Google cloud. So you can get a sense of not only what a pivotal cloud foundry stack and approach would be like, whether it's on premise or off premise, but also what it would be like running on Google cloud with the, uh, the Google cloud services tied in there.

And, uh, you know, for those of them who are just looking to fill their, uh, metaphoric bag of goodies, I believe there's also a free breakfast and lunch at each location. So nice. And did you say Singapore? Have I been saying? I think Singapore, but maybe I said Singapore as in, uh, I didn't know if I've been saying it wrong for 40 years of life. So, uh, yes. Uh, you know, who knows? Who knows? I've never actually been to Singapore. That would, that would be fun. I'd like to, I hear it's wonderful. Uh, and then finally, uh, to not bore everyone with a self-promotion or of in so much as pivotal and, and I guess I identify with each other, or maybe that's, uh, one directional. You, we got a spring days is coming up and that's a two day event and it's going to be in Chicago and New York city and Atlanta. And that's very much so focused on the spring framework and doing things, uh, in that area. And, uh, it, it isn't free cost $200, but I feel like $200, you probably spend more on like, you know, phone charging cables in, in a year than that. So, uh, it's not that big of a deal, but when I was in Atlanta at DevOps days, a lot of people were interested in that because as I'm sure you experienced at, uh, DevOps days, uh, Seattle, a lot of people who come up to the booth are like, yeah, we do spring. So, uh, in those three cities, you can get the latest on what's happening in the spring framework, like spring cloud and spring boot and spring cloud data flow that we've talked about and pretty much any series of technical words you can think of and put the word spring in front of it. I think there's a, uh, a project for that. So, uh, not to nauseate people, maybe we'll come back to the last two items at the end of the episode. Cause, uh, you know, we don't want to just throw out a bunch of ads, I guess. So what, uh, what news have you been looking at, uh, since last we recorded? Yeah, there's a couple of things listed out. There's last week was all the earnings stuff. So, Hey, Twitter did well. Amazon was way up. They hired, I tweeted it wrong. Cause I said AWS added a Microsoft. Amazon added a Microsoft, they had 110,000 employees, which is about the size of Microsoft. So Amazon got bigger, uh, still making lots of money.

Google did well. Microsoft had something like 90 something percent growth in Azure. So, Hey, big companies be making lots of money. So that was always, I mean, interesting to see. I think again, for our industry, it's good to see that people are still betting on these large platforms, especially the big three cloud providers. Uh, the other thing I listed out, which I think I tweeted it yesterday too, was Dom, there was a story about Domino's and digital transformation. And it was interesting because the take was, Hey, sometimes companies and Coté, you probably see this company sometimes think of it transformation and think of internal. They think of the employee experience. They think of, you know, optimizing delivery and all these things that are useful, but still internal facing. And this was a story more about look at Domino's pizza and how they've, they've changed how you order and they've have, you know, this giant percentage of people who order through an app now versus a phone. And so they've digitized the customer versus necessarily solely being inward focusing. I thought that was interesting. And then today there was a story on, you know, Domino's using if this, then that to kind of, you know, turn off your sprinklers when the pizza man arrives. You could hook that up. I didn't, I didn't, I didn't, uh, I didn't go read that story, but I figured it being from the internet, it would be the opposite of what you're saying is that you would turn it on when, when the pizza person came, but maybe. Maybe that's what the read it people would be up to versus the, uh, the nice internet. Yeah. The, uh, I came across that, that, uh, you know, the, I think the thing I came across was, uh, this guy, I think he's still at BMC John hall. And he writes a lot of, uh, a lot of good stuff like this kind of pointing out how mainstream it stuff is, is doing all right. And, you know, one of the points was that the, uh, as one way of measuring it, that the share price of Domino's has had a huge percentage of growth compared to all the tech companies, even including Apple and, you know, again, it's just like one way of doing things, but yeah, you know, you can order pizza from your watch. And if I remember, uh, basically it seems like their idea was wherever, wherever our pizza loving people are, we want them to be able to effortlessly order a pizza. And I remember one of them was like, uh, I forget if this is everywhere, at least in the

UK that you can set it up that by default, when you open up the Domino's app with, if you don't do anything in 10 seconds, it'll just automatically order a pizza for you, which, you know, uh, that, that would be dangerous for people like, like me who likes to select things, uh, when, when I get a pizza, but I can see that it's, it's an interesting study of like, uh, how do we just get pizza to people? It also reminded me of, um, I forget if we mentioned this, but, but our, uh, our, uh, I don't know, our, our, our big brother people, Dell Technologies has a little podcast series with, uh, Walter Isaac. Is it Isaac or Isaacson? The guy who wrote the, uh, most famously this, the Steve Jobs biography, the authorized biography. Um, and, and episode three that they did appropriately called a slice of disruption is, uh, it's all about, uh, how, how Domino's has been doing this and a lot of detail on it. And they asked me to help them out with an accompanying story about DevOps and digital hootenannies. And so you can, uh, you can go check that out. I, uh, there's, there's, there's some commentary about all state and other pivotal customers there, but, you know, to be honest, I'm usually a little like leery of, uh, big corporate sponsored sort of like advertorial stuff like that. But this is actually a really good episode, uh, that, that I listened to because I'm always interested in, in hearing the exact before, after details and, and pretty, pretty rock solid success momentum of companies doing things. And I think, I think aside from some, the usual kind of narration that people like to do the, uh, the business side of that episode's pretty good. So it's definitely worth checking out. Yeah. And I think, again, the fact that the Domino's thing, I think it's just, if you, I know most of the listeners here don't work at pizza places, but they do work at retail insurance, banking, government manufacturing. And I would bet there's a lot of these sort of thoughtful customer facing automations that you could be doing that actually give you really loyal fans versus just hostages as customers. So I like that kind of thinking. Yeah. And, and also, I mean, to that point, as far as the applicability of it, right?

Like if you think about, uh, if you think about the margins of a pizza delivery business, they're incredibly small, which is to say they don't have a lot of cash to just play around with. And so the fact that they decided to allocate cash to, uh, do actual, uh, cloud native kind of stuff, uh, it's, it's good proof that sort of like anyone can do it. Even, uh, even companies with tiny margins. I mean, you know, it's not like food service companies have lots of money to spend on things. So the fact that they prioritize that and it's worked out well for them, for those companies that have, I don't know, say double digit margins and actually have money to spend on things, it should illustrate that it's, uh, definitely possible to do. Great point. So what, what else do we got going on? Yeah. I pointed out, uh, you'll see in the notes, Kubo, if you remember is our Kubernetes built with Bosch thing that we announced at Google cloud next back in, I think it was March. And we've had a couple of releases since then. So this wasn't just some weird gimmicky thing we did just to get some press at a conference. We've been making some updates. So it's the latest version of Kubernetes. We've added some things so that routing can come through the cloud foundry router and kind of expose your Kubernetes apps publicly. So again, when you're looking for that single control plane for all of your different platform type things, Kubo is, uh, continue to be invested in pretty heavily. So I think that's pretty neat to see the evolution there. The other thing I listed out was, uh, so every quarter pivotal does these pivotal partner days where a bunch of partners can come in or prospective partners can come on site and mess with us for two days, kind of pair with us and build integrations into the platform. And hopefully out of that come interesting integrations that you, the customer can use. And so there's a few things I was just browsing around the pivotal network marketplace on Friday and noticed some, some cool new things like cloud flare, uh, uh, Jenkins coming from our friend at alterose pager, do the honeycombs, like some really cool tech that was now in our marketplace for beta tryout. Most of those, I think all of those actually came out of the last partner days. So you as a customer, I guess, keep checking out the pivotal app store, if you will, to

see if there's some, you know, maybe you're using a more classic monitoring tool and want to upgrade what that looks like for cloud native, or you are starting to push apps to the cloud more and you want to do a real CDN solution. It's just, it's a great place to start trying some things out that are purposely built and integrated into the platform. I went to one of those, uh, those pivotal partner days, I don't know, several months ago. Cause it was in Austin and, uh, I basically showed up and, and I was late, uh, which, you know, that's, I guess that's what I do. Uh, and, uh, and then there were actual, like, everyone was like pairing up and actually coding. And so I kind of poked my head in and realized, you know, there was no reason for me to be there. So I had the free lunch and left, uh, but it was, there was actually like a lot of, uh, actual real life things going on there. They're, they're, uh, they're working things, if you will. So if you're someone interested in, in helping out in the, uh, the overall, I don't know, cloud foundry, pivotal cloud foundry ecosystem, it's definitely, uh, it's definitely not just, uh, you know, a mouthful of slides that you go see there, you end up actually doing work. And, uh, so, uh, the last thing is I was just noticing this morning, the Uptime Institute and they're, they're, uh, unless I'm tragically wrong or I don't know, tragically, unless I'm wrong, they're still owned by a four or five one research where I used to be an analyst, but they've been around for a while. And there was an overview of a survey that they did that I thought was interesting and relevant, especially to, uh, the sort of like, uh, cloud native kind of world. And, and, and the, the sort of like, uh, top line thing was that according to their survey, and I think it was maybe like 2,500, 3,000 people that they surveyed and to give you another idea of demographics. So when I work there, I would, uh, I would come, I would talk, work with Uptime people every now and then, and they do a couple of things. One, they just cover research for people who build and run data centers, like the actual data centers. Uh, I don't know why I repeated that. And, uh, and, uh, but they also, they do certification and I think they have their own kind of certification scale of like, like all certifications. It's sort of like how awesome you are at doing the thing. And so they would, I think it's like tier one to three that they would certify your data

center was at, which, you know, all the usual stuff of being unbreakable and whatnot that, uh, and resiliency that you would rate there. So they, they encounter and work with a lot of people who run data centers. So their survey base is probably pretty legit. So anyways, uh, they were saying that at the moment, there's still about a 65% of workloads are running in on-premises data centers. And by that they mean largely not public cloud, whether it's co-load or on actual owned things that organizations own on their own or whatever. And, you know, as I was doing in my little writeup that we'll link to over on my own blog at cote.io, like it's, it's always interesting to see this kind of stuff because, uh, I think, especially in the roles that you and I are in, uh, you encounter this sentiment that it's, uh, most everyone must be running in public cloud at the moment. And, uh, I recall even all the way back to, uh, I mean, I used to be one of these people more or less. I recall back in, uh, in 2012 or so, uh, back when I was at Dell, we were doing a tour, uh, up in new England of all the analysts. And, and I remember there was, uh, well, I received a fantastic CD of, um, who was it? The Allman brothers playing famously playing live at some, some, uh, some, some venue. Cause I think the, uh, George is a colony. Whoever heads up, uh, Forrester is a big, like Allman brothers and live music fan. There's like guitars in the lobby. So every guest receives it's an Altmont or something receives one of these CDs, which that, that's a nice thing to get. Anyhow, in that room with a bunch of the, the analysts, there was a, uh, a raging debate amongst themselves about on-premise versus cloud IT back in, uh, 2012 or so. And it seems like in my own anecdotal experience and by surveys, uh, public cloud still hasn't, uh, gobbled up everything. There's still a lot of on-premise going on, which is what we find amongst, uh, pivotal cloud foundry customers. I think, I don't know if it's every single customer. I'm trying to not be hyperbolic in my phrasing as much as I usually.

But pretty much all customers, how's, how's that for, uh, you know, not being a hundred percent, but pretty much all customers that, that we work with run on premise, but equally pretty much all of them would like to use a public cloud, if not move most of their stuff to public cloud in the, uh, the near and long-term future. So I don't know. The, the upshot is that, uh, I don't think anyone should be surprised that there's so much on-premise IT and, uh, cause there's a lot of it, which doesn't invalidate public cloud. Uh, and I think ultimately like the way that I think about it with myself, when I sit around thinking about cloud with myself, I guess, is like, I'm pretty sure at some point in the next 10 years, we won't have this distinction of, uh, on premise, you know, off premise as much as we do and it'll just be the stuff that, that, that was, that was a good roundup, but it'll just be, you know, like the overall cloud that we use to, uh, to run everything regardless of where it's physically located, except for sovereignty stuff and things like that. Yeah. That 65% actually seems low or I guess high. I would expect it to be more of it was still on-prem. Yeah. It's actually slower than I would have thought. Actually, I would have thought it'd be still more like 75 or 80 or something like that. So that's interesting. Yeah. Yeah. You know, I was also thinking like after I had looked at it this morning, that I mean, I suppose I could try to go get the original survey and dive into it, but you know, I also got to prep for this podcast and do other things with my time. Yeah. That, that feels like a lot of work. Yeah. It's lots of work, lots of work, but, uh, uh, you know, it, there, there probably is some interest. It would be interesting to look at the workload types like package software or batch job software, uh, you know, a whole bunch of things like that versus custom written software applications and stuff like that. Because I think I would suspect if you were to look at a cut of custom written software, uh, that number would probably be lower because you've got a whole bunch of mobile apps and, and, uh, just other things that are new applications that, that might be running on public cloud more. Who knows? And then if you also, if you throw in things like, you know, Dropbox and box and other stuff, who, who knows what it would be? That's right.

Well, we'll have a little cameo from my daughter on here before she leaves the room. She, she, of course, will be a, uh, a cloud native that probably doesn't even identify that being three years old. It'll just all be cloud to her. She'll be a, uh, what would they call that generation? The, uh, the, the millennial Xers that that'll be exciting. Yeah, that's awesome. I know I, it's a, we'll see what names they're running out of leathers, so they'll have to come up with something. Oh yeah. They'll start doubling up on them. Like the airport parking, it's like double a hair. Let me try to take her downstairs and then I'll just edit this part. Hold on a moment. Sure. Sure. All right. I think we should be fine. All right. Well, like I said, I wanted to go over, uh, I don't know, several weeks back, I think that you did this post, but I thought it was a good roundup. I don't know myths to use your title that people have about cloud foundry. And, uh, yeah, I mean, I mean, it is good because of course in our, in our, in our, uh, marketing position, we, uh, you much so more than me, uh, spend a lot of time on counter counter arguments. So I thought it was a, uh, a good wrap up of things. Yeah. I'll be interested in your take on some of these. I mean, this, this really just came from, I mean, look, you and I'm going to be I mean, look, you and I spend a lot of time also reading online and, you know, listening to hear what customers say. And in many cases, these myths are not, it's not nefarious. It's that people, you know, you do drive by product announcements and like, oh, okay, this is what this thing is. And you instantly form an opinion, sometimes not founded in total fact, but you just kind of get the snippet and you got the, you know, the Twitter experience from it. And so the point of this was how do we look at some of the top things I've heard from customers, from even competitors, from the market and just, you know, where are those things that we can help clear things up? I mean, the point of it was not to throw mud. If anything, it was just to actually provide clarity and say, hey, if you thought the product was this, here, it's that. It still may not be the right fit or maybe it is the right fit now, but let's make sure we're all working with the right set of facts. One of the things I've noticed even more having worked here over two years is there's a disadvantage that an enterprise software vendor has in releasing things as frequently as we do, like every three

or four months or so, which is it's, you can't exactly go to the internet and correct all the previous write-ups of things. So you do have to like spend a lot of time saying now this is different or we've added this functionality or fixed this sort of thing. And yeah, I mean, that's a difficulty to do. So I think every now and then it's good to sort of just like come in and say like, here's sort of like the baseline of features and functionalities that we do. And I think getting into it, you know, some of them in particular are the, there's a cluster of them that I would call, you can only run special applications on this, which I think as you get into writing is not exactly the case that you have to be these super cloud native apps. That's right. I mean, if you break everything down, it really comes down to, you know, I think sometimes us putting this into a too small of a box and saying like, hey, yeah, it's just for cloud native stuff or, hey, it's just for microservices or, hey, I can't run this or, you know, it's not for that or it's too enterprisey. Heck, I saw a conversation this weekend talking about that. So, I mean, if we go through a couple of them, you know, in that categorization, I'd called out, you know, the apps that run on CF. It's really easy because we talk a lot about, I mean, you and I, and even on roadshows talking about cloud native apps, because for the most part, it seems like the consensus is that is the right future for everyone to be adopting. It doesn't matter if it runs on-prem or in the cloud. You should be building apps for resilience and scale and fault tolerance and all those sort of fun things and velocity. But at the same time, I mean, Coté, you and I were actually both reading some notes this weekend from the field from an unnamed customer who just finished up, you know, the third week of a dojo with us doing replatforming, and they were just going nuts about how easily this person who had never touched boot had a boot app deployed that afternoon and how they had added all this test coverage and CICD, and they were just blown away that they were getting to prod. So, these weren't Greenfield, let's build sprinklers to talk to my pizza sort of apps. This was like, hey, these are just existing legacy apps, and we're doing some light modernization

so they run. So, when I look at CF, it's not just microservices, although we talk about that a lot. It's data-driven apps, it's event-driven apps, it's background jobs, it's all the sort of things that make up most portfolios, and that pool of things that don't fit is really shrinking every single release. Yeah. I think one of the things that, you know, it's one of those things where it slightly hurts us commercially to kind of have this position. But I think our position is that by default, right, you should do things in a cloud-native way, right? And the whole point is, as you're highlighting from, as I recall, that large manufacturer, that it's probably better to do things in this new cloud-native way than in the previous ways of doing things, because it's been demonstrated from, you know, your unicorn-y Spotify's and Netflix-y's all the way to, like, you know, these manufacturers and insurance companies and banks, that when you do things in this new way, it's better, right, by all productivity and business value sort of metrics. And, you know, if you don't want to do things that way, then you'll have less better sometimes, right? Now, in other instances, the improvements that you get are just all about, like, fixing parts of your process. They're kind of like local optimizations that you can do. So, you know, we have one of our more public and larger customers is Comcast. They often say that they have, like, 900 or 1,000 applications running on Pivotal Cloud Foundry. And from talking with them, you know, some of those applications, I mean, not all of those are brand-new cloud-native applications that are Greenfield, because they've only been doing it for, like, two years. And, well, I don't know, maybe they're awesome, and they could write over 1,000 applications from scratch. But some of those, many of those have just been migrated. And you see that from other people like Liberty Mutual talking about migrating some of their existing applications. So you definitely, there are lots of people who run existing applications. But, again, what we like to highlight most often are the newer, brand-new applications that

get huge, huge improvements. And they really focus on kind of like our high-level pitch of things, is that once you put all of this cloud-native stuff in place, the idea is that you get, as I like to say, you get a more programmable business, where you can start to do new business models and new things. But, yeah, a huge part of it is making sure that your legacy stuff doesn't drag you down. And there's plenty of that scurrying about in cloud land. That's a great point. You know, then the sort of sister argument to that is, but, hey, these apps are kind of Cloud Foundry-specific. We sat down with an analyst a month or so ago, and it's somebody who we're friendly with, and we do a lot of briefings. But this person still thought, like, hey, I build a Buildpack app, and then I deploy that Buildpack app to Cloud Foundry. I think, wait, wait, no, wait a second. These are just apps. Like, start new Spring Boot Project, start new Node.js Project, .NET Project, and then literally just push it. Like, the Buildpack is actually how we take your source code and kind of create the necessary, you know, pull in dependencies, lay down the app server. Like, you build the little blob that becomes the container. But that's always just, that's build process stuff. Your app is just your app. And so that sometimes gets confused that Cloud Foundry injects some sort of weird hooks in there that lock me in or make it really hard to move my app elsewhere. Like, bring a Docker container. Bring your source code. If you don't want to have any specific CF hooks, you don't need them. Almost just like dropping things into EC2 in AWS. Like, I could just throw my generic app in a virtual machine, use no first-party services because I'm scared of lock-in or things. Or I say, let's use some of these first-party services and take advantage. It's all up to you. So Cloud Foundry is opinionated, but at the same time doesn't impose too many things on your app that keeps it from going elsewhere. Yeah. And, you know, we referenced him earlier. Like, if you ever see a Josh Long talk, like, I mean, it kind of follows the line of what you're saying is he just basically, I think he has IntelliJ installed on his machine. But he basically just starts from scratch. And if you know his talk, he tells you multiple times that you just go to start.spring.io.

And everything that he builds is extremely portable and, you know, based on very open spring things. And then, you know, I don't know if he ever actually gets to the point, much to the sort of, like, market-y mind hat that I have on my hat rack, much to my chagrin. He doesn't always get to the point of running it in Pivotal Cloud Foundry. But, you know, obviously running Spring Boot applications at Pivotal Cloud Foundry is great. But there's plenty of things that are extremely portable in ways of developing applications that aren't some weird, funky thing. And then also, admittedly, I haven't, like, looked at build packs in a long time. But as I recall, when I looked at them, they're basically pretty generic manifests that, you know, again, follow a 12-factor sort of externalized configuration and dependency mapping stuff to externalize it. But they really just sort of, like, point at your source code and your application and talk about the services that you want to use. And I would suspect it's not that complicated of a thing to, like, convert what you have in a build pack to whatever other format that you want. They're extremely lightweight. And then, you know, so it's not necessarily that they're, it's, this is, you know, another point is lock-in, which is, like, a whole other rant. But it's sort of like, of course, if you want to move your application to a platform that doesn't use build packs, you cannot use the build packs. But it's very easy to translate them around and move them. They're just, they're just little YAML files. Well, I mean, and a build pack just, you know, it's a set of scripts that more or less define what should your code look like in prod when it packages it up and says, hey, look, here is the Tomcat server or here's, you know, if I don't have it embedded, here's the different things I need to actually run this thing. It's late binding. It's saying, do I want a developer to necessarily build the container or do I want the platform to? And those are both your choices. You just choose which level of responsibility you want to take. So none of these impose anything on your app. It's just, it's fine. It's something that people have kind of gotten used to now that we want to make sure people see it's extensible as well as it's opinionated.

One of your favorite things being an MVP type is the .NET support, right? That's right. So, you know, that's another one where, I mean, historically, coming from the .NET background myself, there weren't a whole lot of places to run Windows apps, mainly because licensing was such a pain in the neck for everybody that every platform would just advertise what they did with Linux. Like, hey, run Ruby, run Node, run Java, you know, and that's it. And so you would use Azure or you would use some really sort of niche providers that could do Windows stuff. Now, Cloud Foundry always, for years, actually had support to kind of bolt on Windows. And we had done that for a while. And then .NET Core came out, which lets you run kind of .NET apps on Linux, which is cool. And many vendors now tout their .NET support when really what they mean is .NET Core support, which is a very small percentage of developers so far. But then, you know, what Pivotal has added is this is full-on Windows support automated the same way you do Linux. So that if you have .NET framework apps that run on Windows, fantastic. You have .NET Core apps that run on Linux, awesome. You have BuildPack support, you have OS support. And so this is arguably the only platform that you can put on any cloud that's fully Windows or Linux compatible. That's nice. At least it gives you some choices as you try to modernize some of your existing .NET apps, which, as you know, Coté, I mean, most of our customers are nicely split between a whole host of Java app, .NET apps, and then weird commercial stuff that defies classification. You know, my thinking is over the next year or two, we'll see a huge amount of .NET applications in the cloud native space. I mean, definitely, definitely in most of the customers I talk to and hear about, as you were saying, they all, one, are really interested in that. And it's almost like they have this sort of like green room, if you will, this waiting room of .NET applications that are just eager to move over once they get comfortable with it. So that'll be interesting to see how that pans out. And hopefully it's the continuation of, you know, what we've seen at Microsoft as a company for the past several years, which is like, to put it in my own words, their summarization

is like, we'll run whatever you have. As long as we're involved somewhere in the stack, we're happy for it, which is fine, right? I mean, that's what you would want ultimately from a company like them is that they want to fit into whatever IT stack the way that their customers want to, instead of imposing a, you must use all of our stack sort of position. That's right. And that's actually a nice lead into the other one. I, you know, at least the last of my baked in ones, we can throw a couple of my bonus ones in if you'd like. But the last baked one in was sort of that story itself is, you know, Azure, Google, Amazon offer a ton of great services. And so the myth could be, if I have that, what the heck do I need a platform for? Like, those are kind of platforms. Isn't this just excessive? Are they just doing turtles on turtles or whatever the heck the analogy is? You know, and I think, you know, our answer is sometimes you do want to have that sort of application-centric construct that is taking care of security throughout the stack, the dev experience, and you want the portability, which is nice. But for our customers, many of them run on public cloud. You know, we are a giant partner of Amazon and Microsoft and Google. So that's great. And our customers see the value of laying a platform on top that is app-centric, that does not expose the developer to any sort of leaky abstractions where they're figuring out and fiddling with DNS, or they're updating server OSs, or they're patching things. Instead, they just literally see a push, and it doesn't matter if their target is on premises or in Google Cloud, the dev experience is the same, and we continue to add on services and availability, all the sort of Cloud Foundry value. There's a lot of value, even if you're using native first-party cloud services. So there's plenty of good case studies we could talk to people about. But we see tons of great scenarios here where it's additive. It's not an or. Yeah, yeah. And I mean, that raises the other thing I would talk about. So first of all, I think, yeah, it reminds me of, this must have been from last year sometime. I remember I was, we were in, I don't know what this is called, but like the Air Force

had brought in, it was like a public, have vendors come tell us how to, or advise us on how to solve our problems. And one of the interesting points they brought up, I mean, you know, being, being the Air Force, they had tons of stuff, um, is that they wanted to establish a common application development, um, and development's the wrong word, a common application platform or, or way of doing things across all the various types of infrastructure, private and public or whatever that they had. And, and, you know, I think obviously this opinion is self-serving, but I think something like Pivotal Cloud Foundry and more broadly Cloud Foundry provides like this, uh, common mindset of doing things in the same way that, uh, the, uh, the, uh, the duopoly, if you will, or, or the, what emerged in, in over the past 10 or 15 years is that most everything was in like, I guess it's a triopoly was basically like Java, uh, .net or PHP or Lampstack. And there, you know, those are just very common ways of doing application development. So instead of having 20 different ways of doing it, you had at least those three. And I think one, uh, it's nice to have that common way of doing your application development and operations across whatever type of cloud you want to run in and therefore having the support of, of the, uh, let's see, we got like three public clouds that we support, if I recall, and then, and then basically on OpenStack and VMware. And, uh, I think that's about it that the infrastructure that you can run on. So wherever you're moving this stuff, you can, you can move your, you've got the same application model. Uh, I struggle to use the right phrase for that, but like, that's a nice amount of portability. And also, um, especially like in the case of, you know, the air force or other big customers, it, it hopefully gives you a, uh, some long-term risk management as far as over the next five or 10 years, that, that programming model will be supported and evolve. And you won't be sort of like stuck in a cul-de-sac of, I don't know, using natural or ADA or, or, or, or, you know, cobalt or whatever. So that, that's one thing of, uh, that's nice of having this separate abstraction layer rather

than natively using, uh, the cloud things. And then as you brought up and admittedly, I haven't seen a lot of people doing this, but I think, I think it'll be an interesting, uh, option to see over the next few years is in theory, you can also start to mix and max all mix, mix, I guess you would max it, but mix and match all of the, uh, all the various services that you have across the public cloud. So we've talked about this here and there and on the, on the podcast, but like, you know, if, if, if you like the way that Google does all of their data analytics and big data things, but you want to run an Amazon or Azure, you know, you could combine those two things together and have a common, uh, sort of substratum for all your microservices stuff and not have to put up with, with learning and managing all the different ways of doing that. If you've got this abstraction layer. Or the scenario for that works well, you know, our friend Tony from Home Depot, who we had on the podcast last December talks about, look, they're using Google cloud. They use working on premises. And what that means is that you can also potentially do local development that uses a Google machine learning experience, right? And then when you push that same app using the exact same pipeline to Google cloud, the app obviously works the same. So it also gives me access to the services in every place, even if I'm developing it and starting on prem or moving to cloud, like nothing changes in my deployment pipeline. That's a really powerful thing for shops that just want to get good at software and don't want to be awesome at the infrastructure pipeline stuff. And then finally, I mean, I kind of alluded to this earlier and we've talked all about it, but just to kind of round it up into a topic, I think, I think it, it, uh, it kind of gets to this idea of lock-in, which is a very, uh, a very common anti-pivotal, uh, message that, that, that we encounter. And, you know, lock-in is like, for me, uh, having followed all of this stuff for, uh, like 15 years or so, it's very much so a Rorschach test of like what it is the person worried about or talking about lock-in is, which is, it tends to say a lot more about the position that people have, uh, than sort of like the, uh, the reality of stuff, which, which sounds like some, uh, some vague allegations or whatever allegations, but some vague

commentary, but to make that more specific, right? Like, uh, you know, there's several layers of lock-in. One of them is, uh, you could call it like, uh, back, back in the, uh, in the Sunday, Simon Phipps had a good way of rephrasing lock-in, which was the freedom to leave. And the idea was how difficult would it be for you to take whatever it is you care about your data or your applications and, and move them from your, your current stack to a new stack. And so that's one way of rating lock-in is like, can you move these things, right? Like how, what kind of effort and risk would it take? And the interesting thing about that positioning is it has nothing to do with open or open source. It's more about effort to move things, right? Which kind of gets to my second way of thinking about lock-in, which is like, you know, you should always ask someone who's telling you that some rival platform is lock-in if that means that if you can't, you can't move their applications to whatever it is they're selling. And I would bet they would be more than happy to say that you could move whatever locked in applications you have to their platform instead. Otherwise they probably wouldn't be there trying to sell you essentially. So, you know, that's a good way to test out like freedom to leave essentially. Now, of course there's, there's all sorts of ways to counter counter argument that, but it gets to the point that like a lot of lock-in discussion should more be about as the word we keep using portability of things and analyzing what the portability of an application is across any platform that you might have. And then, and then, and then, you know, the other, the other lock-in thing, and we've kind of highlighted in the, the pivotal cloud foundry world, how, whether you're using an open source thing like the spring framework, or you, you have, you know, you can have a very heavy touch series of build packs, or you can have a light touch series of build packs, but the portability of that stuff is as portable as you want it to be. Right. Like you can directly rely on and integrate to public cloud services, like we've mentioned, or you could directly integrate with proprietary sort of pivotal cloud foundry things or not. There's all sorts of options that you have.

And that, you know, to kind of, uh, uh, uh, a way of when one of the better ways of thinking about securities is essentially like, you've got to do a risk benefit analysis of should I lock into these proprietary things, which to be clear is not the only option you have in pivotal cloud foundry land or, or public cloud land. Right. Or should I take the time if it does take time to be more portable in the way that I'm doing things? And, you know, it all depends on what your needs are and, and, and how things are there. But, you know, the upshot of, of the conversation I tried to have with people. And I remember back in the red monk days, we would discuss this a lot too, is like, you're always choosing a type of lock-in and just be aware of more of that idea of freedom to leave and portability that you have. And think about like, what kind of, I don't know what the opposite of lock-in is, but think about what it is you need versus, uh, what sort of, uh, what sort of commitments, if you will, you're willing to make, whether it's an open source thing or a proprietary thing or, or anything like that, because whatever platform you choose, there's a, you're, you're going to limit your portability to some extent. And it's just good to, uh, analyze what your needs are and all of that going in. Yeah, it's well said. I mean, everything from owning a house to choosing Red Hat Linux, I mean, they're all lock-in at some level, right? I mean, everything has, everything has a switching cost. And in some cases you're willing to take it. There's a reason that, you know, AWS Lambda is still taking off, even though you can't move any of that stuff anywhere else, because it solves a real problem. And people who are using it are going to accept any risk they take, because for now it can add differentiating value and it is worth whatever pain they think they will face later. When we were hanging out with, uh, one of the Gartner analysts a couple of weeks ago, and he mentioned, we were talking about lock-in and he said, look, lock-in bites people when there seems to be either steady or declining value, but increasing cost. And so if that Oracle database continues to cost me more every year, hypothetically, and yet I feel like I'm actually getting less out of it, that's when lock-in starts to bite me. If something is awesome, I rarely am worried about what my lock-in cost is, right? I mean,

you don't care that you're locked into Austin if you love living there. If it was a complete train wreck of a city, you would worry about wanting to leave and think, my gosh, now I own a house. How do I get out of this? So nobody cares about lock-in when that thing is adding tons of value. So pick things that you think will add value over time that should appreciate in value over time. And hopefully PCF is one of those things that gets better every month and your costs only grow as your business does. Those are hopefully good criteria to use to pick these platforms. Yeah. Well, why don't you, uh, why don't you pick one of your bonus items and then, uh, and then we'll close out with other two, uh, self-serving news things and then, uh, and then we'll be done. Yeah. Yeah. Yeah. You know, I'd list a couple other things for Coté's fun and amusement, whether it's pivotal web services, not being production grade, like people run production apps there, enjoy it, love it. The only other one was operating systems. Do they matter? You know, do I care what OS runs Azure? Do I care what OS runs Amazon? Do I care what OS runs PCF? No. Like we use Ubuntu and we have this crazy hardening process to make sure this is a rock solid secure OS under your platform, but it really doesn't matter if you love and have certified again, Red Hat Linux or using any other Linux distribution, like it's OS it's embedded. It's embedded in the platform. It's updated as part of the platform. Technically the OS doesn't matter when you have a properly converged platform that takes care of underlying automation for you. So sometimes we can get overly excited about the things that run under platforms, but at that point, you know, it's voyeurism only. It doesn't really matter. And speaking again of, of my red monk days, I remember back then there was this sense that, uh, I think, I think I, and other people that are used to phrase it this way is the operating system is, is becoming a piece of middleware, which is to say when the operating system matters, it's becoming a piece of middleware otherwise. And, and, you know, I was just looking at some surveys this morning, but I think pretty consistently when you look at, uh, operating system use by use instead of revenue, uh, it's typically some Ubuntu variation, uh, that you see out there and maybe core OS shows up a lot, but, but like, it's just Ubuntu rules everything for, for whatever reason. I think it's

because, uh, historically and probably currently they're the most developer friendly, uh, operating system. And also, I guess, per canonical strategy, the most sort of liberal, if you will. So those people who developers who build platforms tend to gravitate towards Ubuntu and, and, and make that choice. And it's not saying it doesn't matter, right? It's saying it matters to fewer people. And so I think fewer people should care about the underlying tech. We say the same thing with container orchestration or OS or these things, they're super valuable tech. It's just a smaller population than currently should probably be caring about it. Yeah, exactly. I think, uh, I think that's the case. Well, as promised the, the other two items that, uh, that I wanted to go over is we have, uh, before we wrap up, we have cloud foundry summit coming up, uh, June 13th to 15th. I guess that's officially next month now, right? Since it's, uh, May 1st. That's right. And, uh, and I I'll be doing a few interviews, which I, which I'll publish here with, uh, let's see, Greg from Comcast and I think Opal from, uh, from all state and, uh, just kind of going over there, both speaking there. And so we'll, uh, we'll, I'm just going to, I don't know, talk with what's going on at Comcast and all state and what they'll be talking about, but it's a good, it's a good conference, uh, both from a technical side and a, uh, you know, business side. Cause as, as, uh, as we'll talk about with those two and, and, uh, and as you'll see from others, a lot of the discussion at CF summit is about how organizations are using, uh, cloud foundry and, and being the cloud foundry summit, it's not just about pivotal stuff. It's about all the cloud foundry variants out there, how they're using it to get to that idea of a, of a programmable business. And it's one of the good momentum, uh, sort of moments you get to, I, I, in preparation for talking with Greg, I, I pulled together some highlights of all the, uh, most all of the Comcast public talks, uh, that I could see. And I'll put a link to that, but you can get a sense of the kind of, um, I don't know, momentum and lessons learned, um, that goes over there. There's, I was reminded of a particularly good talk that, uh, not the singer, but, uh, that James Taylor of Comcast went over at SpringOne platform,

kind of talking about, um, what I would call the business it alignment philosophy and thinking. And he has, he has a, a pretty good rant and a structured talk about that with some good examples from, from, uh, Comcast, but anyways, uh, and there's also a bunch of technical stuff. If, if you want to go over that, you can tell that I focus on the other thing more. Uh, but if you're interested in going that you can use the code CFSV17COTE, which is a lot to remember, but it's basically Cloud Foundry, Silicon Valley, 17 Coté, right? CFSV17COTE. And, uh, I'll, I'll put, I'll put something, a link to that in the chapter notes. If you're, if you're looking at the chapters in the podcast and the show notes, but you can get 20% off when you register with that. And then finally, as always, we have a tremendous amount of, uh, books available, the, the O'Reilly books and some early editions of like cloud native Java and, uh, another cloud native programming book that Cornelia is working on. And, and the one that, uh, that we just published, uh, the booklet that I worked on, uh, or a pivotal site. If you go to pivotal.io slash eBooks, you can get all of those. I think you only have to lead in yourself once. And then we have magic digital marketing software that remembers who you were. And then you can just go, uh, download all of those and, uh, check. That's right. And it turns off your sprinklers. Hopefully. And, and so, uh, yeah, hopefully we'll see you at some of those events, uh, that we went over. And, uh, if you're, if you're at DevOps stays Austin later this week, or Ozcon, you come by the pivotal booth and we've got all sorts of gigas to, uh, to give out and things like that. And as always, uh, this has been Pivotal Conversations. You can get the latest episode, uh, for this by subscribing to it, which you can find in iTunes or overcast. And I think in stitcher or wherever you get your podcast. And if you're more of a traditionalist, like I am, and you like to get the RSS feed, you can go to soundcloud.com slash Pivotal Conversations, all one word, and you can find the RSS feed there. And you can find all of our past episodes and listen to it in the web. If you're, if you're one of those folks. Uh, and then

finally, uh, we trail sort of like a week or so after we post these, but if you go to pivotal.io slash podcast, you can see all the show notes that we have. And as mentioned, I think last time and a little bit here, I'll, I also go in and add a chapter marks into the podcast. So as there's things we talk about in the podcast, if you, I think only overcast supports this now, but you can open up, uh, the, the little listener there and you can see links to things we're talking about, uh, for that as well. And with that, uh, thanks for listening and we'll see everyone next time. Bye bye.

Tanzu Talk · 🍕 Get more pizza.