Pivotal Cloud Foundry 1.8 with Jared Ruckle (Ep. 35)

Episode 22 · 53:02 · SRT · archive.org · all Tanzu Talk

Well, you know, I just recorded another podcast and I asked this question and usually I assume that everyone who listens to any one of my podcasts listens to all of them. My son's been doing word problems in math, so I think I just formulated one there for logic. But I'm going to go ahead and ask this question again because I'm very curious. So for each of you, do you use pencils or pens? And what is your rationale? Why don't you go first, Richard? Yeah, I believe I'm a pen person because I'm not seven and I don't have pencils sitting around. So I do typically use pens because if I can't get it the first time, that's not worth doing. No, no erasers for me. That's perfect every time or I destroy it. That's right. How about how about yourself, Jared? I think in the corporate world, there's actually a lot more martens for error than in the school world. So I'm a pen guy. Also, if you make a mistake, you just cross it out and you kind of move on. There isn't quite as much focus on penmanship in the corporate world from what I found. That's true. I'm drawing a race marker more than any of them now. Yeah. I think if I was in a meeting where whiteboarding happened and someone started writing in cursive, I would just scoot back from the desk and be like, all right, we're done here and just leave. That's it. That would be such a strange occurrence. I don't even see cursive in my normal life, let alone in a corporate world. We spent a lot of time learning that when I was in school. It's a weird vestigial thing. It is. So we've done a little bit of introducing. Why don't you introduce yourself briefly, guest? I'm Jerry Ruckel. I'm on the marketing team here at Pivotal in beautiful Seattle, Washington. I work on Richard's team helping tell the Pivotal Cloud Foundry story to the world. So I wanted to have – well, I should say we wanted to have you on because you wrote a post about the most recent release of our Pivotal Cloud Foundry, version 1.8.

And when we get back – when we get to the second half of the show, we'll go over that. But thanks for being on here. Yeah, looking forward to it. My pleasure. Yeah, I like how you said that it was your idea. Like, I'm keeping him from – I'm trying to really keep Jared in a headlock and not unleashing him on the wild, but you fought me on that. Well, as everyone would presume from a company that talks about Agile and DevOps and culture, we try to maximize the output of employees. So I know I owe you a few favors for using up some of Jared's time here. I'll have to – I'll ship you a case of pencils. How about that? I like that. Better stay. Well, you know, I was joking before this that it was something of a slow news week, at least from the infrastructure perspective. We already talked about how Oracle is going to take out AWS last week. So we covered that. There was some more on that. You know, you can go read up about that. Very thrilling, exciting stuff. But there was – I think for me, the two most interesting things broadly – I think that, you know, there was – there's a new version of Windows Server coming out, which is always interesting from an infrastructure thing. We can – you can – you can school me being up in that neck of the woods about what's in there. But I think the most interesting practical thing is that apparently 500 million accounts were stolen from Yahoo. This is like in 2014 or something, right? Right. Right. Yeah, I need to – you know, I don't really use my Yahoo Mail anymore, but I do use Flickr. So I probably should go update my password and update it. Yeah, definitely scary stuff. I mean, obviously, there's plenty of jokes around. Like, it was great for their ads, you know, because they had a bunch of impressions by everyone logging in to change their password. And, you know, it's kind of a subtle way of also saying you had 500 million accounts, so well played. But it's, you know, not good stuff. Use a password manager if you don't already. Yeah. You know, that's definitely probably everyone's biggest takeaway is change passwords often. Yeah. Yeah, as we are fond – well, every now and then, you and I quipping on this podcast. Like, it's always be careful. You've got to be careful about throwing your rocks and glass houses because you, too, may one day be working at a company in the waning years of its revenue growth.

You don't want to make fun of them too much. No, and these breaches, sometimes they're internal, sometimes they're whatever. So, again, it's never fun. It's something that you may face yourself in most companies. Now, you two seem like you probably know security much better than I do. But now, if I turn on the old 2FA stuff, I'm pretty much cool, right? Like, that's – whenever something gets breached, I'll be safe because they can't do their second factor. I mean, I should still change my password. But it's sort of like we'll more or less guarantee my account can't be hacked. I mean, you're better off. There's always the social engineering sort of stuff where, hey, call up and say, my God, I just dropped my phone in the toilet. I can't get the two-factor off. Can you just let me buy this one time? And, hey, I know your birthday because I trolled you on Facebook. So, there's still ways that that's not any sort of foolproof thing. And if someone does steal your phone, cut off your thumb or something, then they can also get in there. There's another way. The old toilet defense. Always a winner. There's another thing that I saw what people are also doing is that for the answers to the secret question, what's your mother's maiden name, what's your pet's name, is that you just have 15 strings of random text. So, it's essentially another protected password instead of what those values would actually be. So, that's some extra skunk works advice there. Yeah. You know, and maybe it does this automatically, but I was hoping – because I noticed that too recently, and that's clever. I was hoping the old LastPass and 1Password would work with that. And if they don't, I'll have to submit a feature request because that would be nice. It's just additional passwords. Yeah. I was – it does – well, what? Anyways. Yeah. Breaches are very annoying. I was talking with my wife about it because she has the thing where she uses her Yahoo account for her junk mail that, you know, I have other – you know, I have my infamous Johnny Legion account that I use for getting free PDFs. But she uses Yahoo for that, and she actually has changed her password in the last year. So, she's following the advice of rotating.

So, good for her. Yeah. Good old passwords. I actually, you know, while we're on this topic, I just enabled the unlock your laptop with your Apple Watch thing, which it didn't work for the first three days, and miraculously it started working, which is kind of mysterious. But, yeah, it's actually kind of nice. I just – for this podcast, I just sat down now, and my computer was unlocking, and I was freaked out a little bit, and then I remembered it's because I have that turned on. But that'll be fun. So, as I mentioned, there's also the general availability of Windows Server 2016. Now, everyone who listens to this podcast knows that I don't know that much about Microsoft stuff, but you guys do. What's going on in Windows Server 2016? Yeah. I mean, it's the usual – you know, some updates on – it's always easier. I mean, you never see someone say, this is the second best version of Windows we've ever released. So, it's always, this is the best version of Windows. You know, you hope so. There's some simplicity, things around storage and networking. There's some support for some of the container things natively within Windows, which can be pretty exciting. So, there's a – it's the usual upgrade thing. I believe that Windows continues to bleed market share to Linux, as more people see that as an automation-centric OS to use. And, you know, people were getting a little excited today. Docker also announced some general availability for Docker on Windows 2016. And, again, it's really easy to get excited about that. So, I did some quick, you know, serious research using the Googles to find out, you know, what is the adoption rate of Windows versions? Because we can get excited about these new versions. But, you know, according to numbers from July of this year, Windows Server 2003, which is unsupported, still has an 18% market share. You have Windows Server 2008 is running 45% of workloads, and Windows Server 2012, you know, a four-year-old OS, is still only 24%. So, you know, we can be excited, but you're going to be 10 years out before Windows 2016 is seen as a mainstay in your enterprise. Does that kind of stick into a gloopy sort of bell curve there, if I heard the numbers right?

That's kind of nice. Almost, yes. I mean, when you see Windows 2012 is used as much as Windows Server 2003, that makes you think the apocalypse is nigh. Well, first of all, that's sort of like, that's some activity after my own sort of tech news reading heart. Like, I don't, maybe I complain about this too much, but I hate those kind of like slapdash kind of tech journalism stories that have no context, right? So that's like an obvious question to ask is like, what is the, what is the, the, the sort of penetration and adoption patterns for this thing that came out? So good job. Gold star on that. You know, I want to be like you when I grow up. So I got to go double click in. Oh, go on, go on. Indeed. So, so yeah, that's exciting. I think, I think, you know, I think when I was a young programmer and I've commented on this several times when we cover Microsoft software news, like I was, I was, you know, like a PHP programmer and a Java programmer. And it's not until I became sort of like a, a real person in, in, with cognizance in the outside world that you realize like, oh, Microsoft is kind of a big deal in the, in, in the infrastructure software world. So that's great. It's good. They have a, a release out there. It's good. I think in, in light of our hot takes last week, which I enjoyed, you know, I would say if my hot take on this is that if you as a company are waiting for innovations at the OS level, then you are thinking at the wrong level. You need to be thinking at the, I mean, that's where PCF and platforms matter because your deployment of OS level updates is, is depressingly slow. So if you're looking for lessons to help you innovate, then you're probably not doing it right. So did you, did you read up on, uh, I just saw this before, but what's, what's the, uh, the Docker partnership thing? Did you, uh, yeah, I mean, it uses the native windows 2016 containers to let you now build and run Docker containers on windows, which is novel. Now I get PCF containerized windows for, for years now using other mechanisms.

And really the goal is do I, can I isolate workloads and do some resource management that that's the goal here? So, you know, we've been doing that for a while. It's really cool that it runs natively in 2016. We've actually already tested and will support PCF on windows server 2016. So you're already set there. We're not using the native constructs yet. We're using what we've built around that, but all good stuff. So as, as that continues to mature as technology, you'll probably see us figure out some ways to support the native bits. Yeah. Yeah. I, you know, I, I, I like this, uh, this new Microsoft that, that very, uh, I wouldn't even call them a fast follower so much anymore. They're, they're pretty much on schedule for everything. And then they seem to, uh, they seem to fully to use an older and even an old term again, they seem to fully embrace, not so much extend like, uh, other technologies out there. And it just works like, I mean, for, for those who don't know, obviously they, they're an investor in pivotal. So, so they like the, like the way that we do things and they work with us a lot and they, they have their own way of going about things. I was just at a DevOps days, uh, Dallas or DFW last week and Jeffrey Snover, who I think is like head of engineering for STG or whatever they call it, or some highfalutin thing. He's one of the only fellows there, but the PowerShell person, like they have PowerShell, like they're, uh, I don't know. Like I, I think, I think if you were to, uh, as it were, be an alien who came to the planet and observed all of Microsoft's infrastructure stuff, it would seem pretty like, like everything else. When I mean that as, as, as like a, uh, a good compliment, they've really embraced the heterogeneous data center thing that you didn't really, for example, see people like Sundu way back when, who's kind of, uh, just, uh, didn't acknowledge the existence of that whole world. Very effectively. No, they've done a very good job. Now, of course there can't be any Microsoft news that doesn't get commented on, on Reddit or hacker news with people still referring to, you know, 1993 era Microsoft to putting the dollar sign in the Microsoft name and hating them. But, you know, as much as a company can transform, it should be, uh, they should be complimented for, for going a place.

They could have gone sideways six years ago and instead they reinvented themselves around open source and agility and other stuff and kudos to them. So, yeah, hopefully there's some sort of, uh, Cade Metz or Steven Levy type of person writing a book about it. That, that would be nice. The, uh, the new era of Microsoft. I like that. So, uh, let's see. Also, it looks like also in Microsoft land that, that Azure has spread out to, uh, to many more regions. How many regions do you think they have going now? Yeah, they are up to 34, which, you know, Scott Guthrie, the, the big cheese of Azure is often prone to say that they have more than Amazon and Google combined. And, you know, the interesting thing, everybody can keep expanding. That's fine. What was interesting here is A, it's Germany and Germany is, is well renowned for having some extremely stringent data protection rights. So this one was interesting because they actually have this, these locations running under a trustee of sorts to make sure that they're, they're following all the right data protection concerns. So if you're wanting to keep data in Germany, which is often a huge criteria, then, you know, it seems like Microsoft went the extra mile to actually make sure that you have a lot of confidence in that. Yeah. Yeah. I mean, I, I think, I think that's, that's kind of, that's kind of a representative data point of, uh, when we were talking about the, uh, the public IaaS magic quadrant out there. And, uh, one of the ways of summarizing it is like, uh, well, there's two ways is like, good luck if you're not Amazon or Azure. Sure. And the, the, and the reason for that is the, the scale and barriers to entry are so massive that it's, it takes a long time and a lot of money to build up to that. And this is a good example of if you want to, if you want to be the world server, you've got to go out into the world kind of ironically for how you, how, for the way that people consume cloud, where you're not supposed to care about it and all of that stuff. But all of that gets shifted to the cloud provider and they have to worry about, you know, every single country, every region and, or I should say worrying about them and building it out. So it's, uh, that scale is, is pretty amazing.

Yeah. And these clouds operate, I mean, I'll, I'll, I'm going to pepper Jared with this question later, but I mean, what's interesting is, you know, German now, Germany now behaves a little differently. China is usually a different type of cloud environment for vendors, obviously have others. So sometimes multi-cloud can actually all happen within the same provider because in essence, you're dealing with multiple data sets, sometimes even different policies. So multi-cloud doesn't have to span providers or even sites. No, that's, that's, that's a good point. And it's also to, to use the, uh, funny homogeneous heterogeneous stuff from, from the early days. It's, it's almost as if the, uh, countries are defining the heterogeneity now. I mean, not completely, but it's, to your point, it's interesting to think about how part of what as a large organization you have to worry about with incompatibilities and integration is due to, uh, geographic things. Not necessarily if you're using, uh, you know, Unix versus a mainframe versus windows versus Linux, but I don't know. That's just a fun thought exercise. Speaking of fun thought exercises, there were, there were rumors that the last of, uh, uh, at the end of last week. And I have to say the, uh, especially if it's from Bloomberg, like the, the rumor mill for tech acquisitions has, has been pretty, uh, accurate over the past couple of years. It's, it's always panned out pretty well, but it looks like maybe everyone's favorite Twitter is, is trying to sell itself and be acquired by someone. And the, uh, the interesting thought experiment is like, who would he, who, what would that be like? And it, and it is, I mean, before I get y'all's thoughts on it, it is a little, uh, I wouldn't say unnerving, but, but I, I usually don't care about what happens in the consumer space. But because Twitter is such like a, I guess, important mission critical part of my career, I actually care a lot more about this than, uh, normal things. I would like, it's, uh, I, I, I gotta, I gotta know what to go. I, I, I at least need to know when, and if it hits an iceberg so that I can start establishing myself somewhere else. Because I do this analysis every now and then, and like, aside from when, uh, our own corporation helps promote my own sort of content on my own, the only place that I'm really successful in driving a lot of traffic is in Twitter and the podcast that I do.

So, you know, I got to keep this stuff up or, or, uh, or I'm going to have to downsize my life or something. So hopefully someone responsible buys them. Yeah. You'd be dusting off your Google plus profile, which we know that drives a lot. Exactly. I had to go look this up last week, but I'll, I'll have to go back to friend feed. Although Facebook, I, Facebook bought them and shut them down. So I don't know, but I don't know what's, uh, what's y'all's fantasy buyers for something like Twitter? I mean, you saw the list, I think of who's sniffing around here and everybody can make up the list of whether it's, it's obviously people like Google or Salesforce was in the mix Verizon. So, I mean, there seems like there's some bad places for them to go that it might be morphed into something that fits that company's worldview. I don't know. I feel like I'm kind of rooting for Google on this one. Because I think that they could do some cool stuff, integrating search into the firehose of Twitter and probably wouldn't muck with the spirit of what that thing is versus, you know, Hey, I love Salesforce, but you could see this becoming a very, something that integrates more into their worldview of customers and relationships. And like, I want it to just be a platform where people can scream about the weather and yell at each other, not something that business value. So, you know, I'd probably rather go to a company that just sees this as interesting data versus forcing it into an experience. You want old man yelling at cloud, not old man providing cloud business. That would be much better. Sorry, I interrupted you, Jared. What were you saying? Yeah, I think that, yeah, I agree with Richard that I'd probably be rooting for Google here as well because it's such a transformative platform. And I think that, you know, Coté, you mentioned some of the things of what it's done for you personally. I think a lot of the hardcore Twitter users feel the same way. And something like Google would probably make it more likely to be kind of left alone as this, you know, expression platform where it's so transformative and how news is transmitted and how real time it is and the citizen journalist kind of aspect of it. And it going to a business software firm, you know, seems a little weird. Yeah, yeah. Man, who knows? I mean, I think just to play armchair, like, iBanker or strategy person, like, I think Google is a good, like, second choice because of our assumptions that they would just, like, not have to be profitable, I guess.

Or that they could be, like, you know, barely profitable just because, you know, Google seems to be happy because Google has lots of money and they can run things like that. On the other hand, they sort of, like, seem to have run Nest into the ground and other things like that. But it seems like long-term-wise, something like Facebook would be most interesting because I think Facebook has a pretty good history, like, with Instagram, for example, of not only not ruining but probably improving and making better a business of things. And the synergies there would be amazing, right? It's just like Facebook already is really good at selling ads and hopefully they would be good at Twitter. And the only risk is that probably the reason people like us like Twitter is a large part of it is because we don't like Facebook or we don't want to mix those two streams together, right? And so that would be a danger. They would have to, like, Instagram it, so to speak. But I don't know. We'll see what happens. Despite me saying that I think, like, Bloomberg and others are good at the rumor stuff, like, this is one of those that, like, it almost seems maybe this is, like, the old guy in me, like, not getting used to the future. But, like, Twitter being bought by someone? Come on. How could that happen? But we'll see. It seems like such a standalone brand. I know. Social media is just exploding and Twitter continues to grow slowly. So, I mean, it seems like the drumbeat has been there for a while that they need to figure out either a business model or find out a big parent who can take them under their wing and let them just keep growing organically. Yeah. No, that makes sense. That makes sense. Well, at the opposite end of success with glowing organically, which is to say, in a good way, we have a new version of Pivotal Cloud Foundry out. I think it GA'd a couple weeks ago. Is that right? Yep. Yeah. So, it's formally, I think, titled Pivotal Cloud Foundry 1.8. I think we're pretty good with our naming. And so, you know, you wrote a post on it, Jared, a little while ago. But before we get to that, why don't you introduce yourself?

What's your background? How long have you been here? Like, you know, who are you? Yeah. I think probably the best place to start, I've been doing, you know, product marketing, product management for many, many years, longer than I think that I ever would have thought that I would have been doing that. Been at Pivotal for about two months, almost exactly. Yeah. Working on Richard's team with product marketing, telling our story out there. This software remains to still be very hungry and is eating the world five years on. So, a lot of cool things happening in how we can help our customers get their software to eat some market share, I guess. So, how would you define what product marketing does? Not just here, but just in general. And I ask this very selfishly, because every time I see my dad, who worked at IBM, so I don't know why he asked me these questions. But I think he's just, like, gaslighting me. But, like, it's a little weird to describe what product marketing does. So, I'm trying to get some cliff notes here. Like, what do you tell people you do? Yeah, it is a little bit tricky because you're not a developer, you don't code anything, and you're not a salesperson, so you don't sell anything. So, what is it exactly? Kickboxing, if I remember. And so, I go back to the pragmatic marketing school of thought, where product management is defined as voice of the user, and product marketing is voice of the buyer. And you spend a lot of your time trying to figure out what happens in that buying process, what type of personas are aligned in the sales process, you know, what artifacts resonate with the persona, what problems do they have, and how they're trying to solve it. And I think that it's really gotten interesting, that distinction in recent years in this subscription economy that we're in, where you don't, you know, buy a piece of software, and then, you know, try and get it to work for three years, and then maybe renew something.

That buying and using criteria and behavior have a lot of kind of overlapping ebbs and flows as people are, you know, constantly kind of, you know, re-evaluating, you know, what tools they want to use and how they want to use them. Yeah, you know, I hadn't thought of that. I mean, that is, and it sounds cheesy to, like, say it out loud, but whatever. Like, there is, because of the subscription-based way of doing things, there's much more of a relationship people have with their software. In the sense of needing to, like, check up on it and think about what they're doing with it. I mean, it's different than, like, you bought this thing, and it's just sort of dropped off and see you later. I guess it's a relationship with the vendor. Like, you potentially have a more in-depth relationship because you're dealing with them more, which certainly is the case with us in a nifty way. Also, you're like the 50th person to name-check the pragmatic marketer stuff, so I should go look into that one day. That sounds really useful. Yeah, it was, where, you know, nobody goes to college for product management or product marketing. You sort of come up through the developer ranks or sales engineer ranks or, you know, through the marketing ranks with some type of, you know, technical savvy. So a lot of times you really need a playbook for how to make these things happen, and they've got all kinds of great, you know, slogans for how to deal with these things. You know, product managers have all the responsibility but none of the authority, and just all these kind of great, you know, bumper stickers that really reflect what life in tech is like where you don't have, you know, infinite time or infinite resources, and that's why I think the adjective pragmatic is such an apt one. Yeah. Well, let's see an example of product marketing in action. So what's up with 1.8? How would you sort of, like, summarize what's going on in this release? Yeah, it's interesting. I think for me, coming from more of the infrastructure-as-a-service or software-as-a-service world, this whole version number thing is quite the throwback. It's been an issue to kind of get your sea legs around the 1.8 and then all the other variations of version numbers.

But once you get past that, one of the themes that kind of stood out to me when I was writing the post, you know, came from SpringOne platform that we had in July in Las Vegas. And one of the keynotes really focused on, you know, time to value, really this notion of being able to reduce the delta in time from when a new feature is ready and when that feature can actually be shipped into production. And so in trying to look at the story and look at those themes and where our buyers may be and, in fact, where a lot of our users may be, it's really this idea of how can we get time to value for more of our applications. Because a lot of the platform scale that we, you know, expound, a lot of customers can realize you really get a lot of those efficiencies when you can run more types of applications on a platform like Pivotal Cloud Foundry. So I think that some of the features that we talked about in this release really supports that idea of having, you know, more and more types of workloads on the platform with some more advanced networking features with TCP routing that allows you to have some legacy workloads or maybe some bring your own container type of opportunities or even some Internet of things type scenarios coming onto the platform that use this, you know, different type of, you know, routing protocol versus HTTP. Right. In that same type of vein, there's also PCI. You know, we were just talking about, you know, Germany and data sovereignty and data compliance. When you're talking about really big companies, you know, coming to cloud computing, they have to contend with these regulations, you know, HIPAA and PCI, you know, a lot of things in the financial services sector. And I know we at Pivotal invest a lot in trying to help customers, you know, comply with those without compromising the pace of their software development. And so having things like security as just part of the platform that maybe works in the guts that you aren't really exposed to as a developer operator is a good thing. And some things we've done with with the IPsec add on this release really, I think, supports that notion of making security just part of a platform.

So you can go secure and go fast where those aren't necessarily, you know, diametrically opposed kind of kind of ideas. Right. Yeah. I mean, it seems like our our constant sort of theme, if you will, is whether you want to call it time to value or speed to market. But it's just it's just the good old fashioned IT thing of faster and and more specifically, like speeding up your release cycle and the way that you software and deploy it and everything. And and then, you know, a lot of what you went over is the the the other side of what we do is like, I mean, we have kind of a we don't always put it this way, but we have an audacious goal of. So here's the new platform for how you do all your software, right? Like basically a general purpose software runtime for all the software that you have. And as a consequence of that, right, like on the one hand, we want to have and we do have like all the if you want to do stuff like all the bleeding edge, like cloud native way, like there's definitely that in there. But as as we go more and more into our releases and have more and more customers, we just have to like add in and I don't mean this in a judgmental way by saying have to or whatever, but we have to add in a lot of standard things and just all all all of the things that that you need for a general purpose platform. And it's it's always interesting to see how we balance those two things on on each release. Yeah. And I think that the other the other big thing, too, just getting back to the features, is this log integration that we have with our PCF metrics module where, you know, it's great to go fast and release new software all the time. But there's always that day to consideration, you know, thinking about what it's like to actually run this stuff in production. And how can you you get that developer operator workflow to be as as tight and crisp as as possible? And with with metrics, giving you some information about how the application is behaving, you know, monitoring metrics information in this new release. Now you've got actual logs from the application mashed up in there.

So as you're looking at these graphs that are essentially the health of the application, you can drill down into these very slick looking charts and see what was happening in the logs during that actual period of time. So this is, I think, really a key thing for a lot of our customers, because developers and operators have the same set of facts as the product management team, you know, puts it for that offering. When something goes up in the night or there is an issue, a lot of times those teams are bringing two sets of facts to the table. And now with it being integrated into, you know, a module that's connected up to the platform, you've got everything right there. So people have the same set of facts. And when it comes time to troubleshooting an issue, they can just review the facts and get on with it. They don't need to debate who has the best data because it's the unified thing that they have. Right. Yeah, I mean, you talked, you touched on TCP routing a little bit. I mean, that's pretty new for us. I mean, PCF has typically been about web traffic, you know, HTTP traffic into an application. So with TCP routing, the whole idea is you can just do TCP directives through our cloud load balancer, things like that. What else jumped out with you on that? You wrote up some examples of that, but as a customer, a developer, whatever, why do they care about that? Yeah, I think that it's really just kind of what we talked about with being able to support more different types of workloads. I think there's a specific type of IoT protocol that works with TCP routing. And then this idea of legacy, where you look at what workloads enterprises have in their data centers today, the stuff that actually makes all the money, being able to think about how those systems, which were sometimes, you know, built, you know, 15 years ago, you know, a lot of those were using some different type of networking protocols. So being able to carve off some of those, you know, monolithic services using TCP routing, bringing those into the platform can help you start to get, you know, more of this microservices, you know, goodness versus the legacy construct of having to go months in between releases a lot of the time. You know, you also touched on, you know, going back to the world of point releases, you know, when Jared and I worked together for a cloud provider, and we'd ship every couple of weeks, there was no concept of a version of the cloud.

I mean, it was just whatever was out there. But we purposely at Pivotal don't make a big deal about releases, we do a blog post, we educate our field and our customers, but we don't do big PR stuff, we don't go brief every analyst, because I think we see it as there's just a drumbeat, we're still shipping pretty regularly. What even between the quote unquote point releases, you have new tiles coming out from partners, you have new plugins, you have new extensibilities in the open source. So, I mean, how is this different from you from maybe, you know, classic point release stuff to always shipping as a cloud provider to what this kind of hybrid thing is? Is that weird for you? Is that if you think that's just a good new reality? Yeah, I think it's a good new reality, because it's, it's great for the customer. It's great for the customer to have, you know, new, new chunks of features coming out constantly and sort of, you know, smaller bite size bits that, you know, reduce the risk of, you know, migration or reduce the risk of things that you're operating. I think there's been a lot that has been, you know, talked about, about, you know, what this new Agile thing means for developers and operators. But it's also a different, you know, world for a lot of other parts of the organization, you know, marketing teams, you know, sales teams, customer care teams to kind of adjust to this sort of, you know, agile new reality where you've got, you know, probably a lot of big companies with big releases where they'd have, you know, a few big bangs every, every, every times a year. There's a lot of pomp and circumstance around those releases with, you know, trade shows, you know, confetti parades, all kinds of, all kinds of things. But then, like we just talked about on the Windows Server 2016 thing, that's kind of a big bang release that in reality won't have its impact felt for a few years down the line. With platforms, all this cloud, cloud native stuff, we want to get the stuff in the hands of customers, you know, much faster so they can realize the value and frankly, so they can give us feedback on, on where to go. So it, it is kind of a brave new world, but is only, you know, going to get, you know, more like that versus less like that. Yeah, I mean, we have a pretty good collaboration between product and marketing. And so you were a lot of at least observing the sausage making of building PCF and getting it ready and kind of going through staging environments, running in PWS first so we could get some feedback.

Again, I mean, building distributed systems is hard. You also think there's a parallel our customers are interested in because they're also trying to build complex software. And maybe we can even do more about explaining how we build it because what we're doing to build PCF may mirror some aspects of what they need to do to ship their own software. Yeah, there's definitely a lot of that where a lot of companies like Pivotal and I think a lot of others, you know, will do a lot of these executive briefings, EBCs, and essentially, you know, lift up the, lift up the kimono about how, how we work, how we build software, how we're able to, to go fast, what the culture is like, what the tooling is like, what the processes are like. And we even with Pivotal, I think we take that a step further with, you know, Pivotal Labs where we'll actually do, you know, long-term engagements with customers that teach them, you know, about how to kind of do this, you know, on their own and then, you know, go forth and prosper with this, you know, new methodology down the line. But I think that's a big part of the pitch and it's interesting that sort of the more automated the world becomes, the more important this human-to-human interaction is about, you know, doing things. And I guess that kind of gets back to the empathy idea that you hear a lot about at conferences and trade shows and stuff as well. So, so I, I, I, I had to, uh, uh, despite the fact that we're now raising up to the culture, empathy, uh, person level to go back down into the technical details. Now, now you've covered a little bit, but I have to admit, I have no idea what IPsec is because, uh, I, I don't like security. It's annoying. I mean, I like the effects of it, but I don't like knowing about it. So, uh, so that's good for my future employment opportunities. Uh, no security companies apparently, uh, but like, uh, like what is IPsec? What, what, can you tell me what that is? Yeah, I think that, uh, maybe the, uh, security for, for Coté, uh, book would have a couple of different chapters. Um, there, there's kind of this, you know, one way of thinking about, uh, encryption at rest.

So when you've got your data in storage, what kind of protections do you have around that? And then once your data starts, uh, you know, was it wasn't around what's the encryption in transit, you know, type of story. So the IPsec add-on is essentially a module that you install, uh, deep in the, in the core of pivotal cloud foundry that will encrypt the traffic of, of the data that's going as it, as it whizzes around of cloud foundry. So I think it's release number 1.5.37. And I mentioned that only because, uh, the minutia of that version number, you know, 3.7 versus 3.5 or some other variation. Each one of those releases will have really key features that will help you in the PCI, uh, auditing type of vein. Every customer, you know, is different. Their applications are different. You know, every PCI auditor is different, but now having a, having a story and an easy mechanism that can help with, uh, encryption in transit for your data inside of cloud foundry. It's something that makes that PCI story, um, a lot easier for our customers that are dealing with the storage of, of credit card data and so on. And I, and I guess that's an important nuance is this not only stuff outside of, uh, pivotal cloud foundry, but also when it talks to itself, so to speak, it's encrypted. So that means that you're, you're, uh, as we were joking about earlier, like you're more protected than, than, than if you weren't doing that, essentially, there's not a, uh, an internal weakness. Right. That's right. And, and again, there's really no kind of, you know, uh, drawback to the developer operator productivity, which I think is a really key, key part of the security investments that pivotal continues to make. You mentioned the, the, the rotate credentials idea. That's, you know, one of the, the big pillars of the pivotal security story, um, you know, repair and repay being other kinds of elements that those are all things that you need to almost think about security differently with your, you know, cloud native applications where you, you can't really have it as trying to keep things out. But just remaster the entire security experience for your platform altogether.

And you really diagnose what causes these, you know, bad actors to get into your system and, um, you know, really think, uh, you're quite, quite, quite differently about how the system can protect itself from those threats. Yeah. Yeah. All right. Well, that makes sense. And then the other area, you know, I just, uh, before this recording, I just saw that, uh, over at red monk, what, one of my old bosses, as it were, Steven O'Grady has this post. I haven't read yet about, uh, where's the databases and DevOps, but, uh, you know, we, we don't, we don't talk about it a lot. Maybe we need to update our diagram, but there's, there's, uh, there's some database improvements for like preexisting data sources. We have in this release, right? Yeah, this, uh, all goes to the version number kind of fun as well, where, um, uh, at, at PCF, we have our own, uh, you know, flavors of open source services, things like, you know, MySQL, things like RabbitMQ and, and Redis. And there's some interesting, you know, point releases that we have for the pivotal flavors of those that include the latest point releases of the open source versions of those particular products. So there's a lot of nice, you know, enhancements, um, latest features from the open source version that we'll package up into, into those different tiles that our customers can take advantage of. And I think this is another great example of, you know, how open source can really work well for, for big companies. You mentioned Coté earlier about, you know, maybe some objections that sometimes people will have about open source, um, by having a vendor that's committed to open source, like Pivotal is, but then helping take some of the rough edges off and upgrade in a regular cadence in a way that's contextual for the enterprise, um, can really, you know, make some of those barriers, you know, melt away. Yeah, I, I think, I think ongoing over the next few years, the way, uh, the way we handle things like whether it's MySQL or Redis or things like that and roll up those upgrades and make it as frictionless as possible for, for our customers like that. Uh, well, hopefully we'll be successful at it, but our, our success of that will help prove out the, uh, the kind of cloud platform model, namely being like, um, you shouldn't have to worry about that stuff, right?

Like the only worrying you should have to do about your database usage more or less, and by you, I guess, I mean the product teams, right? Of course there should be someone worrying about the, uh, the finite attitudes, if that's a word of it here and there, but there shouldn't be as much needing to worry about upgrading it and doing all of this stuff. And so it'll be fun to see how, how collectively the sort of cloud native world wangles out updating, uh, those various middleware services. So it's, uh, I don't know, it's encouraging to see that that's, we're not, we're not just sort of running away from state, which, which is like the hardest part of, of, of a lot of, uh, programming that you do, or, or I should say it's becoming one of the hardest parts because, uh, we've solved so many other parts of application development and delivery. Well, I think that, that's a huge point is, is that's where data microservices keep picking up so much mindshare as people are now thinking about, gosh, it's great that I can ship my, my core web app all the time. But what happens when I have a schema change in my database, what happens when I'm, you know, adding things, data microservices and DevOps with database, how do I do that without breaking everything? How do I kind of protect people from downtime or completely hose you? So I'm glad that Pivotal with some of the Spring Cloud Dataflow stuff and trying to do some new patterns around data there and different thoughts of data microservices that we all have to get better at that. Because if the database slows you down, it really doesn't matter if the rest of it's fast to some extent, because you're still stuck on that. Yeah, it's, it's, uh, at least for me, it's sort of an interesting, uh, idle thought experiment to think that the, uh, that the, uh, the idea of no SQL came about like what, like five years ago. And, and I think, I think maybe it's just now that there's sort of, uh, extremely wide mainstream worrying and thinking about it. And to some extent, and I, and I, and what you're saying made me think of this is, is, uh, without changing your application model, it's really hard to just slot a no SQL type of thing in. I don't think anyone says no SQL anymore, but it's hard to think of using a non-relational database unless you also change over how you think about your application layer.

And so I think the, uh, the application stuff has caught up to that. And so now there's a lot more interest in all of that. Like I was, I was looking at some chart, I should go find this, some chart of like all the non-relational database usage out there. And many of them have been over the last few years, years rising in popularity quite a bit. Yep. Yeah, no, no, no doubt. But we should cover that in some other episode. We'll get some, uh, some data person. I think the only coverage I've ever had in pivotal conversation podcasts about data is it's really hard. So, uh, we could delve into that. Yeah, that's probably, it's probably all we need. It's peak data right there. That's right. I wanted to ask one more question about this release, but we go into one other topic, but, um, on-demand service brokers. So we built some capability that is, is new for, for cloud foundry. Could you talk a little bit about what an on-demand service broker is or what, what was the status quo that this changes? Yeah, we've, uh, talked a lot about how, uh, pivotal cloud foundry is, uh, an opinionated, you know, platform. And I think one of the really, you know, uh, linchpin, uh, opinions that we have is about, you know, service brokers and, and how those are, uh, how those enable, you know, the wide range of applications on the platform to, to talk to third party services, doing it in a very, uh, flexible, uh, extensible way. Um, and, and previously what we had was, you know, multi-tenant, you know, service brokers, which were okay for some scenarios. Um, but it was, uh, there'd be some, you know, noisy neighbor things that would come up every once in a while, a few other drawbacks in terms of the operator developer workflow would need to be pre-provisioned and so on. And so thinking about, you know, on-demand service brokers as, as a way to, um, alleviate some of those, uh, behaviors that we had seen in the past and have a much more flexible model where services could just be, um, created on the fly, instances of those services could be created on the fly and bound to an application, um, when they were, you know, really required at the, at the, at the time that it was really needed. So there's a couple of things there that offer a little bit more of, you know, flexibility

for, for customers. Um, I think the other exciting thing is that this is something that will, uh, dramatically, uh, lower the bar for getting new tiles into Pivotal Cloud Foundry. This is something that we had heard a lot from our partners, people wanting to add their own third-party tech on top of Pivotal Cloud Foundry and this on-demand approach, um, will really open up the, the floodgates, I think, for a lot of additional, um, tiles to come onto the platform, which is, uh, a nice win for the, for the ecosystem. Yeah, that's great. The single instance stuff, the current model's fine. Again, as you mentioned, a lot of cases, they're multi-tenant, but single instance. So if somebody wants to beat up the RabbitMQ environment, it would be nice to be able to just on-demand spin up my own environment. So that, that's huge stuff. So we have the framework there now, and then we're going to see more and more services adopt that, right? Yeah, that's right. And there'll be a lot of, I think, exciting things to come as we start to get, um, additional tooling around how this, uh, how this simplify, how this process will be simplified. And we start to see some additional third parties come on, come online. So, so watch this space. Yeah. So last week you, uh, published a provocative blog post about opinions and everyone has them. And specifically it kind of came out from, well, I guess tell us where this came from and what the spirit of the post was about. Yeah. You can't really have a bland, uh, blog topic about opinions. It would be a little, a little weird to have that. Um, but I think that, yeah, back to this opinions idea, I think that we have these opinions about, um, you know, Pivotal Cloud Foundry and that, um, you know, strong opinions help you go, go faster. And a little bit of the, of the origin from it was, um, that we've, we've seen a lot of people with the, the do it yourself route, um, not really go with an unopinionated type of platform. Um, and then they kind of, you know, have the, the, the slog, uh, there, which I know, uh, Coté and Richard, you talked with, uh, uh, uh, Walburn about, uh, the other day. Um, but there's also starting to see this notion of the opinions being adopted by some other open source communities, which is always great to see when there are others that are kind of, you know, thinking about the patterns that we've built in the Cloud Foundry and kind of, you know, taking that, taking that on.

I think there's, um, some news that's been out there. There's a nice piece in the, in the new stack about, uh, the service broker work that the Cloud Foundry Foundation has been doing with some other open source products with, you know, Kubernetes and thinking about how we can extend the service broker API models, some of these other type of type of projects. And, uh, also very recently we've seen how, um, other projects are wanting to, to do containers and get some of those benefits, um, but not necessarily tied to a specific, you know, proprietary container format. So thinking about what, you know, Cloud Foundry has done with, you know, garden and run C and thinking about that container tooling versus specific vendor distribution, people are starting to kind of, you know, come, come online with, uh, opinions that we've had for quite a while. And I think that's just kind of your goodness for, for the industry when people start to think in, in similar ways that have been validated, you know, technically to be really successful. And it also helps, um, our customers as well, see the opinions that Cloud Foundry has, how it's gaining momentum from other places and how they can in turn, you know, use that to help with our digital transformation. So, so on, on, on this topic, what do you think, uh, what do you think people's hesitation is about buying into a platform? And let me do the thing where I ask a question and then talk for a little bit. Uh, like, like, it seems like no one, as, as they say, their debates are tonight, so I can use all this posturing, right? No one would agree that like, you should write your own operating system. And very few people would say you should write a file system and on and on and on. But then the, uh, as, as, as we know, as a, uh, as a purveyor of platforms, there's a lot of people who are like, I should build my own platform. And, and I don't know. I mean, obviously you've thought about this. Like, what do you think, uh, what do you, what do you think their, their thinking is in, in, uh, in an empathetic way? I mean, obviously their thinking is totally wrong and they just need to call up their local pivotal rep and buy from us to be snarky about it. But like, what, uh, like, where are they coming from with that? Yeah, yeah. I think a lot of it is kind of the, the, the maker culture that I know that you guys have touched on in the past.

Um, I think there's also a lot of excitement around, you know, some of the, the open source tooling that's available now with, with containers and infrastructure as code and all the things you talked about in the, in the recent podcast is that you've got, you know, dozens and dozens of products now you can stitch together and as a, as a, as a developer that is, you know, in this full stack way of thinking, you get really excited about thinking, oh wow, I could, I could really build, build this and really help our, help our business. Um, but there's really that, that reality of the, the day two operations that come into play, where if you build a platform, even if you're successful, a couple of years down the line, several million dollars into it, you still have a business to, to get on with, with transforming, you know, so I think it really is this kind of, you know, maker, maker desire that exists inside a lot of, a lot of companies, um, where they get a lot of, uh, enthusiasm about some of these new tools that have made it far easier than ever before to experiment with some of these things. But, you know, a little bit of this experimentation, um, gives the idea of possibility that doesn't necessarily mean that you should, you know, go through and you'll fully fund it from a business point of view. And as you know, Matt, Matt will tell you, it's now 2016 and there's a lot more polish. There's a lot more options now, um, than there was even a couple of years ago for something like this. So, um, it's kind of this, I think, enthusiasm for, for building and making, um, but fixated at the, at the platform level rather than kind of the, the application level. Yeah. No, you, you know, I, I, this past weekend, and this shows you how exciting my, uh, my inner thoughts in life are. I was thinking like, you know, what would I do if I had any choice? I should go get an economics degree. That seems fascinating, even though it's called the dismal science. And then, you know, like most people, what I'm really interested in, I guess is what you might call like the, uh, the micro micro economics, right? Like how incentives work and how people decide to do things. And I don't even know if that's economics, but that's all that's interesting. Like, I, I, I don't know what to do with interest rates and stuff like that, but it seems like, uh, it'd be interesting to have like a sort of like super micro economics of software development design. And speaking of like snoozers and being fun at parties, but it'd be fun to study like

the, the way that people think and the decisions that they make about their, their code and architecture and things like that. Cause I think, I think as always, a lot of it is sort of like fashion driven and just sort of like, not that that's a bad way of putting it, but the incentives are not always straightforward. Just like you were saying, it's like you put it a very nice way. Like, uh, to put it another way, like, well, what do you expect programmers to do? They program. So if, if you put a problem in front of them, their first thing is like, let's program, which, uh, but then it's, it's the interesting question is like, but then there's these other things that they don't program. Like they don't sit down and write a string tokenizer if they're a moderately good programmer. So like what, when does that switch happen and how, what is that thinking? So, uh, I'll write that down on my list of, uh, things to do when I win the lottery to, uh, how to spend my time anyhow. Nice. Yeah. I think we should, we should really get a GoFundMe going for you, Coté, to get that, uh, micro, micro economics. I would, I would love to see, seriously, I'd love to see some of the data and the output around companies with this do it yourself approach versus the, the, uh, the, the buy approach. And that's a little bit of that ROI kind of conversation you're talking about a little bit earlier, you know, Coté. But yeah, I, it's fascinating to see what that economic, you know, data is, what that output, what that output actually would be for those two different scenarios. There you go. Well, on that note of the dismal science, it's a, it's rainy and overcast here. Are you guys sunny up there or is it rainy? No, it's going to be like 72 and sunny today. So don't tell people don't, don't move here. Very nice. Yeah. I, I have, and, and, and I, I'm trying to empathize with you guys. I've had this, uh, this feeling that I need to just nonstop drink coffee today. So I get it. I get why you guys are so big into coffee up there. It's, it's a, it's, it's a good protection against overcast skies. Now I guess they drink a lot of tea in London. So that's their, their fix, but man, you, you gotta, you gotta have coffee. So, uh, Hey, thanks for being a guest. Uh, this is fun where, where, if people want to, uh, you know, follow more stuff that you

do or interested in finding out more, where, where should they go? What should they do? Uh, I guess maybe not go to my Yahoo email account. Um, best option. Uh, best bet is for me is on, uh, Twitter, uh, at Jared Ruckel, J A R E D R U C K L E. Uh, yeah. And hang out there, um, do a lot of blogging for pivotal and, uh, try and keep tabs on what's, uh, what's shaking in the industry. So I'd love to connect to people that way. Yeah, no, you're, you're, uh, you're, you've brought some really good posts to the blog. It's, it's always good to get like new people on and see what they post. It's, it's, uh, it's fun to have good stuff in the blog. So that's, uh, you can follow along in the pivotal blog as well. That's, that's good stuff. And as always, you can follow, uh, Richard and I in Twitter. I'm Coté. What are you, Richard? At R Saroder. And, and this has been, uh, thanks for listening. This has been Pivotal Conversations. It's always, as we were joking about earlier, it's always good to, uh, add in some, uh, some reviews on iTunes or just recommend it to your friends or just talk to us wherever and, and, and tell us that you're listening to it. It's, it's nice to hear. And if you want to find the, uh, the show notes, we'll put links to the, uh, the blog posts that we mentioned and, and, uh, and the news items. If you go to pivotal.io slash podcast, you can find us, uh, Pivotal Conversations there. And you can always go to SoundCloud as well to find the, the, uh, the secret backend where we host our podcast. It's soundcloud.com slash Pivotal Conversations. And you should just go ahead and subscribe and get it downloaded automatically. And with that, we'll see everyone next time. Bye-bye.

Tanzu Talk · 🍕 Get more pizza.