PurePerformance - What happened in 2022 and where 2023 is taking us!

Episode Date: January 1, 2023

What a year 2022 was! We had 25! episodes with amazing guests from all over the world covering topics from Kubernetes, OpenTelemetry, DevOps, SRE, Cloud Migrations, DNS, Value Streams all the way to Persona Driven Engineering and drawing parallels with Digital Marketing. If you are new to our podcast check out the playlist and listen to some of those we mentioned during our episode!Now its time to say Thank You listeners for the continued support. After 5+ years of podcasting we still see rising numbers of downloads which is the best motivation for us to keep going. Stay tuned as we are going to cover industry relevant topics going into 2023 – or is it year 53? (only those will know that listen to the full episode)

It's time for Pure Performance! Get your stopwatches ready, New Year special of Pure Performance. Happy New Year! Happy New Year, Andy. Man, you are slow. What's happening with you? Were you partying too long or what's wrong? Oh yeah, you know me, I was just, you know, me and my family, we had a jug of green alcohol and we were just chugging it all night long

and rolling around. There's a song, it's called All Night Long. She gave me kachi all night long. What is kachi, should I ask? It's actually a special, I think it's a special massage. I think we, it was a really popular song a year ago. Is this a family show or what? What's going on here?

I don't know. How did we end up here? Because it's New Year's, right? It's a new year. Even though we are in the old year, recording it in the old year. Are we giving it away? Oh, man. Are we time travelers? Yeah, we are time travelers. year recording it in the old year are we giving it away oh man are we time travelers yeah we are time travelers um you know it's 2023 for anybody who wasn't sure

but i thought of this one today it's actually only the year 53 in epoch because you know it's such an it's such an arbitrary calendar. Like all the calendars are arbitrary. So like if we look at, happy year 53 to all you developers. Right? Thank you. Yeah, so useless piece of information,

but thank you so much for it, yeah. Speaking of useless information, boy, do we have an episode for you today. We do. Actually, can you remind me again, did you say 52 or 53? 53, because it's 1970. 1970 is the January. Okay, because it's interesting, because I was just making my accounts, and we have 52 weeks in the calendar year,

and we had 25 or 26 episodes of pure performance last year yeah which was fantastic and in preparation in the long preparation that went into this particular recording today i was actually really fascinated if you look back a year ago and this was kind of the last episode before new year's so december 20th in 2021 we had Christian Heckelman we call him Wurst a lot because that's his handle, his Twitter handle it was really cool because he did

back then he did a presentation at KubeCon which was the top rated session of all of KubeCon in North America it was called how not to start with Kubernetes and that was really awesome we did a blog post with him of all of KubeCon in North America. It was called How Not to Start with Kubernetes. And that was really awesome. We did a blog post with him.

Really great stuff. So folks, if you are just getting started with Kubernetes, I think this is an episode to go back to and also look at the blog that we wrote. It was just Google How Not to Start with Kubernetes and you'll find it. Yeah, it was a great one. And then I think we just both said, right, we didn't remember that we had so many great sessions

because we really, we had 25. We talked about open observability. We had Doten Horowitz, who runs his open observability talks. That was great, getting his point of view. And I think he's definitely like Henrik, right? Like Henrik Rexert, he's really out there and educating the community around

what's happening on open telemetry, what's happening with everything in the cloud native space when it comes to observability. Another favorite of mine, which was another colleague of ours, Steve Tech. Hi, Steve.

He's not only a great dancer on the dance floor after perform which he is legendary for that he's known for that

but he's also it was really great because we did a recap actually at early March March 7th it was right after perform

perform last year or this year depending on how you see it 2022 was all about game changers and with a recap on things we learned so

it was a pretty good one in case you want but it also reminds me that Perform is coming up again coming up again coming up again

yeah yeah it's going to be a very interesting Perform I think we know it's going to be interesting we know it's going to be interesting I was trying to be coy I was trying to be a very interesting perform, I think. We know it's going to be interesting.

We know it's going to be interesting. I was trying to be coy. What I can tell you is that we have and I got a... It's actually strange that I don't know him, but Tony Hawk is one of our keynote speakers.

That's right. He's pro skater extraordinaire. He's done like a loop-de-loop on a skateboard. I think he was the first one to do a 720 or might have even done more in terms of expense. He was the first rock star skateboarder. There were famous skateboarder people when it was still underground and all,

and then he just broke it open. So pretty cool that he's going to be there. He's been in some of those jackass movies too. when it was still underground and all and then he just broke it open so pretty cool that he's gonna be there he's been in he's been to some of those jackass movies too I think if you ever seen any of those but anyway there's a couple other cool people coming right Kelsey's coming back that's right that's right I saw that I was like man twice now and I'm not not there yeah and yeah I hope to get the, I mean if everything goes according to plan, I

might be with him on stage again. We'll see what we do this year. This year with a real-life audience at the Cosmopolitan in Vegas. Yeah, exactly. And then, I mean, you know, we don't want

to spoil too many secrets here, but it's definitely going to be an epic performance. I think it's going to be a magical moment for Dynatrace and for our community. So in case you listen to this and you're wondering what are these guys talking about, perform.dynatrace.com. It's not too late yet to book a ticket to Vegas and see us live. Or if you can't make it to vegas we do it hybrid that means you can also join at least the keynotes i think a hybrid and then uh what else do we have we have hands-on training days monday and tuesday so if you go on site you get the latest and greatest hands-on material uh to master your skills in observability and then we have the breakouts obviously and the expo area and everything else brian the only sad thing is i think you are not going to be there this year correct correct yeah we have uh yeah i mean as a lot of people may or may not know i've got a uh yeah mentally challenged daughter special needs and all. And my wife's already sending me on some trip like right before. And it's hard for either one of us to be a single parent for too long on that.

So I just can't go back to back. But in a way, it might be good because there certainly is a lot of spikes going on with flu and everything else. I'm like, well, it's a good excuse. But yeah, hopefully we'll be back in full force next year um i was just even just in my head when i said back in full force i was thinking we need to get a light up disco floor in front of the pure performance set up next year and we can have like people dancing the whole time in front you know why not yeah but the good news is right i mean as i said this is henry said yeah

exactly as sad as it is that you are not there uh we will have henrik henrik rex said with isn't observable there on the floor yeah making sure that uh people that are in vegas get the latest on everything that happens in observability space and mark tomlinson has also agreed to you know support him as good as he can because he will be, he will be there as well. Mark Tomlinson, our dear friend.

First friend of the show. First friend of the show, yeah. And friend. And friend. Yeah. Yeah, it's been a crazy year.

A lot of fantastic guests, right? Yeah. I mean, if I, and I'm just looking at the list here, there's a couple of things

that what strikes me, Brian, is that when we started the whole thing years ago, it was all about performance engineering. And we talked about, I think the first couple of sessions were top performance problems in Java, top performance problems in.NET, how to do performance testing and performance engineering. And if I look at things now, we talk a lot about site reliability engineering. I think we definitely covered this topic a lot. We had topics around value stream analysis. We had topics around serverless. We obviously covered Grail, right? I know which Grail was something we covered,

but then, and this was really nice the last time, in the last episode, we really breached into kind of digital marketing and business analytics with my friend Bernhardt. Rightt right right which i thought was him so we really spread out because you know markets are changing and the i think the way what our listeners hopefully also want to hear you know they want to go with the with the flow and they want to learn something new. And performance engineering has also changed, and our tech stack has changed, and the people that talk about these technologies,

there's also new people all the time. And that's also great that we have a lot of great guests on here. We also had one of my favorite highlights was getting the custom-drawn picture from Steven Townsend. Oh, yeah. That was awesome. So if you don't know what I'm talking about, if you look for the episode, the SLO Dilemma from August 1st, 2022, Steven Townsend.

He's in New Zealand or is he in Australia? He's in New Zealand. So he drew a fantastic MS Paint picture of our picture. Definitely a... I mean, there's a lot of great stuff for these episodes. I was just looking at the pictures and I saw that. I was like, oh, that's right. That was awesome. So I don't know if you're listening, Stephen,

but if you are, good on you, mate. Oh, that's Australian, though. Sorry about that. Close enough. But what I want to say, right, if you're new to the show and if you want to, obviously today's a special show. We'll just do a little recap

and give a little outlook on what's coming. But if you are interested in anything related to performance engineering observability site reliability engineering devops we have all these topics covered in episodes and most of the speakers right they are really practitioners so they are working for organizations in in their respective jobs we had actually one of the cool episodes we had and so many cool but um with diana nida she is the sre lead at a very large fintech in canada right and she talked about how she became an sre in fintech yeah and and that was that was awesome right and like how they are what are the best practices on cyber liability engineering

at a fintech organization, at a financial institution. So that was really good. And you know, the last one I want to point out is besides going that how did I become an SRE at fintech,

we also had that great episode with Philip Krenn about why is it always DNS, TLS, right? It was just like so, it was like such back down to, if you talk about like the Java performance problems and all that, it just like came back right down

to all those root problems are still there, no matter how complicated the stack, no matter what new problems come in, the old ones still haven't gone away. Yeah, that was the episode. Just have it here. Why is it always DNS, TLS, so bad configuration changes

that bring down systems yeah big shout out to Philip Grant from Elastic and maybe

to kind of end it at least when I look at it the second to last of this year December 5th was interesting

because it was about SRE for non-unicorns aka enterprises with James Brookbank, where we talk a lot about site reliability engineering and we always reference, obviously, how Google has been so great in doing all of this. But then a lot of people say, well, this is great if I'm getting very excited.

But then the excitement stops short because it's hard to implement these in non-Google-based environments, maybe. And I think James gave a couple of great pieces of advice because he, before he joined Google, he was actually working at HSNBC, I believe, another large

fintech and implementing and it was really great. And he also wrote a book, it's called Site Reliability Engineering for the Enterprise which is free you can download it

on the web it was a great great episode bringing it bringing these principles to you and not just

saying it's just like these lofty things that Facebook and Google are doing but you can do it too yeah

that's always been the issue, right? It's the whole unicorn syndrome. But this is the same thing that goes back to the, what was it? The one second page load, the Google effect. And it's like, no, that's Google though. And that's what was awesome about this episode. Because it's like, yeah, you're not a unicorn. You're not one of these things.

Like, let's talk about the realistic components of it, just like the realistic speeds for realistic, you know, real websites. So it's always great to bring it back down to earth. I want to say thank you to another group of people that I've interacted a lot with this year. And it's not that known probably, or maybe not to the peer performance listeners, but I hope to get some of them on performance. So in the last year, I kicked off a group. We call it the,

the Dynatrace Cloud Automation Guild. It's basically a, it's by invitation only, but the group is growing. It's a group that meets every other week and we are sharing best practices on how to implement observability, how to implement

DevOps, how to implement SRE, security, all these concepts, how to do it right. Meaning, how is a bank,

how is an e-commerce store, how is a government organization, how are they basically implementing and integrating in our case you know mainly dynamic ways but in general observability into their end-to-end the processes and i want to say

thank you to everyone on the guild that has presented because we had just like we have a podcast every other week we have a guild session every other week and we always have presenters and they show how to do things back to the guild so thank you for everybody who presented and i also know we have a lot of topics coming up in the next year we already have some speakers for the first for the first two months at least of the year and and that's great and i'm very grateful because i learn a lot but more importantly everybody else in the guild learns a lot. And if you hear this, and if you're a Dynatrace customer,

because this is, it's for Dynatrace customers only because we are sharing a lot of internals. If you're a Dynatrace customer and if you want to become part of the guild, if you want to learn, but also share on how you do things right, then reach out to me.

My name is Andy Grabner. You're pretty sure you'll figure out a then reach out to me. My name is Andy Grabner. You're pretty sure you'll figure out a way how to contact me. I have one last thing to do as well. I want to thank all of our listeners. Because obviously we couldn't be doing this without you. As we've said multiple times on the show, I think Andy and I both agree that one of the main reason we love doing this is we get to just talk to all these amazing people, learn so much from everybody, and that gives us then the chance to give back to the community by sharing with all of our listeners, and hopefully you're

all finding it as fascinating as we do. But we've been doing this for quite a lot of years now, Andy, and obviously if no one was listening, they'd probably pull the plug on us. But they haven't because you're still there. So thank you so much to everyone. We hope you continue to enjoy it and hope it brings some value. And, you know, our comedic endeavors bring some entertainment to your life. We try to do a best yeah hey but um uh looking at the year 53

of whatever strange strange yeah epoch 53 so i would say p.e post that buck there you go 53 p.e p.e so i'm what i'm looking i think a couple of topics that I want to get covered. And folks, if you listen to this, let me know if you are interested in the topic, if youical side has forced a lot of organizations to rethink a on their spending in terms of energy carbon footprint i mean carbon footprint has always been a big topic but i think now it's more obvious than ever at least here in europe that uh that electricity prices go through the roof and we need to just make sure we are operating more efficiently also in that respect so i want to i want to really like to talk about finops uh i want to make sure that we are covering uh aspects of uh of operating uh financially aware and and just covering this what what can people do to to optimize their or to cut down their spendings by still, while still, you know, delivering great digital services.

That's a big topic for me next year. Yeah. I am also really curious on, I think Kubernetes will be there

for a little while longer, but I'm also wondering what's after Kubernetes because eventually Kubernetes will be replaced by something

else. Is there anything else we should talk about? Yeah, I mean, I don't know if it will be replaced, but I had some colleagues who went to AWS reInvent, and they were pushing

Lambda like crazy. I mean, obviously, just like any phone company, we have a new pricing plan. We want to push it on you because we'll make more money. So I don't know how much of that was AWS's attempt to try to swing more people there without them thinking or if this is going to be the year of serverless or not. But I do, to your other point though,

not going to serverless, is there something we aren't quite aware of yet? That's on the rise. So yeah, it's interesting. And the other thing that I've, last week after recording, so that was mid of December, I was invited to be part of an interesting webinar and it was called, Is DevOps Dead?

And it was like a little battle between two sides and I was on the side of DevOps is not dead and then the other side, they said, you know, DevOps is dead and there's something else. And I got to say, right, while I was defending DevOps, because I think the principles of DevOps, which means that you want to automate and optimize as much as you can in order to get value out to your customers as fast as possible, that you're breaking down silos that exist that are holding you back or that are causing friction. I think these basic principles of DevOps still exist.

On the other side, DevOps has been around for 15 or so years, and maybe it is time to especially educate people on a new model that kind of expands DevOps to the modern cloud native world. And the term that I hear a lot and also kind of my quote-unquote opponents used in that webinar was platform engineering, platform ops. So really, you know, building platforms that enable engineering teams

to get value faster out to production. In the end, it's the basic core concepts of DevOps, right? Automation, automation, automation, observability, and breaking down barriers. But as people are coming in into the market fresh and they look at maybe some of the material that was out there or that was discussed five to ten years ago, I mean, the world has changed

and the tech stacks have changed and the organizations have changed so maybe it's also time while the basic principles are still the same maybe we need to talk about something new maybe it is platform ops yeah it's it's it's like a name change it feels like right like when we went from apm to observability right and a lot of people like to say, oh, APM was just looking at the code, but all the APM tools out there were doing some level of logs, doing infrastructure, doing this stuff. And it's like, okay, now it's observability. And it's advanced, though.

It's talking about a lot more of this. Okay, name change is good. And I think you have a good point with this, though. It's not like the core concept has changed. But it's just, as I think you said, everything that was written and focused around was a lot more specific to the DevOps. I mean, even during the DevOps cycle,

we started talking about DevSecOps, DevSecBizOps, started pulling in all these kind of things. And now it's just like, okay, platform definitely could make more sense. It'd be interesting to see where that goes. Some of this, though, kind of things. And now it's just like, okay, platform, definitely could make more sense. It'd be interesting to see where that goes. Some of this, though, kind of feels to me like, oh, is it rock and roll? Is it country rock? Is it punk rock?

Is it industrial music? Is it goth, right? It's all rock and roll. I mean, it's all DevOps, really. Or maybe DevOps was a rebranding of something else. I don't know. I think DevOps might have been similar to the advent of rock and roll. Because before that, it was really just this waterfall.

Like, you know, the waterfall was your jazz, right? And DevOps became your rock and roll. Yeah. And now it's just all these sub-genres and sub-genres. Oh, if we take into account this and take into account that. If you listen to punk rock, let's say the Sex Pistols. When that came out, and I'm not going to take a huge sidetrack here, but when that came out, it sounded groundbreaking.

But you listen to it now, I'm like, yeah, it sounds like rock and roll. It doesn't sound like anything special. It's guitar, bass, drums, and it's just rock, right? And it feels like same things going on with this. But again, there's new ideas, new concepts. If you're going to get people to think of it the right way, if you're using the word DevOps, they might be thinking, well, I'm not really a developer. I'm looking at the cloud platform. Yeah, go ahead. Give give it a name what's in a name well the names are important right because well no i mean but it's in the end though it's it's

serving the same purpose yeah it does yeah but still it's especially as people are coming in new we need to make sure that we everything we learned over the last year we don't want to confuse people and i think that's also something and maybe another thing I'm looking forward to in 2023 we've gone on several journeys from our perspective from Dynatrace on open source, whether it is

that we are heavily involved in open telemetry, that's one thing obviously, we have launched open feature which sees a lot of involved in open telemetry. That's one thing, obviously. We have launched Open Feature, which sees... How's that coming along? It's amazing. We have I think it's 23 different companies.

I'm not... Don't call me on the number. You need to probably look at it. But we have a lot of different companies that are now actively and two vendors that have integrations now with OpenFeature. So feature flagging vendors that are using that as a standard. We have eBay as one of the, I think, companies with James is his name,

kind of promoting it and using it internally and really showing the value of OpenFeature because it basically unifies and provides a standard for implementing feature flags in um in your source code because a lot of organizations just as we at dynatrace we have multiple different homegrown open feature flagging frameworks and in one application it's an if statement and the other one it's a custom developed library and the next one it's maybe a commercial library and Open Feature really aims to standardize this. It's similar to what

Open Observability is doing or Open Telemetry is doing in standardizing the way we collect observability data. Open Feature is standardizing the way we are we're thinking about and implementing feature flags.

I think that's and so that's a big thing. So I talked about this, but I'm also, from a naming perspective, captain, right? We obviously have... Captain has come a long way.

We're now an incubator project, but we also learned a lot over the last three and a half years. And we just launched the Captain Lifecycle Toolkit, which brings all of the lessons learned brand new to the cloud native world because the first iteration of captain we did not know everything we know now and also the world was different githubs was not that established

yet as a as kind of the effect of standard for delivery and this is why kind of our next incarnation or iteration of Captain with the lifecycle toolkit is more cloud native, has GitOps at the center. And based on the first feedback we've received, it could be a really cool story in 2023 because we provide observability and orchestration around your GitOps practices on Kubernetes. Because GitOps itself is mainly focusing on deployment, on delivery.

What's missing, however, is the whole lifecycle management of your deployments that are done with GitOps pre- and post-deployment checks and giving you visibility and observability into your deployments. And that's what we're doing now with the Kepton Lifecycle Toolkit. So check this out, and I'm pretty sure we'll have we will have some links links and also episodes in in 2023 oh yeah i'm sure and it wouldn't be pure performance without some captain episodes right exactly yeah exactly yeah it's been a while since we talked about it too, so definitely time to bring it back. Yeah. And what else do we, will Twitter still be around? We don't know.

I don't want to open up Pandora's box, but it's going to be interesting to see how all of this unfolds. Yeah. Yeah. Yeah, yeah, yeah. And I was just, I'm just looking at actually a tweet from Kelsey because I think, here we go. This was an interesting one. And he actually,

he retweeted it and he made this tweet on January 31st in 2017. Almost five years ago, six years ago. Wow. He said, at some point I made the decision to focus on foundational concepts, not features of a particular implementation. This is when my tech career took off. And that's actually interesting, right? At some point, I made the decision to focus on foundational concepts, not features of a particular implementation and my tech career took off. And I thought that's an interesting statement to make.

That's pretty deep too. It kind of makes you have to think about that one a lot, you know. And it comes back to what you said earlier. Isn't it funny that we talked about DNS and TLS? It comes down to the fundamental concepts,

because if you don't understand the fundamental concepts, it doesn't matter how they're implemented. You need to understand the fundamentals. You need to understand certain architectural patterns, which ones make sense to adopt and which ones don't make sense for a certain problem that you want to solve. And then the implementation itself, I guess, shouldn't matter.

Yeah, and I think it also goes into some of the theory practice as well. Like conceptually, you know, what is it that you want to accomplish at the end? Same thing that we talk about with observability too, right? Like our SRE is what is the end goal? What is the concept, the foundational thing we're trying to achieve here? And you can fill the rest in from there. So I think there's really two angles from that. What are the foundations of the technology constraints, the network, all these other kind of things that you have to work with?

And then also, what is it that you want to achieve in the end without all the minutia of the detail? And then, just like I think you said, or maybe someone else, I forget if it was you or someone else, if you define your SLOs based on your end user, your first level SLOs based on your end user, sub-user facing ones will start filling themselves in. Yeah. Right?

There was James from Google, actually. Right. Okay. So, I don't know, that's at least the way I'm interpreting some of what Kelsey was just saying. Just from like three seconds ago, because this was one you hadn't even mentioned to me three seconds ago. I have no concept of time. I babble. Anyway. Hey, so 2023, we're looking forward to recording a lot of sessions. I know we have a lot of our upcoming sessions already scheduled. Just to give people a little bit of a heads up. We will have learning from incidents. It's going to be an interesting session with Laura Nolan.

We will have a session that's going to be interesting. And it kind of comes back to what you said earlier, that AWS reInvent, they talked about serverless a lot. We'll talk about observability of the mainframe in terms of

cloud-native, because for me still, it's still out there. No, not only that, but it feels, as much as I know

about the mainframe, it's all like high efficient functions that are executed and you also pay for them. And this is the same with serverless, right? And then we have an interesting one with Nico Meisenthal.

By the way, the mainframe session is going to be with Christian Schramm, one of our colleagues. And then we have one with Nico Meisenthal on hijacking a Kubernetes cluster. Nico has been touring the cloud-native space, the cloud-native communities over the last couple of months. And one of his favorite presentations or most highly rated sessions was hijacking a Kubernetes cluster.

What does hijacking mean? Is that like a hockey reference? How you can hijack. Oh, hijack. Okay, okay. You were doing the J as a little bit of a hard. check me like hockey is like a hockey reference? How you can high check. So basically, okay, okay. You're doing the J as a little bit of a hard like, yeah, I get it now. It might be because I had a glass of beer. And I think it's just the it's the accent. I thought you were saying high checking. And I was thinking like, is

that some kind of like hockey reference where like you check someone up high, like, I'm gonna take a destroyed Kubernetes and get in the penalty box. No, hijacking. Hijacking the Kubernetes ecosystem is what it was called, you said? Hijacking a Kubernetes cluster. So basically, you should check out some of his recordings. He's doing live demos on how he kind of finds security holes into Kubernetes clusters and then from that point on takes

over one component after the other in a Kubernetes cluster until he has full control over it and then he can use it for whatever he wants, Bitcoin mining or Turn it into a Doom server like playing Doom online. Or instance. I don't know, yeah. Oh, speaking of which not to give, I'm not going to mention the details

but did you see that recording of the... There was a... For some of the stuff that's coming up in our world, somebody created an app and it had like Doom in it. They were playing Doom on that thing. I didn't see that, no, no. Yeah, anyhow, sidetracked there, but I'll I'll send you the link. People will know what we're

talking about, you know, we're basically making DOOM 5 as Dynatrace, that's gonna be called DOOM 5. You're gonna be able to go around and you're gonna be able to find your performance problems and shoot them. Ah, look at that. Good old American style. Exactly. See I'm not even drinking, this is the danger. Let's get back to the topic. Yeah, that's why. What are you looking forward to in 2023?

Oh, well, thank you for asking. I'm looking forward to, well, I don't know if it's looking forward. Well, first of all, I'm looking forward to, hopefully with all the lingering chaos obviously you you mentioned earlier some of the economic uncertainty all this stuff going on you know globally not just here in any one single country but um you know as we're starting to pull further and further out of the pandemic you know just in general it'll be nice to look forward to maybe things returning to whatever the new normal is going to be but with regard more specifically you know one of my pet peeves that has come up a lot a lot lately is i'm really hoping in 2023 that people will start considering more than the architecture they want to build when picking what they're going

to use. And I brought this up to you a little bit earlier, and you had a fantastic phrase you coined out of it was, you know, choosing the ecosystem over the ego system. So what this really boils down to is the pattern that I see so often now is that people will pick some fancy setup. We're not going to just do Kubernetes. We're going to do some obscure bespoke Kubernetes. And now I'm going to make things up here. Like we're going to run serverless functions in Kubernetes, right? Whatever that, you know, just the most absurd things.

They develop everything, they get it all set up and running, and then they go, oh, now we have to consider observability. Now we have to consider security. Now we have to consider all the other things, the ecosystem that has to live around this code. And guess what? Because of these choices we made, we're up shit's creek. Everything's going to be really difficult. Maybe it's going to be all open telemetry. Nothing wrong with open

telemetry, but as soon as people start looking at that, they're like, oh, wow, that's a lot of work. We just want an easy button. Well, maybe you should have thought of that before, you know, to sound like a parent, maybe you should have thought of that at the beginning, right? So I'm really hoping we see a trend of taking off where people are going to consider the entire ecosystem that is needed to run, maintain, observe, and secure their applications and their code before committing to that. And maybe it turns out you have to take your second favorite pick. Because if the effort to do observability or anything else is going to be just way more than you want, we talked about, I forget who we were talking about.

It might have been, we were talking about serverless with who's a cam. It might have been when we had cam on. Or it could have been the Akamas guys, Steffano where you know when you're doing serverless you don't get to see anything how things are running right you also don't know how necessarily to write things optimally for that system right you don't know how memory is being used in serverless you don't know how so there are different and there are different ways you can write things that if you know what's underneath will take advantage of that platform. So if you can't observe that, if you can't get anything from this, you're just stuck. So anyway, I'm ranting

here, but it's because I see it more and more and more and more. And it also even goes back to the first episode that you mentioned from last year, or one of the last ones which is was the how to fail kubernetes was it or how not to start yeah how not to start with kubernetes right and that's you know let's just move our let's move our monolith to kubernetes right the things you don't do when you go through yeah so it really comes down to education of picking what you're gonna to pick for a very good reason, knowing all the variables, taking them into account, and not making a choice to feed your ego, but to feed the ecosystem.

You know what's funny? What? I have another Kelsey Hightower tweet for you on this. Oh, yeah? He tweeted this yesterday. And it says, the no-stack developer. And no-stack developer

criticizes the stack without understanding the stack. In addition to complaining, no-stack developers engage in full rewrites, but don't finish them before quitting or getting fired, leaving an organization with no-stack. It kind of goes into a similar direction.

Well, this is something that I can't claim this. I think it's because of some of our guests like Kelsey and some of the other people who've been on that I've always been driving the idea that you pick your stack for very good reason. Yeah. Right? And this is just, you know, we picked up on that and then you see it out in the field. So yeah, I mean, of course, Kelsey is right on top of it. I think we even heard some of this from him before. Even the idea of like, you know, keeping everything simple, right? Don't complicate your setup. Get it up and running.

If it's working fine, you're done. Right? Only complicate it if you need to. But everyone's looking at this cool new thing, the cool next, the next the next and if i do this it'll help build my resume right but that goes back to the ego system right um best thing you know as even even going back to when uh i'll make a covet comparison right when people were talking about the masks and and the vaccines and it was like well if that stuff that stuff works, people won't notice because you won't see a bunch of people dying, right?

Or the numbers will dwindle. But you only notice when things are bad. So if you're running, so the equivalent being, if you're running your architecture, you keep it simple, you're not doing anything fancy, but it's running beautifully, no one's really gonna notice because you didn't get a hero moment. It's just like, yeah, everything's everything's fine. Well, everything's fine, because I made all the right choices, you

know, but that doesn't get noticed, which is always the challenge that I get it. But yeah. Anyhow, I just Yeah, I just really hope people consider more and really think things through. As awareness grows around the ecosystem that you need to run your code, my hope is that people will consider and make choices based on that, not just Amazon says, serverless is going to be great. Let's just do all serverless. Another one, here's the crazy one.

I don't know enough about serverless, but I hear people, so Dynatrace style, we have active gates, nice utility. Well, there's been several people who get really upset that we don't have a serverless version of the active gate currently. You got to spin up an EC2 instance. Like, well, we're all serverless.

We don't do EC2 instances. We don't have VPCs, we do none of that. We're, well, we're all serverless. We don't do EC2 instances. We don't have VPCs. We do none of that. We're like, can we get a serverless version? And I'm like, yeah, it doesn't take any effort at all to run an EC2. But at the same time, I'm thinking like, first of all, that's a decision.

Like, oh, we're not going to run anything that's not serverless. Okay, why? Because we want to be all serverless shop. Okay, that's not a reason to not run. But then there's also the idea of if you're running something that's running 24-7 and doing constant work, isn't it going to possibly be cheaper to do an EC2 instance as opposed to serverless? I know that was true in the earlier days of serverless, I don't know if it still is.

But there's just these things that you hit where you're like, there's no logic behind that choice. Yeah, it's almost like a religious thing that we said. I'm only Linux and I hate Windows for whatever reason. I'm only iOS and not Android. And there is, for both Lambdas and for EC2 instances, there is no server, no physical server you ever have to touch. So in the end, it's all serverless. Yeah. you gonna do what are you gonna do yeah yeah hey let's end on a high note yes i want to say i want to also say thank you to you brian because i always try when i post these recordings on linkedin to to say thanks but's always the same thing. I always have to thank you

for being the magician behind the scenes and for making this possible. And I hope, my wish is that we will do the same episode again in 350, 365 days, and we look back to a great

year 53, or whatever you want to call it. But we're looking forward to the next year but we will appreciate that we had another 25-26 episodes with a lot of cool guests. And speaking of cool guests, I think I need to really thank

you because for people who don't know Andy's the face man who gets all the cool guests for us because he's out and about and has all the contacts in the world. So all these cool guests would not be possible without Andy. I wouldn't get to do my mixing on these episodes if you weren't pulling your magic and your connections and getting all these fantastic people on.

So really, really thank you. Because otherwise, everybody, you would just hear Andy and I talking to each other episode after episode. And if you're still with us on this one i think we're a good team that's what's nice yes all right with this i my beer is empty i need to fill it up with my glass is empty i need to fill it up because i have another meeting coming up in eight minutes i need to run downstairs. All right, enjoy. Get another one. Bye-bye. Thank you, everybody.

Happy New Year. Have a great 2023. Or 53. Have a great 53, yeah. Or there's so many calendars. Yeah, anyway. Thanks, everyone. Ciao, ciao. Bye.

Bye.