The Smells of the Kubernetes Community

Episode 262 · Fri, 05 May 2023 13:56:33 +0200 · 45:26 · SRT · archive.org · all Tanzu Talk

Well, it feels like the old KubeCon was last week, but it's actually two weeks ago, I think, which makes, I've actually had that thought several times this week, and it made me think, well, what did I do last week? I can't really remember. Do you, do you, do either of you two keep like a, like a daily log, journal, a diary? Not so much like, you know, your reflections on life and the passage of time, but more of just like what I did today, like, so that you can look back and be like, what did I do last week? Because I don't remember. It looks like you do though, Ed, what's, what's your method? You know, mine is clearly very analog. Hopefully there's nothing confidential or even readable on there. Oh, it's, it's blurred out. So you're in luck. It is blurred very technically. Yes. Um, it's just, it's a way for me, I suppose, to try and focus. I'm not sure it makes me any better. Um, and my filing system is just a big pile of A4 somewhere there. You know, I've got it for like the last however many years, right? Six, seven years. Here's what I did this week. Here's last week. And that way, if, if I fail to record it and whatever system we're meant to be recording in for whichever team you're in, I always, always know I've got something I could go back and say, actually, that's what I did over the last however many months. Um, interestingly, I find it's, it has to be a mix of work and out of work as in stuff I do for the kids, stuff you do. Oh yeah. I kind of keep a note. So that way, if I end up looking at something and going, I didn't do much that day work-wise, but there's a huge long list on the right. It's like, right. I got distracted by car breakdowns or whatever it might be. Oh, right. Last week, my car broke down on the school run. So that was a bit of a hindrance, but, uh, Wow. That's so admirable. I wish I did that. I try, I try to do exactly what you're saying. I have a, uh, depending on what editor I'm using that quarter, because you know, I'm always rotating my tools way too much. Uh, but I, I try, I keep, I keep a file called daily log and, uh, I like to put in, this is great for my to-do list management. Cause I have one to-do list manager. And then I also put to-do list things in the daily log file. Right. Uh, because, you know, I don't know about YouTube, but the thing with a to-do list manager is after a few months, everything is due today. And so it's more of just like a reminder

place to have everything. Whereas in the daily log, it's actually stuff I want to do that day. But then what happened recently is I start rolling over the to-do list stuff. Uh, so, you know, the eternal to-do list problem, but I, I, you're, you're inspiring me because I always keep my, my personal journal separate from my daily log journal, but I think you're right. It just has to be the same. Like you got to put all the stuff in there, right? Like, you know, that bit of paperwork for me in that it's what I actually did in a day. No, that's what I should have done that day or anything else. And also the only thing that's ever worked for me is a bit of a four on a pen. Yeah. My laptop. Anything that makes sense. You know, like you said, I've got to-do lists, I use to-do list. That's great. But stuff doesn't get entered into it. Whereas a bit of paper, I'm just like, I just spent 20 minutes doing this and jot it down. And it kind of just works. It's the only lightweight system that I've maintained. A little captain's log, personal entry. Yeah. How about, how about yourself, Ben? It sounds like maybe you, you don't have a daily log. I've read so many times that it's a really good skill, a life skill to have, you know, that it's good for your mental health as well. It allows you to, you know, focus, like you said, and also remind you of the things that you've been spending your time on and maybe give you an opportunity to change the balance and stuff. So I get it. I get, I get the theory and the, and that's the practice. I just, I just don't do it. I don't even have a piece of paper on my desk. Yeah. You could say that, but I haven't replaced it with like a to-doing app or anything like that. It's like, okay, if people want stuff, they'll ask me for it at least twice. Yeah. There's, there's my, there's, there's my prompt. Get a to-do list, Ben. I'll ask you again at the end and then you're done. Yeah. You know, I, I, I used to use... I wasn't an ask though. It was an order. Uh, I didn't do orders. Uh, no. I used to have this, I used to use this to-do list thing called OmniFocus, which is just like,

like if there's something beyond a Swiss army knife, then that's what it was for like, uh, you know, for, uh, for to-do list management. And I'm realizing it had, it had three categories of a due date. One of them of course is nothing, just there's no due date. And then, and then of course there's the actual due date. And then it had this other category called defer, which basically, which I never use, which basically meant don't show this until this date. And, and so I think maybe that's what I, that's probably how I schedule things more is, is I, I defer them. Like I don't want to look at it until a certain date. And that's why it always piles up with, with today. It's like, it's not like it's due today, but I need to be like aware of it. And, and at some point, at some point that CFP is actually going to be due today. So I want to make sure that I see it far enough ahead of time, but you know, to-do list management there. This also reminds me when I was, when I was a younger programmer, there was one person at the first programming job I had who showed me, man, I can't remember the abbreviation. It wasn't called TPS, but it was some system for programmers where every 30 minutes you were supposed to check in and log what you had done. And like after you can imagine, as, as maybe, you know, uh, Ed, after like several years of this, you could go do an analysis of how you're spending your time and how you're doing, you know, doing well at it. And it, uh, it seemed like something that you would have to use if you were like developing missiles in a highly regulated environment. And, uh, people would come in and want to audit how you spent your time and things like that. But I don't know, seemed awesome, but a lot of work. I do recall. Yeah. Some, some nineties management technique around that. Yeah. Yeah. Maybe it's like TST or TSRP. I don't, I don't know. Anyways, speaking of how we spent our time, Ed and I were at CoopCon and I know you watched it online, Ben, uh, which, which is exciting, but let's, because I know what I did. Uh, so I want to entertain

myself. What, what did you do while you were there, Ed? Wait, wait, wait, wait. Do you though? Do you really know what you did? Well, maybe that's, that's more appropriate because I don't know what I did. I, uh, I can't go over it. And now we get to the truth of it. Fill him in on what he did. What was actually going on? Um, so actually, I mean, so my experience, it was interesting. I guess that's my first conference in probably three or four years, I think post COVID. I don't think I've done any others. And I couldn't remember when I went before one from actually, it's actually the one thing for me, aside from all the different technical KubeCon specific stuff was just nice to be mingling again. You know, I know a couple of people did get COVID there, but it didn't feel like a big deal. It just felt like life was normal. You can go back to a conference. It was a nice conference in that case. And in terms of being a big conference, you know, 10,000 people sold out. Um, it really reminded me of early VMworlds. And I say that, you know, I haven't been to VMworld for quite a few years. I used to go all the time before I joined VMware, but once I joined VMware, there was no need to go paradoxically. Um, but it had that energy, it had that buzz in the air. This was my first in-person KubeCon. I've done a couple of the virtual ones. I don't think they count. You know, there's no buzz in the room when it's virtual. Um, but exactly. You can feel the energy just looking across floor. I mean, yes, it's a trade show, but it was still actually my first day. I got a bit of that buzz day one buzz. And it was like, Oh, I haven't had this for ages. So excitement. I don't know if you can say that for a tech conference, but you know, I'm a, I'm a geek, so it works for me. Um, so yeah, I like that. I spent a lot of time kind of wandering around the solutions exchange. Um, because for me, KubeCon is an interesting one, I suppose, in that a, there are a ton of brand new people. Um, and I've been brand new vendors, brand new solutions, apparently 58% of the audience were new as well, which I still find surprising. Um, but either way, I guess that's sort of where all the energy comes from. So there was a lot of like, who are they, what do they do? Um, and a slight feeling for me of I'm beginning to get a handle on it. Um, you know, I was sort of new to

the cloud native stuff a year or two ago. I used to be able to turn up to VMworld, know all the vendors, know a lot of the people, um, drop right into that whole community thing. And that was a brilliant sort of conduit to information and conversations and what's coming up. And I'm still kind of gaining that in the Kubernetes world. Um, so it was a nice way to sort of see who I did know, who I didn't know, how many I recognized, what was going on. Um, and to try and get a feel for themes, right? Which I don't think you can ever do if there's 10,000 people and more breakout sessions and there's never time to watch. You know, I've got a list of sessions I will probably never watch. I'd like to. And that was when I was there. I, each day I was like, right, what's on my schedule? These sessions, some of them I couldn't get in because it was just super busy. You get to them. I got to one session about a minute beforehand. It was full. I, so I learned my lesson. I went to next session 10 minutes early. It was already full. Uh, and it was like, right, okay, maybe I'll just try and watch some of these virtually from a desktop somewhere because that might be easier than walking across a conference hall and not getting in. Um, or I'll catch up afterwards, which we will know probably never happens. Um, but yeah, you know, I, I sort of got the same themes as everyone else and it could be that self-fulfilling thing that I went there to look at several themes. So therefore the conference seemed to be all about them to me. Oh yeah. Yeah. And, and then, and then, you know, based on what you're saying it, you know, if, if you're looking around for a bunch of separate parts, then you're like, Oh, there's a lot of separate parts instead of like, which, which of course fits with like an infrastructure thing, like Kubernetes, right. Where there's, uh, all these different components with like, um, Oh, you know, depending on how community versus vendor you want to be with competing, like, uh, modules to fill it, competing plugins. Boy, look at me using vague words. I don't even know the Kubernetes terminology for like, you know, this thing fills in this component of, uh, of something, but yeah, it, it does seem like, you know, there's, there's like at least two options for each way you would do something in Kubernetes if, if not more. And, uh, and, uh, did you go to the community

pavilion or whatever they called it where there's like, I mean, it's a bit mixed because I mean, it's interesting, I suppose, KubeCon being more open source focused. So actually a little less vendor focused, you know, there's not a lot in way of both sort of product updates, even in the keynotes, there is no big vendor coming on stage saying, ta-da, version X or something like that. Yeah. All sort of iterative updates. Here's how these projects are going. Here's how in general this ecosystem is working. Um, so going over to the sort of the area where it is all braced on the, the open source projects was quite interesting. Less vendor pitching, obviously a lot more, here's where we fit in a lot more kind of better together stuff. You know, we work with this, we're hoping to do that with them. Right. There were a couple of sort of ad hoc talks on that floor as well, which were interesting. So working groups will meet up just at those booths and it might make it to a schedule or something. Some of them were more informal. So it's not on a KubeCon schedule, but if you're hanging out in the right Slack channel, somebody says this group will be at this booth at two o'clock and we'll have a chat. And I dropped in a couple of those for like the, the app, um, SIG group. And it was literally maybe 10, 12 people tops. Um, just having a discussion around the booth and I felt a bit like an interloper because I'm not really someone who hangs out in those special interest groups or anything. Um, but did have an interest in them. So you're sort of listening on the edge thinking is someone about to say, who are you, what are you doing here? But actually they're all, it's all very welcoming. And if they're in, if you're interested, then you can just listen in. Um, I then tried to get into some of the follow-up sessions, but they were full. So, um, but yeah, I liked that sort of, uh, the project part of the floor. It was interesting to just see what was going on. Um, and that the advantages of like 40 booths are all tiny and you only have to walk a hundred meters. I cover a lot of ground versus some of our conferences or others where it's in hall B across the road, catch a bus, you know, if you go to reinvent, plan half an hour to reach that one here, they're all there within three minutes walk. Yeah. Yeah. And, and there's also tons of stickers over there. That was very well coordinated. Did you get any of them? I didn't know. I'm a sticker free guy. I'm weird in that way. I see

all these people throwing for every sticker. I kind of get the urge and then I know because they'll sit on the desk and I'll bin them. That's no good. You, you, you should get one sticker for your laptop. That's like, I don't, I don't know what they do, uh, you know, over in y'all's neck of the woods, but like no bills posted, like just like, like have, have that on there. That would be hilarious. Well, yeah, I, I, uh, uh, I spent some time hanging out with my friend, Matt Ray from my other podcast and he was at the, to the point I'm about to make, I'm going to mix up the names of things. He was at the open cost booth and he works at Kube cost. And it was, he told me this would happen, but it was hilarious. Like I was there, you know, we had lunch and I was just hanging out and maybe about the time I was hanging out, maybe 15 people came over and always the first thing they said is what's the difference between open cost and Kube cost. And it was just, it was amazing. And then he would go over these two things. So it is like, but then also from, from that booth, I could see the Carvel booth, you know, we got a lot of VMware people working on there and, uh, that was pretty, pretty busy and interesting, but yeah, it's, it's fun to see the, um, activity amongst all the open source projects. And I was talking with someone, you know, cause we're, we're in, in the business here, you're always speculating on this stuff. I was thinking like, you know, they should put the, the way they had the boot, the open source stuff arranged. It was sort of like in its own little grotto, if you will. And, uh, I don't know, it seems like it should have been more, more front and center. Maybe not, you know, we had a great booth of VMware near one of the two entrance doors. So those are, those are very sought after probably charge for positions, but you know, you could put the, uh, you could put the open source people in the center of the area there. And that might be nicer. Who knows? That was lunch. You can't do that. Oh, that's true. That, that, that was lunch. Yeah. Yeah. That's probably more the, the money side, isn't it? The vendor. Yeah, of course you have to walk through it to get to the other stuff. Yeah. Yeah. Which is totally

cool, but whatever. That's fine. You know what they should have is they should have like a sponsored lounge area in the middle of the open source thing where you can like, you know, sponsored by well, VMware, of course. And there would just be like really comfy couches and like things like that. And lots of plugs that you could hang out in. And it would be like, it'd be a way to like have, as people have discussions with the open source people, they could like, you know, little groups, they could move over to the comfy couches. And then it's also just like a nice little space there, which, which should be fun. That's my suggestion. I'll file that at suggestions at CNCF.io, which probably has one of those replies that says, no one checks this email box. Please do not reply back. I will say they, there were a couple of things that cropped up and I can't remember where, right? There's too many Slack channels and forums and things, but somewhere I remember stumbling across a kind of feedback forms for things. And it was people complaining about sessions being too full. And, um, at some point someone was being overly efficient at closing doors to rooms and saying, no, this is full. When it wasn't, there were loads of chairs empty and people were complaining. And actually very quickly, people from the CNCF came back to it. People saying, hi, I'm partly organizing this and the other, you know, this bit's out of our hands. It's being done by sort of health and safety on site who run the conference. For sure. And they got it all sorted and they got people, and you know, within a matter of hours, everything was being addressed. I got the impression they were very on top of things actually. Yeah, they are. They, the, the, the CNCF and the Linux foundation people are, are, as you say, very professional. Like I was there, uh, on, on a media pass through, through my other podcast. So I, I went to some of the, uh, you know, like the press conference and then like the separate media sessions. There was one, you know, someone at the new stack wrote it up and somewhere else, but there was one on platform engineering. This is where they had vendor speakers. Cause they had, they had someone from Red Hat, Hashi, whatever. And, uh, there was one other person, Git Labs.

And I think, I think they were on that panel talking, but then they also had like a, uh, you could call it a customer panel, someone from Audi or however you pronounce that. And of course, CERN, CERN's always showing up to these things. Uh, and, uh, I think like some American bank or something like that was there on the customer panel. And, you know, one there also and everything on there. Yeah. Who, who was it? I forget. I forget who it was, but like, I should, I should go back and watch that recording because like, I was, I was kind of like zoning out a little bit, uh, because it wasn't like they were, they were basically just talking about how well things were going. And I, I like, once I kind of hear that, like my mind kind of disengages, cause I always want to hear the, like, you know, things we're doing best practices we've come up with and things like that, but they seemed happy with, uh, their Kubernetes usage, but also see if there's anything. I was going to say useful, but advisable to extract. What's that? I think I might've seen that, that panel. Um, was it part of the sort of keynote, the sort of start of the... No, it was like, it was like in the media conference thing, but, but anyways, those were run very well. And, and as always with the media thing, there was a, a better buffet lunch than the box lunches. They had the, uh, they had the, the, the well cooked tender chicken, uh, chicken thighs section. You know, that's a very popular conference food to have the chicken thighs. And of course the salmon with creamy sauce on it and, uh, some other unique things. Uh, I knew I should have come and found you at the conference. You should have upgraded my experience. Yeah. How about yourself, Ben? You were watching remotely. So you actually, what are some sessions that I, you know, just putting it out there, seriously jealous you folks getting to go and to, uh, to experience that, you know, I have serious FOMO of missing out on that, but, um, I did get to see it sort of from afar and, and a lot of the, um, the session videos have dropped this week.

So I've been, I've been, um, going through and picking out ones that I think are particularly interesting. And, um, of course, cause VMware had a lot of people there. Um, I also saw the, uh, sort of, um, collaborative notes, if you like, you know, everyone's into one doc and just sort of puts their notes down and stuff. So that was brilliant. Cause that sort of, you know, gave you an idea of what the, what the themes and stuff were that, that went, that was to interrupt you. That was really good. It was like, I think that is the most extensive trip report notes I've seen ever from a company. And, and it really added to the, I mean, it added to the, the overall sense of like, there's a lot of stuff going on here because it was anyways, but that was, that was nice. It's a better way to take a pulse of a conference, isn't it? The one person's opinion because everyone sees different sides of it. So you get 20 people saying, here's what we see. And you get a much more rounded view, even though it's probably still VMware centric, shall we say. So I think that was really cool. And I think what I sort of, sort of took away from that and sort of felt anyway, just, just looking through the, the, the video playlist, for example, from the CNCF site is that this conference seems to be really more of a CNCF conference. They still call it KubeCon, but Kubernetes is becoming, um, predictable and mature. And it's getting to the level where you don't have to talk about it as much. It's getting to the level right now. And it's getting to the level of design. It's getting to the level of design. It's getting to the level of design. It's getting to the level of design. It's getting to the level of design. New, uh, has a lot of, um, interesting things going on, things changing fast and, and, and changing for the better very quickly. And I think that came through in the talks as well. You know, the, the talks and the themes there, like I said, looking through the sort of show notes doc that, that we made afterwards. The themes were things like, um, wasm coming up a lot, eBPF coming up a lot and backstage coming up a lot.

You know, these are all sort of side issues, if you like, to Kubernetes itself, you know, Kubernetes is becoming a sort of a, um, a nicely stable sort of predictable background. Not that there weren't talks on that. There were some great talks on that as well. I saw a really good one that was about how to kill a Kubernetes cluster, which I thought was a brilliant talk. I saw that. It was the first session I couldn't get in. I got there and I was like, what? And I, at first I saw these huge crowds and I was like, are they there for this session? I thought this was a bit, bit niche, like an operator one almost. And there was queues out the door. And I was like, well, I'm not getting into that one then. But obviously the idea, obviously a lot of people have killed clusters. And I'm like, how did they do it? Yeah. And it was a, you know, it was a developer talk. So it was about a sort of, you know, first of all, how an architectural choice can have an impact on your needs from a cluster node. And then secondly, sort of optimized settings and how sometimes maybe the documentation doesn't really line up with what you actually want. So it was a great, great session, a good example of a good Kubernetes session, right? There's still lots of Kubernetes sessions, but there's just a ton of stuff going on all around it. And I think that's the exciting part, particularly for folks like myself, you know, with a developer background and yourselves as well, because it's not just infrastructure anymore. You know, there's a whole bunch of other stuff going on, talking about golden paths or talking about, you know, how to do things like service mesh through eBPF and stuff. So there's loads of stuff going on. I'm really jealous I didn't get to go, but I'm very thankful that the videos appeared online so quickly while it's all still sort of fresh in everyone's mind. So go ahead, have a look at the YouTube playlist. And we've picked out a few I think we can put in the show notes as well. So, yeah, go check them out. Yeah. And, you know, you're pointing out like a distinction. And by we, I mean, all of us in this ecosystem and community or whatever.

But like, you know, I guess and people do this, but like, I guess we've all got to like start thinking about Kubernetes as a core thing and then and then all the other stuff you layer on top of it. Like it's almost like I don't tell me if you think this is fair. It feels like there's some there's an increasing amount of things. And I think Backstage is in this category that just use Kubernetes. Like they're not really Kubernetes. They just run on top of it. It's more about modernization, I would say. So, you know, Kubernetes is often part and parcel of a strategy of modernizing in particular, you know, business moving to the cloud, doing all those sorts of things. And alongside that, you have to change your development practices quite significantly in order to make the best of it, in order to optimize for that environment that you're working in. And Backstage is one of the one of the really good ways that you can do that across a large group of developers. So I think that's why it's it's there, really. That's why it's sort of mentioned in the same breath. It's because you can run your Backstage on Kubernetes, of course. But that's not really what's being discussed. What's being discussed is things like how do you offer a golden path? How do you document well? How do you make things easy to create and manage and explore? Right. How do you how do you do these things inside the the walled garden is your enterprise when you have hundreds of developers possibly working in in lots of different locations, you know, remote locations? How do you bring all that information together in a useful way? So I think it's it's right to be there. But it it's another case of sort of it's a CNCF conference. Now, it's like you said, there's a lot of open source there. There's a lot of other stuff going on that's sort of around the sides of the of the of the, you know,

Kubernetes singularity that's in the middle. But there's all this sort of galaxy of stuff going on all around it. We see that in the co-located days, right? That's a whole day before the conference begins now. And that's with security already taken out and running to its own conference. And you can easily see then, you know, fast forward a little bit. Those co-located days become more events on their own. And suddenly, actually, the KubeCon piece could eventually shrink down to being a lesser core because it's all around the sort of trimmings around it or like you said, everything's sort of growing up and moving up the stack a bit. Exactly. Yeah, like and as as you're I guess I guess you two are reminding me that it's the it's the cloud native computing foundation, not the Kubernetes native computing foundation. Right. And so like and that like I don't know enough about WebAssembly and stuff, but that would be another example of like, well, this isn't like a Kubernetes thing per se. It's its own thing and it's a cloud native thing. And so therefore it belongs in the CNCF. And then like you were saying, Ed, like eventually there's like KubeCon versus cloud native con. And it's just Kubernetes is just another thing in that umbrella, which which is, you know, that's great for the CNCF as far as its magnitude, which I think, you know, people like magnitude. It's good to be a large, a large growing organization. I got to hang out with some of the CNCF people and they well, one, they seem to have nice jobs. So they're they're well taken care of. So good job on the Linux Foundation. And then two, you know, they're they're all cheery people. It's it's it's fun. They're they're happy, happy. They don't really complain a lot or have bad things to say. They're not like, you know, there's certain areas of the open source world where it's all about like like talking about how, you know, there's evil out there that that must be defeated with the open source toolkit, which is not nice to not be around when possible.

Well, narrowing down to KubeCon, although maybe when we do this survey next year, we should take into consideration Kubernetes versus cloud native. But I think we talked about this a little bit last time, but we finally finally it seems like finally, if you work on it for a long time. But we finally published our own state of Kubernetes survey, which I have written. Only one blog post has been published so far, but I was asked to write three of them. So I've spent a lot of time looking over the PDF and taking screenshots of it here and there. And, you know, well, one, if you're listening to this, you probably would be interested in it. So you should get it. And if you go to the show notes at TanzuTalk.com, you can download it or you just search for state of Kubernetes 2023. But I think I'll just call out a couple of things that I thought were interesting. One, you know, I was surprised that we do this, but we've actually done it for several years. We actually put in like the survey results of market share and including us, like all the I don't know what to call this anymore. Maybe, you know, from your reading, Ed, but like the public cloud hosted Kubernetes services. I don't know if that's managed Kubernetes service. It's just weird to say public cloud, but whatever. All the KSs that are there, the star KS and then the on-premises ones. And I think if you were to look at our friend VMware Tanzu, our Kubernetes stuff, it goes up. I should have the chart in front of me, but like four to five percentage points, if not more, every year. So you see it climbing the rankings there. I think 14 to 20 to 25 to 29. There you go. Yeah. Did I do my math approximately right there? Bang on, bang on. Yeah. All right. Good. I've, I've, I've built up an intuitive sense of charts. I used to, I don't know if I've told you all this, but when I was doing a strategy, I, I wanted to have some programming because sometimes, you know, you'll get a chart and they remove the numbers off of it because they're being.

Yeah. Secretive. Yeah. And, and I was like, you know what we need is we need a pixel counting thing where you could put, you could put a chart in there and it would count the pixels and then you could actually get numbers around it and estimate, you know, to see, that reminds me of something else that I saw on Twitter today, which is a year progress counter. And just every day, that's all it is. It's one gif of like, okay, we're up to 24% through the year or whatever. Oh yeah. That would be fun. Sorry. I interrupt. Go ahead. Anyways. Yeah. I think I, as you know, the rankings, I think, you know, it's a, I think it's always interesting, always interesting. Well, it is always interesting that. That Amazon is always the number one in Kubernetes stuff, because if you watch the Kubernetes documentary, basically Kubernetes was made as a competitive. Gambit against Amazon. And then also Amazon never shows up in the Kubernetes documentary, except as a drone shot of one of their buildings. So it's always this, like this shadow figure in the background of the Kubernetes documentary looming there, like this untouchable unknown thing. This. Well, yeah, but they sold them the drone and the camera. All the memory cards, all the dollies, the lighting. Yeah, that's right. That's right. That's right. Anyhow, it's obvious that Amazon has fully embraced the Kubernetes thing and vice versa, as far as the market. And then I think, I forget. It's close though, isn't it? I'm looking at a chart. It's close. Like, AKS is right there. Like, you can get a piece of paper between them. So, you know, it is close. Yeah. It's not like it's got a runaway. And then going down there, OpenShift occurs. And then I think we're right under OpenShift. And then you have the Google one. Yeah, GKE, then Rancher, then IBM. And then I guess there's another thing to highlight, highlight, to point out about the survey that I think is really valuable for what I care

about is, it's basically, I would say it's for large organizations. And if I remember, it's like, this goes to Ben's joke about like, do I really remember what I did at KubeCon? Do I really remember these three blog posts that I wrote? No. But it's basically, it's organizations of 1000 or more people. And then it skews very heavily towards much larger organizations, which I think is great, right? Like, because it means that, well, it's great, because that's the kind of people I talk with, large organizations. So these trends are a lot more reliable for what big, big organizations, big companies, and other people are doing, which I think makes the market share stuff more interesting. But how about yourself? I know you all have read the survey in detail, probably put it on, print it out on A4 paper, highlighted it, you know, little notes. What did you think was interesting in there? Much same. I mean, I was looking at the market share stuff, because I've seen sort of flex errors recently. And there was some news at KubeCon about IBM making layoffs, this, that, the other. But, you know, therefore, you kind of, I'm always trying to match up in my head. How is this working out? Do these surveys agree? And they never do, right? Different vendors, different analysts, there's always different figures, pick your audience. But then largely looking at those sort of adoption figures was the one that I find interesting. But actually, you know, there's, there's almost never a great takeaway. I mean, I guess, like I said, looking at AKS versus DKS, they've all been in the lead for a long time. Oracle's IBM's always come in somewhere at the back. I like to absorb them all. But actually, does any of them change my worldview or how things work? Not a lot. But yeah, the other thing, I guess, is just that there's still a lot of that middle ground. By which I mean, platform engineering, where does it go next, application platforms, that kind of thing, right? That's still a gap being filled. And therefore, you know, does it actually have that impact on speeding up developers?

And it's whether surveys show whether that, how that bit is developing. And is it working? Everyone thinks there are still benefits there. I'm just not sure they've been fully realized, which I wouldn't expect that space is still immature. We're still figuring out. Everyone, I think, sees the benefits of Kubernetes, whether it's an operations level, whether it's for developers. It's just whether that's how quickly it's going to really work. Yeah. That's where my mind is focused at the moment. Yeah. Yeah. We didn't publish too much on this in the survey, but I look for this in other surveys. Like, it's always interesting to weigh. What's the word? Like, hopes versus reality, like, or, you know, a for ahead of time perception. Like, these are the things we think will like why we're interested in this. And then, like, check in a year later and see what benefits versus achievements. There you go. And I think this year we started with a little bit in the survey, like, asking about, you know, business benefits, which, you know, you were kind of alluding to this. We're relatively small at the moment. But I think it gets to the point of, like, everyone has big hopes and dreams. And we'll see how it pans out. Right? So it's hard. It's hard for people. There's not enough people around who have used it for long enough to have much to say about long-term hopes and dreams to reality matchment matching. But it's a very positive survey, though, isn't it? Like, you know, you look through it. It is. And that's what struck me about it. It's very positive. And there's a lot of progress being made. I think that the two things that stood out for me is a lot of progress. Nine out of 10 saying that they agree that cloud technology, including Kubernetes, is transforming the way the business operates, like, that's a really positive endorsement of all the things you've been saying for years, right, about how moving to cloud helps you transform your business, helps you modernize, gives you that flexibility and that freedom.

So I think it's good to see survey respondents bearing that out across, you know, a broad sample set. No, I think like you're saying, I think the summary that Ed was looking for is the arrow is still going up. Yes, up and to the right, up and to the right. It's a very long, rambling version of that. Yes. It might be a slightly, like, squiggly arrow. But generally, if you were to draw a straight line through it, it's going in the positive direction. Yeah, yeah, for sure. The other thing that interests me is there's a line at the end around three-quarters of folks in the survey saying that they're utilizing multiple clouds. That seems like a good, healthy, large number. And it sounds like multi-cloud is becoming the norm. But it also sounds like there are 25%, you know, a quarter who aren't there yet and on, you know, making the best of different vendors for different strengths and being able to, you know, spread their bet across multiple clouds so that they have some shielding from, you know, price hikes or, God forbid, one of the cloud vendors perhaps deciding not to do it anymore or whatever, right? Because at some point, there could be some consolidation in this market, right? So if you've got all your eggs in one basket, maybe that's not a good thing. So it's interesting to see that, you know, three-quarters are smart to that and are actually, you know, trying to use multiple clouds at the same time. Yeah, and that angle's the, that's the angle. That's the one, the blog post that's actually out, which, you know, again, you can check the show notes listeners to read it. But like, I was, as I was reading over this morning, I was realizing the only reason I ended up writing that blog post is so I could get permission to use this great IDC slide that I found. Like there, there's two, there's two great analyst things I found in this writing up project. One of them is this IDC slide that shows basically, I think it's four quarters of a survey.

I don't know. I'll have to go look, but it's basically four data instances of something like 7,000, a lot of thousands of respondents saying like what percentage of their applications run on what type of infrastructure, like through traditional on-premise managed hosting, public cloud, private cloud, blah, blah, blah. And like to the multi-cloud point, you look at it and you're like, all right, everything runs everywhere pretty consistently. Like things, things kind of shift more away from on-premises, but like quarter over quarter or whatever the period is in this IDC survey, like there's just a lot of, I wouldn't even call it multi-cloud. It's just multi-infrastructure is sort of like the reality of what's out there. And then the, uh, the second thing that, that, well, we actually have this in a license survey, but I finally was, I was digging around and there's an estimate from Gartner about the number of applications that run on Kubernetes, which I think, uh, in, uh, in, in 2021 or so was, what is it like, it's basically, it was 10% in 2021 and they estimated to go all the way up to 25% in 2027, which is 20, that's four years into the future. And so like, even, even at that point, right? Like it's, it's, um, and you know, not to try to pull the arrow down or whatever, but it shows some, a point I think you were making earlier, Ed, that things are like, uh, early, if you will, that even in four years from now, if only 25% of applications are running on this, this thing, then it means that like, there's still a lot of development and work and figuring things out, uh, in the future. And, you know, since, since, since, you know, Ed, how, how did this play out in the, uh, the VM world, uh, world? Cause VMware, like, I forget when the workstation thing came out, that was like 2003 or so somewhere around there. Yeah. Yeah. And then how long do you think it took for, for us to have this, like, ah, everyone uses VMware sort of like kind of forever. Well, I mean,

there's still the odd outlier now, but, um, a good 15 years, I'd say, I mean, 10 years for probably the mainstream, right. By, by 2015, most people had most things in VMs were still some outliers and, you know, a few cases, but mostly 10, 12 years. Um, yeah. And even then, I think if you took the same surveys, how many workloads around in VMs, you'd still be talking about 50, 60%, because you've still got the mainframes, some other stuff, some niche cases, some industries, regulations, there's always pockets, you're never going to get massively high. Um, but yeah, it, it just takes a long time, even once everyone figures out it's good, because a lot of companies obviously aren't replatforming everything all the time. And I, yeah, when we re-architect that system in five years, we'll put on Kubernetes in whatever shape it's in in five years from now, but they're just not in a rush to do it. Um, yeah. So, yeah, I think, I feel like I need a curated museum tour of the VMware museum or something where we get to find out when certain things came to market and, you know, how they did and stuff. I, I find that really fascinating. If there's any listeners out there already know of a curated virtual VMware tour that you can take, maybe it's in some national archive somewhere or something, then, uh, then get in touch and let us know and we'll share it. I remember we've had loads of internal timelines, haven't we? And I remember we had some up at VMware, like, you know, they were the big walking between Moscow name, Moscow name, what is it? A and B or whatever there was, were these posters with various timelines of either product releases or technologies and that sort of thing. Um, and we've got an initiative running at the moment internally, haven't we? To get stuff to the, um, the computer history museum. Yeah. So that we're, yeah. I recently got my 25 year VMware t-shirt, which is, you know, I've not been at VMware for 25 years, but two and a half years, maybe something like that. But, but it's like, oh my word, there's all this history. I need to find out more about this history. It would be a fascinating, uh, be a hell of a YouTube

video as well. I reckon the history. Oh yeah. Yeah. Yeah. You know, maybe we have one of these, especially because of that 25 year thing, but every now and then big vendors will have like, you know, the, the big vendor.com slash history site where I, I think some had one of these that was fun. And, and you know, it's probably nowadays would be one of those things with that really annoying scrolling where things just like pop up all of a sudden, you're just like, ah, I just want to scroll a static webpage. Get out of my face. I remember I saw one, I forget what it was for, but I saw one long ago and that had, um, instead of, instead of vertical scrolling, it was horizontal scrolling on a webpage. And I think, I think for a week afterwards being like a web native person, I was just like, my mind is broken. You're not supposed to have horizontal scrolling on a webpage. I don't understand what's happening here. Like I wasn't told I needed my track ball to, to like read this webpage. I don't think that would work on my cell phone, my, uh, my smartphone. Oh yeah. Because it's got swipe right gestures and swipe left gestures for the OS itself. That would be, that would be madness. Yeah. That would just be crazy. I would never get to the end of it. Well, it's more the problem, isn't it? The history of, um, and your associations, it's like scroll right, but that means discard for me. That's archive, you know, swipe left to tell you the meanings. I mean, I think we're finding out more and more about, uh, ads, uh, Tinder. Or, or at least his email habits. Yeah. Yeah. Yeah. Well, I, I think, I think, I think finally, I'll mention before we close out, there is, uh, uh, I, I, I gave some review input on from some of our old pivotal, not all of them were, uh, uh, from pivotal people, but, but some of the people who worked on this were some of our pivotal friends from, from way back when, you know, two years ago. Uh, but they, they've been working on a, uh, platform, uh, maturity model, which, you know, platform engineering maturity model paper that they've, uh, submitted to the CNCF.

And I think it's, uh, I, I do remember this one because I looked at it this morning. Uh, but I'm, I'm going to, over the weekend, I'm going to think about the fact that I forget things that I work on and, uh, see, see what I can do about that. Uh, you know, I met, I met one of my coworkers, uh, Whitney, her and her son were in town. And I found out that she uses whatever this method is. She uses flashcards, the kind of flashcard thing where if you get it right, it shows you like seven days later. And then if you get it right again, it shows you like a month later that you have reinforced, uh, flashcard learning. And, uh, it's kind of, kind of amazing. She was like, yeah, that's how I learned all this Kubernetes stuff. Just, uh, flashcards. So maybe I should do that. I'm not sure, but this paper was on your shortlist, isn't it, Ben? Yeah. Yeah. I've got a flashcard on that somewhere. But, uh, I, I think, I think this platform maturity model paper is worth looking into because it's, it gets to another area. Like when I talk with, you know, executive people at large organizations, like part of what kind of like the discussion we are just having, like part of what I want to encourage them to think about is planning out how you're going to do all this magic cloud native stuff. And then, and then also learning from the other way that people have like gone through the process, speaking of scrolling to the right or whatever, like, and, and most, you know, most, most, um, what's the word? Pragmatically thinking like, oh, this is going to take a long time. Right. And so like, that's, that's one of the things maturity models are great for. Cause it lists out, I think there's five phases and it's basically like, here's the phases and the activities that you will doing kind of like the smells of maturity, which having a nine and a 13 year old, I'm, I'm re familiarizing myself with the smells of maturity. But you know, you've got, you want to know the smells of maturity as, as, as you're going through there, which I think this paper will give you a sense for if you're putting like a platform, uh, thing into place.

And thankfully these smells are not necessarily bad smells, uh, or, or whatever, but you know, so there you go. Uh, there's your smells of maturity episode. Uh, and, uh, if you want to get the show notes for these, you can go to tonsutalk.com and we also broadcast these live. I don't know. Whenever it comes up, we'll see, try to do them on Fridays in the morning European time. And, uh, you can find a link to the videos there if you want to see the visuals. But, uh, if you don't subscribe to this already in podcast form, you should look up tonsutalk and subscribe, and we'll see everyone next time. Bye-bye. Bye-bye.

Tanzu Talk · 🍕 Get more pizza.