It’s private cloud all over again (Ep. 71)

Episode 57 · 51:25 · SRT · archive.org · all Tanzu Talk

Hey there. I wanted to tell you about a new podcast we got going over here at Pivotal. It's called That Moment. Don't get confused like I was initially. It's not the moment. It's called That Moment. That Moment explores the pivot that changes everything, moments that open doors for discovery and growth, but also brings the looming possibility of failure. Each show features different leaders and innovators sharing their stories and taking risks in business life. It's, you know, classic Pivotal stories. Also, it's fancy with all the crossfading music and optimized editing. It's really some good pro stuff. So if you're interested in those kinds of stories about how people are wangling their way through all this digital transformation, DevOps cloud native stuff, go subscribe to That Moment wherever you get your podcast. But, you know, for your weekly dose of unprofessional old school podcast rambling, let's get on with the show. So do you, in your household, it's well-established. You got lots of rain slickers, probably hiking boots, but do you have a juicer? I don't. No juicer. I haven't really totally embraced the Pacific Northwest yet. I need to make my own juicers and whatever avocado spreads, things like that. Oh yeah. Well, well, you got to watch out with those avocado spreads. You might not be able to afford a mortgage. That's what I've read. Thorough analysis from down under. Yeah. Well, no, but I was at my mother's this past weekend and, uh, she has one of these gigantic constructs. I, I, I even, I shudder to think how much it costs, like a big silver thing. It's a wonderfully crafted machine. Uh, I forget the name of the brand or whatever. But, uh, so, so I was there, uh, uh, on Sunday morning, which is yesterday for us. I went and bought like a bunch of, uh, you know, carrots and fruit and even kale all in. I also bought probably like three pounds of chicken. It was only $26 and I bought like organic stuff. Did you put the chicken in there? Well, I did ask my mom if she had ever juiced meat, but I didn't experiment with that. I, she also has one of those Traeger grills. So that's where I put the chicken to kind of smoke it. I mean, if you, I still need to experiment. If you were pioneering a chicken smoothie, I mean, I think that was going to warrant some discussion here. Yeah. Well, you know, there's only so much

innovation I can do every year and, uh, I don't, I don't want to push the boundaries, but this is just, it's just this phenomenal thing. This juicer, you just like, uh, you just like put a bag of carrots in there and you got carrot juice. It's, I don't know. I, I, of course, my wife, uh, Kim used to be a big juicer person and, uh, had a juicer. So I, I asked her, I was rambling through a question and I was like, all right, here's what I'm really asking. Is a juicer the kind of thing that you get and you use for like a month and then it's just in the way I, you never use it. And she also short circuited the answer and said, we're not going to get a juicer. So there you go. It'll sit right next to the bread maker you have in the house. Yeah. Yeah. Yeah. You know, you've already got that situation with, uh, with, you know, at least under our cabinets, right? Like there's, I think there's a food dehydrator underneath there. And, uh, we got, we got a, we got a full on Cuisinart. We've got the mini Cuisinart. And then, uh, we have one of those, um, what were they popular a couple of years ago? Bullets or something like that. Ninja bullets. I don't know what they are. And then, and then we also have, I bought this, my wife likes pickling. So I bought her this Japanese pickling Tupperware thing, which I think she's forgotten about. I need to get that out and pickle things. That's awesome. Yeah. Making beef jerky and pickles. It's a, it's a party at your house. But you know, you know, not unlike buying a private juicer, there's been a lot of private cloud news recently. Well, a lot, I don't know. There's a lot in your list. And I'm, I'm, I, I actually was, uh, on my other, uh, podcast software defined talk, we were talking about, uh, a couple episodes back, like what it would look like if, if Amazon had a private cloud and many people, including my cohost reminded me that Amazon does have that snowball thing, uh, which, you know, I had thought that it was like a big semi truck, but I saw another article from maybe last year that there was actually kind of like a desktop size device. Maybe, I don't know if it's a rack or whatever that, uh, maybe it was something else. And I remember reading to that article and it's basically, uh, you probably, you know, you're more of a details person than I am. You probably remember the details, but it,

you know, basically has like a terabyte of storage and like probably an S3 interface and Lambda in it or something like that. But that was a, uh, that was a, I don't know, an interesting thing that's out there. It'd be, it'd be fun to go talk with some, uh, whoever would know this, like how that's doing. I shouldn't, I should talk. Yeah. I mean, I think that's more storage stuff. I mean, they have the, uh, totally, uh, they have the gov cloud, which is kind of private cloudy and they have VPC virtual private cloud, which is kind of private cloudy. So I think, you know, they'd be grudgingly embraced the reality that, that most enterprises. Yeah. And so I, I bring that up because I think, I think, uh, well, I, I was most heavily involved in, as I like to joke, when we call it the cloud wars of like, uh, let's call it 2009 to 2013. I, I, I worked at a, uh, well, I worked at Dell doing cloud strategy and software strategy. So I spent a lot of time looking at different options and things like that. And, you know, that I think, I think, uh, I think, um, I think, I think what we have in front of us here in our show notes, including my Amazon one are, are, I guess there's, there's at least four ideas of private cloud. The first one, I forget what, uh, what rhetorical trick this is. You always put a, uh, a pig without lipstick out there just to be eliminated. But the first idea is not doing private cloud, which is always very popular that the old, uh, the old no private cloud or whatever. So we'll just set that pig aside, make some thick pork chops out of it, put it on the Traeger and enjoy some freshly squeezed watermelon juice. You know, on that note, there's a lot of water in a watermelon and it doesn't taste as good as like the watermelon juice I get at the store. I don't know what, uh, I know what the difference is, but, um, anyhow. So then the other one, like, I think, I think, um, I think, uh, the Azure stack, which is supposedly coming out in September probably is, I shouldn't be so dismissive. Uh, that seems to represent like the, uh, uh, you could almost call it the classic idea of a private cloud. I mean, I don't know, you'll have to, you'll have to correct me if you've read through the details, but as I've been reading through it, it's basically like, here's Azure on your own hardware, just sort of like, and it's compatible and it's maybe it doesn't

have everything. Like, I guess to do really fast and effective, like, uh, whatever the machine learning and stuff is, I mean, that's big parallel processing. So you probably need that, you know, that's hard to have in a three node thing that they start with. But anyways, that seems like the first model, which is like almost an exact duplication of, of what you would get, uh, what you would get off premise on premise. Is that, is that your impression of what's in, in a Azure stack? Yeah. I mean, the Azure stack stuff, I think is like, look, I want to run this, but it'd be to your originally when it first got announced, it was, Hey, run this on your hardware. And then in the link, you know, I think that we have in the show notes and others, you know, Scott Guthrie, the team there talked about more like, no, this should be a converged appliance. Cause some of the reason that things like open stack struggled is they were just so many pieces that people struggle to weave together. So what you want is a box from Dell that just has the software on top. And then you pay us per hour for VMs and apps, just like you would in public cloud. So yeah, it's just kind of like, here is cloud in a box, which you hate that term, but that's, that's what this model kind of is. Yeah. Yeah. And that cloud in a box thing, like it's always, uh, it's always frustrating on multiple angles. I mean, one, you know, the, uh, the, the hipsters of cloud, everything's frustrating to them, right? Like they, they just, they, they, they don't want all the olds coming in and like defining something that's actually useful and that works. So, you know, that's that, that seems to annoy them to no end, which is fine. You know, I get it. Uh, here I am juicing stuff, right? Like instead of having my, uh, sunny D, but the other frustrating thing is like, it must be like technologically and businessly, is that a word hard to actually make a cloud in a box? Because I don't think it's ever really been done. It's never been done. Like after all, I mean, that's, I think that's, yeah, yeah, exactly. The hipsters probably say like, we've tried this. I mean, this is Microsoft's third try at this. Like this has not been successful to some extent, because a, it feels like there's a lot of promise. But then when you look at the caveats, it's not worth it. And then you still have to figure out how to update this software. It's not as transparent as the public cloud. And what happens when you have scale challenges and all the promise

sounds fantastic. And then the minute you start trying to do it, you realize, meh, I might as well just keep doing what I'm doing or just use public cloud. Yeah. Yeah. And, and, you know, one, one theory, which, uh, I don't know if I agree with it or not, but like I, I pulled out this quote from, uh, the horror mentioned Scott Guthrie. He's an interesting character to track his career. He'd be a good, uh, if there was enough interest in the world and this kind of thing, he would be the kind of person you would write like a, uh, like a biography about that was a way of telling the history of, of contemporary tech and how, how it works out. But anyways, uh, like he has this quote basically saying to summarize it, that like, uh, when people build their own IT, they tend to want to customize it a lot and that's not going to work, right? Like we're basically going to have like more or less a, uh, again, to use all my own wording, like a black box and you have to work with the vendors to have it all integrated together. And that makes a cloud notion work out. And that seems like, um, I mean, that would be one backward looking theory I would have is that each time, each time people have tried to make a cloud in the box, either they allow too much customization or the IT departments buying them want to do too much customization. And, and that, I don't know. I mean, I could see that that could pan out certainly in the instance of, um, all of the Pivotal Cloud Foundry installs out there. Like, I don't, I don't think there's a lot of room for variation in, in, in the stack, right? Like, like it's basically like whatever infrastructure pools you have out there, just give, give, uh, give Pivotal Cloud Foundry the credentials and slowly walk away. Just like allow it to take over. Yeah. I mean, that's, I mean, in my experience, most private clouds start as IT projects. And the problem is that it does not take into account what public cloud got, right? Which was a good developer experience. And so there's always been that disconnect. And frankly, no private cloud will ever be successful unless you get the software right. It doesn't matter what your infrastructure looks like. It doesn't matter where you stick it. If you do not put the right interface on top, you don't have the right abstractions on top. You're just wasting

your time, at least in my, uh, hot take opinion. Yeah. Yeah. Yeah. And that, that was another thing. Like even back when I was in the, uh, the old cloud wars was the, uh, I, uh, I mean, I think it was a good thing, but I spent a lot of, uh, failed attempts to try to talk with everyone about like, what's the point of using this cloud stuff, which, which is a stack atop of what you're just saying is like the software that you're actually running on this cloud, the workloads, as they would say, are incredibly important to think about because that will drive what it is you're going to be doing with cloud. Right. Like if all you want, like this was a very, uh, there were a couple of popular notions back then that everyone had. One of them was, I don't think they thought of it this way, but it was just like cheaper virtualization. Right. And this is a very like, this is a very like nineties, 2000 view of what, what the it department does or what a significant part of the it department, which is, uh, I have no idea really what applications you're running on top of me, but I provide these SLA. It's a very SRE kind of thing. I provide these SLAs of availability and costs and uptime of all the infrastructure underneath it. And so a lot of the cloud thing I would come across was like, I want to do that same thing. And then I don't really know what people are going to run on top of it, which for storage and things like that maybe is a good idea and stuff like that. But I think it's, it was always troublesome for just shifting your, uh, your off the shelf software onto it. And then the, uh, the other thing I would encounter a lot would be, uh, bursting, which, which is sort of like, which I don't know, I haven't really done a, uh, anything but my usual, like, uh, thinking in the shower at three and when I can't go to sleep analysis of bursting. But I think it was just a variation of like, I don't want to pay a lot for that muffler. Right. It was sort of like every now and then I need more capacity and I'm just going to magically burst into the public cloud to use capacity. Well, it's almost like the whole private cloud value prop. Again, it sounds brilliant. And then you realize like, I have no app workloads that can magically scale geographies with and copy their state and run functionally. Like what

is SharePoint going to cloud burst? Like what did you have that was going to magically work 4,000 miles away in someone else's data center by flipping over a load balancer? Like you don't have those apps. So again, that's, that's, that's exactly, that's exactly the point where we're both making is like the software that you're running on top of all of this, be it the, the platform abstraction layers stuff that you were mentioning or the actual applications that we were both talking about. Like that's, that's, uh, that's both key to needing to do some sort of cloud thing. And then also key to like it working. Yeah. I mean, you and I have seen, Coté, if you look at our favorite analyst firms, you know, survey after survey shows 50%, whatever the number is of shops saying they have a private cloud. I think in my mind, 10% of those shops actually have a private cloud. The rest of them have a virtualization layer or maybe some sort of cloud management platform on top. But look, I mean, private cloud might be getting interesting again because we finally are figuring out the right software abstractions and sure we're biased. And I say, if you stick pivotal cloud foundry on top of open stack or vSphere, you might finally have something useful. And that's maybe we're actually at that point where location can matter less. And I can drop an Azure Azure stack pod in my data center. I can use these other things because the software has finally delivered the sort of experience that developers care about. Yeah. And to that end, I mean, I think, I think it'll be fun to see, uh, uh, what happens with Azure stack, right? Like, I mean, I think, I think there'll definitely, well, without a doubt, there'll be many use cases where it works because otherwise it probably wouldn't be released as a product at this point, but you'd be surprised, but yeah, no, I agree with you. Yeah. No, I, yeah. What, what, what I mean, what I mean to say is I think, I think in this area over the past, what's this year over the past 10 years, I think Microsoft has built up a lot of credibility in a release like this, right? Uh, now, uh, that will also be interesting is what you were talking about as if, as if the notion is if you get enough feelers that the notion of private cloud, like makes sense. And the, in the, the simple classical

definition, right. And that's what, I mean, we, we see, in fact, if I was looking at the stats recently, like the huge majority of our customers run in private cloud mode, I'm putting that in air quotes, whatever, whatever it may be. Right. Uh, and like that seems to be working out for them. And so outside of the, the sort of cloud foundry world, uh, it would be interesting to see how that applies more broadly. And then to the point of analysts covering it and people being confused about it, like it could be, uh, it'd be fun if that was very well defined and actually worked out the way people wanted. And, uh, it's, it'd be an opportunity for, uh, all sorts of interesting shifts in the topology. Absolutely. I'm just stringing together a bunch Yeah. I mean, we all thought, I think even four years ago that the, it was an inevitable march to public cloud. Like there was just no, and I'm, again, I'm a giant believer. I think that's where it makes a ton of sense, but if you do can get it right on prem, there's enough reasons to say, look, I'm going to take advantage of the assets I have. I'm going to take advantage of my own geographies and things like that. Like that's not terrible if you get the software right. We might actually be there. So, so then that brings up, uh, in my, in my, uh, categorization off the top of my head, the, what were we on the fourth sort of thing? And that, and, and, you know, it's kind of related to, uh, Google's announcement around some BPC stuff they have going on. And you do see, I've, I've heard many fanciful terms for this. I used to call it the public private cloud, but essentially, or a managed service managed, a managed cloud is another way of thinking about it, but there are all sorts of schemes. One of the things private cloud might be is like, it's private by network isolation and having your network all isolated and closed off. So you don't, in this, in this public pride scenario, public pride, public private cloud scenario, you don't actually run the data centers or, uh, update them, but it's not like all your stuff is, you know, on the public cloud. It's more isolated. It's like an old colo sort of situation. And, you know, plenty of companies that seems popular, right? And they still, yeah. So yeah,

I mean, a couple of years ago, if you remember, there were a spat of companies offering Heroku did this others where they actually were doing some physical isolation of assets and kind of calling it virtual private. And that was interesting. And then it kind of slowed down, but you know, what Google announced with their virtual private cloud, it's, I'm impressed because that's a hard thing to replicate with your own private cloud. The VLAN stretching, they're doing some of the cross-site networking, the tenancy they're setting up, the expandability of your subnet. Like they're just doing some really cool stuff that would be hard to do in your own private cloud. So if I can use your model four and say, look, this thing is still either network or physically isolated or both, and I can really configure this thing quickly, gosh, that adds a ton of value. So I mean, why do it yourself? Yeah, yeah. It is like, uh, what, you know, I, and, and it brings up another thing I'm, I'm kind of stuck on when it comes to like cloud theory, which is, I mean, I have to assume after reading like the, uh, you know, everything we know about Google and other public clouds, and also kind of reading the SRE book, like a lot of, a lot of, and I'm generalizing here as, as we talked about earlier, but a lot of public clouds, uh, or cloud stuff starts with, uh, Hey, stop tinkering, right? Like don't customize stuff. Like you're only going to have like so many t-shirts in the drawer, sort of like the classic, I don't know if, if, if the, uh, the females in the tech world are like this, but there's a subculture of males in the tech world, I guess, starting with Steve jobs, who basically have one outfit and they have 50, 50 copies of it in their wardrobe. And, uh, you know, it's sort of like, that seems to be the cloud approach is if you want to get all these efficiencies, you, uh, you, you can either think of it as standardizing or constraining, which are kind of the same thing. And, uh, you know, again, across those models, it's, it'll be interesting to see how Azure stack deals with that. And then what the, what the acceptance or kickback is from it. Right. And then in something like a, uh, your public private cloud or virtual, you know, whatever, like that stuff's inherited in there. Of course, that's how a public cloud works out. And, uh,

that's certainly like in our world, I feel like, I don't know, you do a lot more like, uh, sales help and, and, uh, competitive stuff than I do anymore. But like, uh, there's, there's probably a low double digit percentage of objection questions that are basically like, your point is moot, right? Like the thing you're asking about made sense, but it's sort of like, uh, it's like, I, I made this old cartoon way back when I remember those talking dinosaurs and they're talking about one of them is talking about like, uh, here's this new horseless carriage thing I call a car. And then the other dinosaur is like, yeah, but where do I hitch the horses up to? And like, it seems like there's a, a fair amount of where do I hitch the horses up to questions when you have a huge amount of standardization, but, uh, you know, I don't know. There you go. So, so what else, uh, what else do we have going on this week news wise, or I guess last week? One thing I, uh, added to the list was we announced the, uh, the formal project of spring cloud function. So this was something that our, uh, esteemed James Watters mentioned in the wall street journal, what like last December. And everyone was like, Oh, I guess we have a new product, which was awesome. So this is the actual delivery of some of that work. So what this is, is a Java framework library model for building more serverless, like Java apps. And what's cool is that there's a number of unique things in here. So if you do like using Lambda, right? This isn't a function as a service. This is not a Lambda replacement or Google cloud function replacement. This is the app sort of framework that actually runs the app. So it's got some cool things like, Hey, separating how you triggered the thing from the thing. So here's my logic. And just by adding a library kind of a runtime or kind of after I've built it, it would get triggered by a Kafka message or a rabbit message or an HTTP call. And so that's kind of cool to be able to separate that. And then Java is sometimes seen as heavyweight, but the fact that you can pass a string in to an endpoint and we'll actually then compile it into a beam that you can then execute later as a function. That's a very lightweight thing, right? It's not a whole big for that project. So we're doing some things to

kind of make Java more lightweight, more triggered on demand. It's a really cool project. It's not generally available yet, but you can now run the code. I've built a few projects with it just to try it out. So some really neat stuff. I think it's a cool reimagining of what it means to do on demand invocation of Java apps. It's pretty cool to see us innovating there. Yeah. You know, maybe you can tell me what this is, but I've noticed over the past three months to evoke them in kind of a joking way, all the hipsters are just like freaked out about serverless being awesome. They even have like, is this Jeff thing? What is that? Yeah. Yeah. Is that some sort of serverless thing or something? What's the story with that? It is. I don't know where the name came from. Someone will correct this, but yeah, I mean, it's hit a fever pitch. And again, I think it's great stuff. I'm just not yet on the bandwagon that says this is the entire future of computing. I'm just not there yet. Yes. Well, well, they'll do some further research. What's this Jeff thing? I'm writing that in the show notes. Yeah. And how come it's not called Larry? I mean, Larry is always there just waiting to be used. I'm right. I'm waiting for Jessica Conner or Larry Conn for something. Oh, yeah. And then also often dot net land. What you had some news there. There's circuit breakers in steel tone. That's right. Which again, if I said that statement to my parents, they would say, what do you do for a living? What the hell circuit breakers? That's right. As they say in other podcasts, did you just have a stroke? What does that mean? So steel toe is the dot net libraries that we've built. It's open source stuff that lets dot net devs connect to things like spring cloud services like, hey, use a config store, use a service discovery engine, use these connectors to talk to things like Redis and Rabbit, just kind of make dot net apps a little more cloud native. So one thing we've added, which had been a high request was hysterics. This is that Netflix library that does kind of, if you think bulkhead pattern, protect your apps, right? That if this app fails, kind of gracefully catch that and maybe use a fallback and, you know, return some empty results or something like that. So it's meant to protect your apps when you have these distributed systems. So we've added that

now to steel toe. So I can build dot net apps that have this sort of protection from fault, and then it can feed the metrics to the same dashboard the Java spring user would use. So I can see, hey, this circuit's open, or this circuit's closed. And that means, you know, we're using a fallback response instead of the default one. So just some really cool visibility into a really popular Netflix project. Now that's all part of the dot net library. It's neat stuff seeing that pick up. Yeah. And then the last thing. So, you know, we talked about Microsoft sales reorg the other day, and, and, and I finally actually went and read up about it. So I, you know, I'll put a link in the show notes, but I wrote up some little notes and other articles about that. And I think, I think it was, was, you know, from my perspective, like, what's interesting about it is to think about how the, how you can do some sort of reverse tea leaf reading that might be redundant about how, how people organize their sales force, hopefully, is I always, I'm always dodgy on the tail and the wagging tail dog situation, but like, it's good to have one of those, you know, the, the actual way, uh, the market works being, I think the dog, and then how your sales force is organized being the tail. I forget, but I think that's the right version of it. Right. And, uh, there's also, I, I found some, uh, some good IDC hardware, uh, uh, private public cloud, sort of like market share estimates, which those are always fun to look at because it gives you a, uh, I mean, you know, they're like five year into the future forecasting, but it's, it's fun to like see a hundred percent area chart about how something like private cloud and public cloud starts to eat away at, uh, at traditional infrastructure. If you're into, uh, didn't you say a while back that it's sometimes tricky to go do some retrospectives on analyst predictions from years ago? Like, I mean, it's interesting on these, like, what's the 2017 prediction from all the firms? Would we already have windows phone dominance and a private cloud everywhere? Like what's I, I haven't looked myself, but we should be doing some, uh, well, well, since you asked to have a mini,

mini rant about analyst land, which, which I, I feel like I have a, a permission card to do so. Uh, but, uh, you know, I don't, so, so the issue there is, well, there's two issues. One, um, over the course of like five year, 10 years back, I don't think the analysts do that. Um, and I don't know if IDC does this, but, but a lot of analyst firms will retire their research, which to me is just confounding. Um, but, um, which, which is to say it's no longer available. And so I think, you know, I think at IDC who I think they're the best ones for, um, forecasting and market sizing, just because, uh, I don't know, reliability and forecasting is not really a thing. I know that sounds absurd, but that's not the number one metric I would use. I mean, I would use consistency and the time that they've been doing it and that you have the same kind of people working on it. So you get a consistent view of stuff. Like I remember when I went into four or five, one research, uh, um, I replaced, uh, Rachel Chalmers who had headed up the infrastructure software practice. And so at some point very early on, I went to go talk with the four or five, one micro market sizes for the infrastructure software business. And my first notion was like, we got to completely redo this, which, which was just because it was my view of the market. Right. And that would have been terrible because then it would have, you know, there's no consistency, uh, across it. Anyhow. So, you know, I think if you dig, you can actually get a lot of historic data from IDC. Like I think if you get their black book product, uh, I think it goes back a long time and I think you can get past versions of it, but they don't really make it easy. Like you would want where you're, there's just some page you go to and you see like, let's rate how accurate we were, uh, about these things. And at IDC, at least within like a kind of like a two or three year period and especially quarterly period, they'll tell you when they correct a forecast. So they're kind of okay, but it's not that easy. And then like, uh, you know, Gartner might be a little bit like that with their big,

uh, with their, uh, they don't, they don't call it a black book, but they might have that data. But, but in general, like Gartner is the one that will like archive their research as if like the stuff isn't valid anymore, which to some extent it's not valid, but it's really hard to get a historic perspective of it. I don't know. Plus you got to pay for all that stuff, which, you know, I'm sympathetic to that. I think we can all agree on money. Uh, and so it's in, unless you have analyst access, it's, it's pretty hard to go do that. And I, and I think this does create an annoying situation. The, uh, uh, the aforementioned hipsters, which are sort of like the ones who would complain about this stuff the most don't actually have access to the research. So it's sort of like they're theoretic, they're complaining about something that they read about in blogs and Twitter and press releases, which I think I'll, I'll end my rant here. I think that was a, uh, I guess this was in like 2015 and 2016, there was a lot of complaining about bimodal IT and it's just like, well, there's, you know, like a whole bunch of Gartner material on that. And you should probably read all of that before you complain about it, but you don't have access to it. So it was like, uh, sort of tragic, not, not that, not that the complainers are wrong, but it was just more like you got to have a common pool of stuff you're complaining about. Yeah. I don't know the solution to all that, but when analysts are considered justifiably, so, uh, you know, a source of, of knowledge and authority in these spaces, when their research is behind a paywall, it can make it tricky when, you know, the numbers float around in the conference slide. And we're going to help it. That number come from, like, how did you get that? So I don't know how you solve that, but it seems like, yeah, you know, I, I, I was talking with some, uh, let's call them vendor side people recently in the cloud space. And, and of course we started complaining about analysts and everyone's always very kind. They're like, well, except when you did this work, Coté, which is fine. I don't know if that's the case. I go back and read my stuff every now and then it's not always so awesome. But, uh, you know, what I tell people about like Gartner, uh, is they have this thing called the Gartner technology professionals, which, uh, I think that's

what it is. I think it's plural. And, uh, that's the old Burton group. And I'd never, until I came to pivotal and had access to all of this, I'd never really read it except a stray pirated report here and there. Uh, but, uh, they actually produce, I shouldn't say actually, cause I'm not judging the other side, the, what do they call it? The, the, uh, the leader people or something. Something like that. Yeah. Like I, I think the GTP people, like they do, um, from the perspective of like companies, uh, you know, that are considering technology, it's actually the GTP people that do the, the, the maximal amount that you would expect them to, which is what is your advice on how I would use this software and what I would do? Like, that's all they cover. And they have these like 50, 60 page tomes on like, I mean, there's one on like, uh, you know, cloud platforms and there'll be one on Hadoop. And it really is like a very thorough analysis, uh, that you would look at if you were evaluating and even trying to figure out how to use these things. And, you know, they do that on the other side to some extent, like there'll be some advice about, uh, selecting things and shortlisting, like with the magic quadrant. But I generally, what I tell people is generally like industry analysts, a lot of people forget that first word, like their job is to tell you what's happening in the industry primarily and what the vendor sports are and the rankings and the trends and their job, their secondary job in order to inform that is always to kind of like know what you would actually do and how things work, but that's their secondary job. And so you can't really like, you shouldn't really go to them, only go to them for advice about what you would actually do. You should go to also people like the, uh, GTP folks and stuff. Well, we just have to get read on again soon and gab all analysts. Thanks. That's right. That's right. Hey, it's interstitial mid-roll stuff. I just wanted to promote a few upcoming pivotal things here through June, July, and August. We had the cloud native roadshow coming to all sorts of cities. This is a free day long event we do with Google that goes over what exactly cloud native is and how

our customers are using pivotal and Google technologies and approaches. The cities, and this is a long list. So get ready. The cities are Stuttgart, Dallas, Denver, LA, Seattle, San Francisco, Amsterdam, Seoul, Hong Kong, Sydney, and Singapore. There's a big list of dates that you can look at. You can check out the show notes for a link to it, or if you just want to go and Google for pivotal cloud native roadshow, you'll find it pretty easily. We also have spring days Atlanta coming up July 18th and 19th. It's chock full of sessions for developers who want to learn more about the spring framework, cloud native style development, and of course, to be fully buzzword compliant, microservices. If you go to springdays.io, you can get more info. And that's the last spring days we're having so far that we have scheduled this year. So get that one in if you're over in grits and pork land. It'll be good stuff. Finally, why it's way in the future. We also have a SpringOne platform coming up December 4th and 5th. Now registration just opened recently for this. I think you might've missed the early bird thing for it, but that's fine. There's also still the CFP open. It closes on September 1st. Now, what goes on at this conference? Well, it's full of what I would call the suit track and the technology track. In the suit track, you've got case studies and managers and developers as well, but organizations talking about how they've transformed their company and what they're doing with cloud native and their organizations, how they're getting good results by switching over to doing things in more of a pivotal way. But then there's also plenty of events for those of you out there who enjoy more nerding out and doing technical things to check out. I'm one of the chairs of the DevOps pipeline and monitoring track that we informally call it. We've already got lots of excellent talks queued up from the likes of Home Depot, Express Scripts, Allstate Northern Trust, and of course, plenty of pivotal people. I'll be speaking in that track if that's anything. So come on there. Again, it's in San Francisco, December 4th and 5th. And if you just go to springplatform.io, you can check it all out. And with that, enjoy the episode this week. Well, so we don't have a guest in case you can't tell. So other than drawing out our new segment,

as we have successfully done, you know, Richard had the good idea to go over stuff that we're working on, which seems good. Always easy to talk about yourself. So let's start with you, Richard. What's the stuff you got going on? Do you got some sort of desk situation? Are you one of these people who keeps a bunch of logos, icons on your desktop? Or do you have an intricately organized file system? Or third option, do you have the iCloud slash Google Docs view of the world, which is a flat file system, but isolated somewhere? Good question. I mean, my desktop itself has no icons. I don't like a messy desktop. But then, you know, some light organization in a pivotal folder of docs. And I use G Drive for stuff as well. But, you know, try to keep things fairly organized as I'm writing and reading. And, you know, day job is supposed to be leading our product marketing team. But I do like to research and write a lot of stuff. So yeah, the things I'm working on at the moment are, I just posted this morning, a take on cloud native integration. What does that mean? I did some research and did a presentation on that at a conference a couple weeks ago in London, and it was well received. And so just kind of exploring that more, what does it mean to modernize your integration practice? You know, how is that sometimes accidentally holding you back? Because now you've got all these 12 factor apps, and you're doing CICD, and everything absolutely grinds to a halt, because your ESB team is backlogged for three months. So how do you actually think about the integration piece, just like data, just like security? Because if you don't fix all of them, it really doesn't matter how good you are at apps, because it's all going to block up. I'll have to check that out, because, you know, I always like to overly comically simplify things. But I feel like there's a good 20-30% of enterprise application development that's basically like, I have a form that someone filled out, and three or more people have to look at it. And then I need to, as they used to say, mash up some other data in it. And at the end, the form's

either going to have a green checkmark or a big red X on it. And like, that general, whether it's insurance, banking, probably even farm, all sorts of things, especially government, it seems to repeat itself over and over again. And then to your point, like, a huge necessary component of that is integration with not only data sources, but other processes and things like that. Exactly right. It's the workflow piece. And then all this SaaS stuff is great. But what we know, how the heck are you connecting to Workday? And how are you connecting to Salesforce and using this weird cloud database? And like, that's necessary now. So it was anyway, it's a good, hopefully a good blog post. It's out there on my personal blog, serother.wordpress.com, and probably talk about it more and more. So that was something cranked out, something for the Pivotal blog, probably this week is my take on event-driven architecture and how, what does that really look like technology-wise? What does that look like kind of capability-wise? And actually, speaking of Gardner, I went through a number of their folks and had them put some eyeballs on this to make sure I wasn't talking crazy. And so it's had some good peer review stuff industry-wise internally. And it'll just look at what are the core components in an event-driven architecture? Yes, how does Cloud Foundry help you with that? But frankly, I just haven't seen much in the industry around, not just scenario-based. Like, if you read some of the Gardner reports, they're really, they're fine. But they focus on like, hey, here's one use case of how you can process events through Kafka and dump it out to something. Like, that's great. But if I'm a CIO and I say, okay, what do I need to do to actually be event-driven? Like, there are not that, it's not that information out there. So I thought I'd write that up as kind of a reference architecture from my old enterprise architecture days, like doing reference architectures. So put one of those out. We'll see if it lands with a thud or people go, yay. How do you judge that nowadays? The thud or the yay? I've given up because there's times where I will handcraft a blog post and think that this will go viral. And then you get like one comment on Twitter and a few hits. Other times I'll do like a throwaway thing and that goes nuts. So I've given up predicting what people care about. I just try to include more

memes and cat GIFs because that seems to drive traffic. Yeah. Yeah. You know, I think I'm more or less in the same thing. I think over the course of a year, there's sort of like sheer metrics you can look at. Like, you know, I think that cloud native whatever book I have, like that gets downloaded like 15 or 20 times a week at this point or something. And so, you know, over time that's meaningful. But, you know, it's mostly confounding. Like the things that people end up liking, I don't know what the deal is. No, I mean, the most popular blog post I still have on my own personal blog is something I wrote years ago on where should you host your Node.js app. Like somehow that's like the number two hit every day on the damn blog. It drives me crazy. So, but again, you just don't know what has good lasting power. And that's fine. Many of the things we write today will actually be more popular a month from now or longer. And either way, you just have to put yourself out there. Maybe you're wrong. And sometimes I appreciate that too, because someone will yell at you and you have to read the comments. Yeah. I think the number one thing I write stuff on nowadays is just not word of mouth on the classical thing, but people actually saying something, whether it's online or at conferences or whatever. Cause that, uh, like, uh, I don't know for the, for the kind of people that the kind of audience and market or whatever you want to call it, that we work with, there's not that many of them. And so like, like the, uh, the actual numbers of people looking at something are very small, but, but the, you know, if, if, if the right one to three, I don't know, managers or developers or enterprise architects at some gigantic bank, look at it and like it, then it has a huge effect. And so it's, it's the, uh, I don't know what you call that kind of model. Well, it's kind of the aggregator I've noticed. I wrote a post on how enterprise architecture adjusts to the cloud and that thing blew up because it got picked up by a few different like EA sort of aggregates who shared it out. Or, you know, if something hits the DevOps weekly newsletter, it blows up. So it's just interesting how certain things that collect these musings we throw out into the ether that, that, that, that seems to help more than just who gets

the most retweets on the first post. Yeah. So what else do you got there? Yeah. So I'm going to Denver next week to deliver another.net cloud native bootcamp. I put together this event with Zach Brown here at pivotal where we would walk through kind of what does modern.net look like? What the steel, how do you use steel toe? And then we spend the second half of the day actually building apps. So we did this in Seattle a couple of weeks ago, and it was awesome to have 20 something people in the room who hadn't touched any of these tools. Some of these were admins. This was not just devs, but kind of cranking out little bits of code, pushing in the cloud foundry, doing some of this crazy, you know, backing config store and service discovery stuff. So we got a request to take it on the road. So we'll do Denver next, probably one more city, and then I'll hand it off to people who like traveling all the time. But it's really neat to see this conversation move forward. Cloud foundry is awesome for .net. There's some cool libraries out there and just kind of unlocking some people who thought that their career was stalled because they were .net shop. And instead we're, we're opening that back up again, which is awesome. So, and then coming in September, Azure stack. Yeah, that's right. Then we'll retool it all for private. That, you know, you know, that, that would be, that'd be, you could finally, uh, fulfill one of the reoccurring fantasies I have with vendors is like, why don't you just bring one of those? Right? Like maybe, maybe you can actually, I think I have a picture of, of my, uh, my friend on my other podcast, uh, Matt Ray actually rolling around a, a Dell open stack half rack, but like, if I could use a dolly to bring in PV, you know, Azure stack with PCF on top. And then we all just get into that. That's the dream. You got to get one of those just, uh, come September, call them up and have them ship you on. Right. Dolly as a service. That'd be awesome. So I've got the couple other things, just a quick mention. So for those of you who like spring cloud and development, I'm finishing up my second plural site course on spring cloud. That's that video on demand. training Pivotal is a partner now. It's insanely popular. I've been doing it for five or six years now. And so I'm just about done a spring cloud course on how do you actually coordinate your microservices? That should be six hours of fun for your family. So that's almost finished.

And then I'm doing executive briefing in a few weeks on IOT and cloud foundry and platform. So now I'm in the research mode for a lot of that and thinking about what are the sort of things you really need for modern IOT. So that jumps me back into a learning mode to find out kind of what's the latest thinking on that from both a consumer perspective, but then also supply chain back office stuff. It seems like there's a lot of cool stuff happening there, but what tech do you actually need to be successful there? So that's what I'm with. Well, I'll have to check those out. That's all stuff that people ask me about. And supposedly, given the volume of noise I make after they ask me that I know something about. Although I usually say, I don't know anything about this. And then I just proceed to talk, which is scary. It's scary. Yeah. Yeah. Well, so I, I only have two things I wanted to point people to like, so I most recently, I guess it was just last week, I was talking with one of our more recent customers with a gaggle of enterprise architects. I think, I think that's what you call a group of enterprise architects, right? A gaggle. You might just call it a cab. I was talking with a cab of enterprise architects and, uh, you know, they, they had a lot of questions about what a cloud native enterprise architecture is. And so I, I sort of like, and I don't know, I mean, this is, this is sort of like drives a lot of the way I do things. I was like, Oh, I should, uh, go collect together that crazy board on my virtual wall of all the stuff I've gotten and, uh, and start writing it up. And I think, and then, and then also, uh, if, if, if you were paying attention to my sub tweeting last week, I wrote together a talk abstract that I have now submitted to people, uh, conferences. So I need to start gathering together all of the, uh, sort of theoretic and actual real advice case studies of what enterprise architecture in a cloud native world is. And as you were saying, I mean, you sent over a, uh, a nice list of, of things, which I think that's, that's a good, uh, initial roundup of everything, but there's, I find on this topic, there's no end of questions and no end of

lack of answers, so to speak. Like, uh, it's, it's, as I was telling this, this, uh, this cab of architects, so to speak, it's only been like in the past two or three years where you actually have more than a handful of large organizations doing anything cloud native. So what do they call that in economics, natural research or something like there's only, uh, there's not really enough natural resource to exactly know everything about it, but there definitely are a lot of, uh, emerging things. And I think, I think the intellectual sort of, if you will, quandary that this is, is, uh, if in a cloud native or, or DevOps, if you prefer, uh, method, if we push down 99% of their responsibility down to the product teams and we connect them directly to the business, uh, what enterprise architects do? Cause, cause classically, I don't know. I mean, you know, you having done this, you probably have good input on it, but classically an enterprise architect, a lot of what they do is, I mean, there's two things. Stereotypically they do, which is the same thing. They do governance to make sure you're not doing dumb stuff. And also that your long-term view is good. You're trying to prevent like technical debt, but then they also like fill in all these gaps between all the various roles that you have. And, you know, they try to have the, uh, the lattice work that makes it so the business people talk to the product managers and the QA people and the ops people, and they try to make it all sort of fit together. Hence architecture. Right. I'm supposed to kind of be the connector, the matchmaker in a lot of cases. It's just no, no two shops that I've ever either worked in or talked to do EA the same way. So that's why it's so difficult. Some are like business architects. Like we will change how we do supply chain and how we send products to market. Others are like just they're technical architects and they do reference architectures and they incubate new stuff and they work across teams and there's just no common definition. It seems like at this point. Yeah. And, and so like in, in talking about this with people over the past few years, either directly or indirectly, I think my evolving theory of, of what cloud native enterprise

architecture is, is, um, whenever you scale up to like, you know, hundreds of teams and thousands of products, like to a large organization, uh, there's a lot of white space between these teams and you need to figure, I don't know what it is exactly, but someone has to take care of it. Otherwise it doesn't all like really fit together. And I think a huge part of that, and I don't know if this is like what an EA should do or someone else, but, um, like, I think I'll finally release this, this, uh, conversation I had in the next episode with, with some of the all state people about setting this up is there's this initial year or two run where it's like, you're just setting up all this stuff and doing the change management of getting everyone to use a common build pipeline and getting everyone to, uh, run on a common platform and spreading out the teams and all of that. And I feel like an EA or the cab of EA should have a big hand in all of that because they're, they're one of the, one of the few people who have not only the big picture perspective, I mean, in theory, I've met plenty of EAs who just sort of like, they can, they can direct you to the bathroom in their Ivy tower, but they don't really know what happens out in the city. So to speak, that was a weird metaphor. Uh, but the good EA is kind of know everything that's going on. And also equally importantly, they know all the internal politics of everything that goes on in the organization. If, if they've rose, Rosen, if they've risen to that level. And so it seems like they could be, uh, very helpful in, in, uh, changing things over. That's the thing. I don't know if they do it right. It's a good change agent. If you do it wrong, you're going to get fired. Yeah. Yeah. I always, there's always the fired option. Anyways, speaking of snarky results for stuff. So another thing I started doing this weekend, you know, I, I've mentioned as at my mom's house, uh, experiencing the pleasure of a juicer, uh, or, or I guess, I don't know if I was experiencing it as much as, uh, appreciating its results. So I've been thinking for a while that I should collect together all my, uh, you know, cause I'm, I'm, I'm arrogant like that. I should collect together all my register columns

and a few other things that I've written here and there into, uh, one of those little books, you know, have you heard of books? And, uh, and, and so I started doing that. And I think, I think I've got, I've got all the register pieces that make sense. There's a few of them that are just like too topical or, uh, to put in there, but, uh, yeah. So I started collecting that together and what, what I call like a paste pot book, right? Like I'm sure you've read these books and you read halfway through it and you're like, hold on. I think this is just a bunch of blog posts someone strung together. And so that's what I got going on there. And I'll see, I think I need maybe, I think I'm up to 20,000 words and I think you want to be at like 30 or 40,000 words before you have a book. Otherwise I think they call that a novella. Is that right? That's right. And, uh, I mean, I desperately want the back cover of your book to include your best comments on your articles. I mean, that would just be like, not the, like Coté is a brilliant writer who really should. No, it'd be like, here's a pot smoking hipster who doesn't understand the bobs. Like that's got it. We'll see. Yeah. I've, I've been talking with some people who are like legit publisher or who have connections to legit publishers to maybe do something. But as always, if I get angsty and if anything, I'll just publish a PDF somewhere. And in that case, I will definitely take, uh, the, the analysis that I'm a, uh, a Valium taking DevOps gonzo person is the top quote. That'll be a, yeah, that's good stuff. But yeah, I'll, I'll put a link to both of those. Like I do with these books. I, I have them, uh, uh, publicly available to comment on in Google docs. And I think, uh, it's always good to get input on that stuff, especially the first one. So, uh, before we wrap up, I mean, you've talked about this a little bit, but I'm, I'm curious, what's your, uh, what's the tool chain you use for all this writing? You, uh, you use a notepad over there. What do you, what do you do for all that? How do you, there's, there's, there's really, well, I know the publishing tool chain, cause we both published to the same CMS, but like there's the collecting together your notes and research, right? I don't know. Maybe, maybe you're like that guy, Toby in the West wing where it's just all in your head and you don't actually have to ever write anything down. And then there's the

actually writing it. So what do you use for those things? Yeah. So it usually starts off in one note, but I, I've used that for years. So if I, you know, I like every note that that's where I just gather random thoughts and start to make bulleted lists and that eventually turns into something. And that's when I actually started, uh, Josh McKenty turned me on to using the Hemingway app. So I stopped using so many big words and make it uncomfortable to read. So then I actually write whatever I'm doing, a blog post, a white paper, whatever, and Hemingway. So I get kind of a good feeling. I didn't accidentally get too verbose. And then I ended up probably copying it into a Google doc so I could share it around for peer review. And then obviously we publish the same way. So that's usually how everything flows for me, whether that's a personal blog post or whatever. That's good. See, that's confirming because that's, I don't use the Hemingway app. I go with that every now and then. I'm like, I was joking about this earlier. Like, Oh man, that seems like a lot of work with the way that I write. It's actually pretty nice, but you know, it works, but you know, if anything, it just helps me make, you know, when you get that red mark, when there's your sentence is too long, it really gets in your head. So it really helps you write more succinctly. Yeah. Yeah. No, I know. I know. I know. It's not a lot of work to put it into it. It's a lot of work to, if you believe it's advice, go back and fix it. That's, that's the, uh, but sure. I'm sure it results in clear writing, right? Like, you know, one pair of baby shoes never used. That's, that's, that's what you want when you're in your clean, well lit place. But anyhow, uh, yeah, that's confirming because that's what I always come back to. I always, I always cycle through using all sorts of stuff, but I don't use one note cause I'm not a Microsoft person. I use Evernote, but every now and then I think I'll have like plain text files with Markdown. And I tried to use that Bayer application once, which is lovely. If I had never used Evernote, I would probably use that Bayer application to like collect everything together. And then, uh, I don't know, you got to put it in Google docs. That's like eventually everything. Yeah. Like, like that's, that's, that's like the only way to collaborate nowadays. I mean, the, the best way, I don't know, maybe, maybe, uh, other things are good too, but it's just like, that's, uh, everyone has that. And I think they more or less understand

it. That's right. We're making Google's brain smarter because I'm sure it's reading all the docs. Exactly. All right. Well, uh, with that, this is as always been Pivotal Conversations. If you want to get the most recent episode and find the link to subscribe in the podcast, you can add little comments and see what weird little music spam bots over in, in, uh, in, uh, SoundCloud really like to hear about infrastructure software. And, uh, you know, my daughter uses that as well, uh, I think in the future, but you go to soundcloud.com slash Pivotal Conversations, all one word. You can check out the, uh, the episodes there. And as always, you can go to blog, not blogs. You can go to pivotal.io slash podcast to get the full show notes that we've mentioned, and we'll see everyone next time.

Tanzu Talk · 🍕 Get more pizza.