PurePerformance - Introducing Is it Observable: Observability education fast track with Henrik Rexed

Episode Date: September 27, 2021

If you need to learn how Prometheus, OpenTelemetry, Loki, FluentD, FluentBit .. help you with your observability requirements in the cloud native and non-cloud native space but you don’t have hours or days to dig into the details yourself then you have a new place to go to get educated within 20-30 minutes: Is it Observable is a new educational YouTube channel by Henrik Rexed, Cloud Native Advocate at Dynatrace.In this episode we have Henrik explain the motivation of creating Is it Observable, how to best use the videos and the tutorials on GitHub to educate yourself and gave a glance on upcoming episodes that will also include guests from various tool and platform vendors.Make sure to subscribe to his channel, check out the tutorials on GitHub and give him feedback on twitter (@hrexed) or LinkedInShow LinksIs It Observable YouTube Channel:https://www.youtube.com/channel/UCRiin4u8YZGlVQRZp7qOOawHenrik Rexed on Linkedinhttps://www.linkedin.com/in/hrexed/Is It Observable Github Repohttps://github.com/isItObservableHenrik Rexed on Twitterhttps://twitter.com/Hrexed

It's time for Pure Performance! Get your stopwatches ready, it's episode of Pure Performance. My name is Brian Wilson and as always I have my co-host, the lovely, wonderful and talented Andy Grabner. Andy Grabner, how are you doing today? It's very good, but I'm always surprised of the words you find for me to describe me. I know you're lying and now I actually see you turning red every time you say lovely and intelligent and everything well it's usually like the the nighttime talk show hosts you know it's always uh whenever it's a man coming out they say the talented and whenever a woman comes out they say the lovely which is just kind of

weird so you're lovely and talented look at that and tanned because i just came back from vacation at the time of the recording that i thought it was funny lighting you you look kind of orange you look like our you look like the former guy a little bit okay i just i just happened to be lucky enough to spend two weeks on vacation one week in croatia and one week in the austrian alps and we had 14 days of sunshine nice nice nice i uh i didn't i did not spend time in croatia the last week nor did i go to the austrian alps but i had my little vacation before but you did observe that i look a little funny today i did and i observed like you're i also observed that your camera is a little lower so you're looking down and it gives you a little bit of a maniacal look a little more

angular angled and it's uh it's kind of scaring me. I observe that I'm kind of scared today, and if I had to take a measurement of that scared level, I'd have to figure out what I'd want to use as an indicator for that. But yeah. You also observe that my audio is better today because I'm actually holding my microphone in my hand and not doing strange things that you always blame me for, like making correct choices. For anybody listening, please encourage Andy

to continue with this good audio trend by tweeting at us saying, I loved your audio, Andy. If we get at least one, then I think we can convince Andy to maybe duct tape the microphone to his hand or something for future episodes. Now, the question is, we've just discussed about what we've observed, what you've observed. The question is, what else is observable? Are certain things observable? It's kind of an interesting, strange segue to our topic today and our guest.

Yes, it is. Would you like to introduce our guest? I think I want to introduce him. And I think we will introduce a new funny accent today on our guest. Yes, it is. Would you like to introduce our guest? I think I want to introduce him and I think we will introduce a new funny accent today on the show. Funny meaning in a positive way. Mine's a funny accent. Mine's a funny accent. I know, I know.

But besides you and you with the US accent, me with the Austrian accent, we have Henrik Rexit with his amazing Swedish-French accent on the show. Hi, Henrik. How are you? I have no idea what you're talking about, Mr. Andrade Skramnera. We will have subtitles. We will get subtitles for today's episode. Henrik, it's great to have you on the show. And let's actually dive into it. First of all, when we talked about observability, this is really the topic of the show.

It's actually the topic of your show. It's the topic of your new YouTube channel. Is it observable? Which is phenomenal that you do it because I think you're doing a lot of us a big favor by looking at a lot of technologies. But maybe before I continue, for those people that have escaped your presence on the internet, your presence in the performance engineering space, can you quickly introduce yourself? So my name is Henrik Rechter, as you mentioned.

And I've been working previously in the performance engineering landscape more than 14, 15, 15 plus years. Gosh, that means that, oh gosh. We're all getting old. Oh gosh, I know I'm not talking about my age, but nevermind. And I recently had the chance to, yeah, to be invited to join Linz because I heard that Linz is doing a lot of amazing stuff, beer and

barbecue. So the reason why I said, hey, I have to join just amazing company based in Linz just to enjoy barbecue and beer, that makes a lot of sense for me. There's no other place you can get barbecue and beer. It's illegal in the world unless you're in Linz.

Exactly. Yeah, sounds like it. No, Henrik, I mean, for those people that don't know yet, thanks for you to make the move. I know it's always hard to move after that many years with an employer. You used to work for Neotis, who are now Tricentis, and now joined Dynatrace. Very happy to have you on board. But very, really happy that you are pushing a topic and educating the community around observability. And your YouTube channel is called Is It Observable? And I would really love to give you the chance to talk a little bit about,

you know, why did you start the channel? What are actually things to consider when we talk about observability? What is, you know, what should be observed or shouldn't be observed? And also what are the technologies you have already covered? What technologies are coming up? What will people learn if they watch your show? So,

yeah, I started this, this new YouTube channel because, like you mentioned, I think there's a clear need for education. I think now if you look around us, we have so many new technology coming out, coming in the market, especially when it comes to cloud-native technologies.

And the last few years, everyone was doing it by yourself. But if you look around, if you search quickly on internet, then you can have some just blogs describing something, but you don't have clearly tutorials or content that will help you to get on board or get started to how to collect things. I mean, observability is a pretty large topic because you have several things to consider when you want to properly observe an environment.

So yeah, of course, we think about metrics. That makes sense. We have the logs. Logs could be also considered like a metric as well. And we have the traces and we have the events. There's so many things we can do and then call it to be more efficient and more when it comes to troubleshoot or delivering reliable systems to our customers. So I think it was a pretty hard topic. And I think for all of you listening today, I'm pretty sure that you could get frustrated, especially having new technology say,

hey, could you ingest those traces in this application? And then suddenly you search on internet and you figure out that you're a bit alone there. So don't worry. I don't want you to be alone anymore. I want to help you. I've been also in the park too many years. I've been sleeping outside

and this is why I have a long beard because I just come out from the forests since several months. And now I learned to survive in this wild cloud native forests. And I can give you some tips. But in fact, I had to learn because for me,

from a performance engineering side, I was looking at problems from a different angle. And now I look conservative from another angle. So I think for me, it was also a good excuse to get onboarded,

to get trained and share the things that I'm looking at and I'm learning. So I think that's the objective of Azure Observable. For sure. I think you're actually pretty lucky, right? That you can really do stuff that we would all love to do, but we can't because we don't have the time. But thanks to you, you save us time.

You save us that need to understand certain technologies like Prometheus, like, you know, I'm just looking at your YouTube channel here on the right. You talk about how to collect metrics in Kubernetes. I think this is where you do an amazing job in explaining how metrics in general work in Kubernetes. You give an overview of how Prometheus works. For me, when I watched the first episode,

I think I texted you, I think on WhatsApp, and said, hey, I learned so many things, even though I thought I already know stuff. Then the last episodes, you talk about local collections. You had Loki, Promtail. The latest one is Fluent Bit. Probably by the time this episode airs, there's even more episodes coming up. But it's really great that you take the time and it takes time to dive into these technologies

and into figuring out how to observe it, but what's important to observe and what to do with the data. And then you condense it into like, what is it, 20 minutes, 30 minutes pieces of YouTube tutorials? Yeah, it really depends on the content. And depending on the content, you can have a lot of things to cover. So I was targeting initially a 20 minutes episode, but at the end of the day, it's more likely between 20 and 35 minutes. Because I always add add in the episode, I always have a structure for each episode where I start to first explain

the technology, the architecture. So even if you know the technology, you can skip it and go to the next section. And then I go to how to configure it. And last, which I think is the most valuable stuff for the content, is the tutorial. So I am always creating a new GitHub repository with some reading files. So then if you want to do and play around it at your own pace, at your home, you have normally all the materials to do it by yourself.

And I want to point out to the listeners that Henrik has some lighting gear behind us. So it's not a poorly lit, crummy looking video. It's a semi-professionally done. So it's not going to be, you know, sometimes you see videos out there and you're like, oh my goodness, I can't watch this. But Henrik's putting in the elbow grease and making it look as good as he can, even with his beard. He makes up for it with the lighting. Are you talking about my performance clinics that are typically not well lit because I'm just doing it? No, no, no, no. Seriously, I'm not talking about that.

I think YouTube is fascinating these days. For every 203 things that the internet's destroying our world for, there is one good thing for those 200 negatives, and the ability to learn things on YouTube, whether it's observability, whether it's a lot of the audio stuff in my case, whether it's like, hey, I need to change the oxygen sensor in my car. People put so many great videos out there and then what you notice as you watch more of them is that some are really well made and some

are not so well made but they all teach you but i'm just giving kudos to henrik because i can even see behind them and you'll see in the picture that the lighting gear and and that's always that just always makes it a lot more enjoyable to have good production. Yeah. Hey, and I was... Yeah, sorry. Go ahead. So since I started this YouTube channel, my wife used to have trouble to explain what I'm doing. And then now with all those equipment now, it's even worse.

And my kid is like, hey, my father is playing with audio and video equipment so it sounds quite weird you know just tell your kids you're a youtube star what do they call it yeah you know an influencer yeah a youtube influencer yeah yeah yeah for me it's pretty negative to be an influencer i mean if you look at the influencers that we have around us, especially in the sticker. Well, I also want to give a quick shout out because Brian, you said, right, there's a couple that are doing a pretty good job and we already had Nana from Tech World. She's doing a great one. So shout out to her. Also, I started watching more and more content from

Victor Farcic, who has been putting out a lot of great content on continuous delivery technology, also on Cloud Native. And also, I think a shout out to him because he also looked at some of your stuff, Henrik, and gave you some feedback. I think that's just phenomenal that all of these YouTube stars, whatever you want to call it, that it's not just like I'm doing my stuff and I don't share anything,

but he gave some great feedback as well on what he learned over the years. That's really good. And also just to remind to the attendees today that are listening, if you want to make our videos better and more accurate to your projects and your own personal technical constraints that you can face within your organization, yeah, send me a chat through YouTube or LinkedIn or Twitter. I'm interested to collect the technology that you would like me to cover. Because at the moment, I'm driving the backlog based on well-known technology of the market.

But again, if you think that there are more technologies that make more sense, please let me know and I will try to do my best to cover those new episodes as well. I've got an idea for an episode, Henrik. Can you observe Schrodinger's cat? Observe Henrik. Can you observe Schrodinger's cat? Observe what? Are you familiar with Schrodinger's cat? I don't know. The chaos theory. Is it chaos theory? Yeah, Schrodinger, the idea that you put in quantum physics, you put a cat in a box with a radioactive isotope. And in order to tell if the cat is alive or dead, you'd have to open the box and look. But mathematically, it could be alive,

both alive and dead at the same time and the only way to know is to open it, which then changes the outcome. So the question really is then, is that observable? But because by observing it, you change the outcome. So I think you need to spend an entire, I'll send you, you can read it up because I'm sure you can do a full episode on that

because no, I'm kidding. But a bad attempt at a joke and you never heard of the concept. So it kind of backfired. So I will shut up now and we'll move on. Let's get into the meat of the topic. No, no. So from the show perspective,

I just launched the channel. So I'm with much focus on building relevant content for everyone. But then I wanted to also cover weird topics, more funny topics. But I think I will wait a bit to get followers to start doing crazy stuff

because otherwise people will try to... You won't understand the purpose of the channel if I stop you. Hey, Henrik, you earlier mentioned, I think this is something that we need to repeat as well. You have a GitHub repository or a GitHub work. It's called Is It Observable?

And I already see that you have, you know, six episodes on there. The next ones are PromQL, Kubernetes events. So it's really great that you are, you know, as you said earlier, not only covering the technology in a YouTube fashion where you explain the technology and then walk through what you've learned and what's important and how is it observable, but then you also provide the tutorials to really, that everybody can kind of explain and explore it. So thank you so much for that. That's really great.

One question that I have for you now, for people, especially that are here for this the first time, what is the thing that you, like coming to the first episode or maybe the late, whatever episode you want to pick, what is the thing that you thought,

ah, this is so great that I learned. And I would love that people watch this episode because I think this was an eye-opener for me. I think, I mean, for every episode, I learn something, to be honest, because when you have to build a tutorial and find the right approach, sometimes you are doing something in one well-known approach and then suddenly you discover that you can do it in another way. So especially in Prometheus,

all the notion of those service monitor and that will help you basically to configure properly Prometheus. So I think that was very interesting. But otherwise for the other parts, I mean, logging, I was not an expert in the logging space, to be honest. So I had to jump in that space and I learned quite a lot. All those, the pipeline that you have to design to ingest and transform your log streams before you store it.

That was really interesting, to be honest, because I see a lot of value for a lot of organizations. I mean, we all want to get logs and we all want to attach some context. And sometimes you just read a blog and you understand the concept, but if you don't play around it, then it could be very virtual as a concept.

And last, I think the episode that will come up soon, because it's not published yet, it's about Ingress and Istio, where I knew a few things around it, but I really wanted to dig in Istio because I was really looking for the observability capacity that it offers as well. And I was really, really impressed, to be honest. I never thought about all the things that you can do with Istio. So that was really, for me, a pleasure.

And I had fun building that episode. And I hope that you will also have fun consuming that episode for those who are listening. Are you just covering Istio, or are you covering service meshes in general? Or is there another episode coming up with other service mesh technologies? So because it was the first time,

I first introduced the concept of service mesh, and then I jumped in Istio, because otherwise it's going to be very brutal for users to say, hey, here it is, Joe, and then you introduce all those concepts related to service mesh. So I had to first introduce service mesh before jumping into Istio, for sure. For the other service mesh,

like Linkerd and the others, I can basically continue doing in other episodes. But for the moment, I thought, hey, I want to do observability and start with Kubernetes. In Kubernetes, there are so many different angles and different solutions. I thought that I want to do this series first

and then maybe we can improve it by adding more and more solutions. The next, what I would like to also address is everything related to site reliability engineering. So SLI, the SLO, how to get started. Because this is what we, I mean, if I want to start and use SLI and SLO, I want to have someone saying,

hey, you want to get started? This is how you should do it. This would be the process. So this is what I'm trying to build. And same thing, there's so many things like chaos engineering, and there's so many products out there. And I usually take advantage of each individual episode to pick a solution and deep dive on the solution,

or two maybe, depending on the episodes. And this is really interesting. So sometimes you can see that the solution makes sense. Some others are more experimental. But at least it's a way of seeing it lively. I like to read blogs, but sometimes when you look at the blog and say, okay, it's cool.

The picture, I love it. The screenshots of the overall solution, I love it. But how do you get to that point? And this is usually the question you can ask yourself when you read an awesome blog on a new solution. Hey, I wanted to ask you about the SLO one. It's funny because just before the recording of this podcast, I was on a call with a prospect

and they were interested in the SLO topic. And I doubt your video will come out in time, but they said to me specifically, do you also have any guidance for best practices? Because they were new to the slo thing they've heard about them they've looked into them but to your point with why this is such a great topic or all these topics are great is because yes we can understand what an slo is and tell you you know what to calculate or if you need to but where do you start right if you're just getting into this

idea um you know what are best practices for starting out with SLOs? And I think I'm assuming, or I guess I should ask, do your episodes cover things like best practices for getting started? Start with these. And I think you mentioned it already early on. Here are the things that you should care about so in in the concept of something like an slo episode these are the kind of things that would help a a person know what to start with setting up correct yeah correct because i i think that uh there are lots of content today uh explaining what is an sli and explaining what is an slo so i think providing another video explaining what it is, I mean,

does it make value? So I thought that bringing the, covering the angle on how to get started and what would be the recommended process to do so, I think makes more sense for our community. And what I'm trying to do in the next coming episodes, I mean, if possible, of course, is to have guests to the show. So for the SLO, SLI episodes, I will have special guests that knows very well the SLI and SLO. And I think Andy and Brian, you have interviewed, that knows very well the SLI and SLO.

And I think Andy and Brian, you have interviewed, it's Steve McGee from Google. So he would be part of this episode. Yeah, it's great. We had him on a podcast a couple of weeks or months ago. Yeah. I think that's an important point you make too, that oftentimes there are things out there that explain what something is but nothing to tell you what how to get started or what to pick if you even take a if you think about not even

from an slo point of view but monitoring your end user and this is something that's been a topic a hot topic in web performance forever is what metric do we use do we use page load time do we use? Do we use page load time? Do we use visually complete kind of thing or time to first paint? And having a discussion around how can an organization decide what to pick there? Because web performance, it all depends on what you're doing, what's important to them.

These are the things that are not out there. There are videos, there are explanations about what these metrics are, what they tell you, but how to have that conversation or how to have the thought process around what's going to be best for our organization and most meaningful for us is something that's often forgotten. And we see it all the time where people ask, you know, well, what should I do? I'm like, well, I don't know. I'd have to have a room full of people to talk about what's important to the organization and these considerations.

So I think that's a really great angle that you're adding to it. Because everyone's just left assuming they know what to do with the information, which most people don't, I think. Because it's this brave new world of information overload and what do we pick and how do we start? Yeah, that's usually the biggest challenge. In fact, it's interesting because we had a few meetings with the customers on that particular topic of SLI and SLO, and everyone says, hey, could you give me a standard SLI? Can you give me a standard SLO? And I think there's a misunderstanding. There is no standard SLO.

I think you have to know your system, know your users, know your environment, and you will define your SLO. So I think, yeah, it makes sense. I think people understand the concepts and definitions, but when it comes to say, okay, now I'm going to start, and yeah, it could be very difficult. I mean, even in video games,

the first level is like you learn how to just use the remote control. So yeah, think about, is it observable like the first stage of your best video games where I'm going to teach you how to control your character and behave in this cloud-native world? Hey, Henrik, as you are covering more and more technologies, it means you are pulling in more and more data,

whether it's logs, metrics, traces, events. Have you also, when you walk through these technologies, do you also cover what happens with this data? Where does the data end up? What do you do with the data? Because just observing data is great, right? It's the first piece of getting the data in and observing it. But what do we do with the data? Where does it get stored? Do you cover this as well?

So I explained the way the storage has been done, if you take the logging, for example. And then having ingested some metrics or some logs or some traces is great. But then understanding what you can achieve with this, it's another piece. For the moment, I try to build graphs and explain, I mean, for the things you should pay attention in communities, for example, but for the others to be completely transparent is depending on the topic, which it could be very big. You don't have the ability to cover things you should consider. Those metrics are really important.

In the case of Istio, for example, you should do this and this and this, and that would be the best dashboard that you can do to be able to utilize the data that you have ingested. But that makes sense. I could definitely do a series for each new episode where I say, now that we have ingested the things, this would be how you should consume it in a smart way. And also, as you said, even though I think it's probably very hard to say what is a good threshold, what is a good,

I don't know, behavior, because it always depends on the application but certain things are probably like for instance right if you are coming back to problem patterns that we've seen over the years making too many round trips to the database and if you can measure this and this is something you can say this is not a good thing or if you're running constantly in garbage collection issues and you can measure it and observe it through certain, let's say, metrics, then, so I assume for Istio, there are some metrics that are, let's say, indicators of bad behavior or good behavior.

So I think that would be great if you can at some point cover this as well. Yeah, at that moment, I was very focused on how to connect things together and what would be the areas where you should bring your attention because there are problems related to those components and those components need to be properly observed. And then there are a few open source solutions out there,

which are amazing. So for me, it also was a great excuse to say, hey, can I connect this solution with Tenedris? And so then I started this episode to say, let's figure out if it's possible or not. So for me, it's an excellent learning curve, to be honest. And I think also, I try to bring the angle of making things connected together and take the best out of all those

solutions than to have all the metrics in one place so you can have a real observability of everything. This reminds me of the old question if a tree falls in the woods, does anybody hear it? If a process falls over, does anybody know? Or does anybody feel it? You can collect the data but are you

doing something with it? Is it bubbling up? Are you aware of it? Which is the second half of it. Observability goes hand in hand with, as Andy said, how is it being stored? How is it being processed? Is it dashboards? Is it being analyzed through some automated process or not? Whatever that is. But I hadn't thought about the concept of the tree in the woods thing, but it's a fun debate. Hey, Henrik, now I'm not exactly sure when this airs,

but I know you are pushing out new episodes every other week. I think Thursday is an observable day, correct? So it's not, yeah. I try to find a way of attracting people to the channel. So that was my communication of last Thursday. So it's every two weeks, I deliver new episodes. So every 15 days. And the delivery date will be Thursday.

Okay. So every other Thursday is an observable day. Observability Thursdays. Observability Thursdays. Yeah. That's nice. And what's, you know, looking beyond now, at the time of the recording, it's August, I assume this will air sometime

maybe in September. What other technologies besides the ones you've already mentioned do you have in mind? What's coming up? So there's lots, lots of soft technologies that will come up in the next coming weeks. So I think

because I covered Prometheus and we know that if you want to consume, it's great to collect things that you mentioned, but if you want to consume the data to create alerts

or with the alert manager of Prometheus or Grafana to make dashboards, you need to understand a bit PromQL. So that's why I decided to do an episode dedicated to PromQL. And so this one will be in the end of the month. And then I have now currently a new episode coming in with the communities events. There is so much great information that you can gather from only the events

because every community's object is going through various states and there's a lot of events coming in and coming out. And I think for me, it was very interesting to learn. So I would say it's a bit the second angle of the first episode on communities. So I remind you concepts related to communities in that communities events. Otherwise, open telemetry is a very, very hot topic.

So I indirectly covered it with the events because in the events, I built very interesting tutorials to connect K-span from Weaveworks that transformed the event flow into traces. So I connected K-span with that. So that was a way of introducing the OpenTalent Tree Collector. But I'm planning to have a series of episodes on open telemetry. So first, what it is, then going more on the practitioner aspects of how to instrument,

what are the things you should instrument or not on your code. And of course, the other angle is how once you have ingested those traces, how you should consume those traces. And again, I have tons of other technology more or less related to the cloud. So Google Cloud Functions, Google Cloud Run, all the various technology concepts that are available in AWS as well. So there is a pretty large backlog as of now. But yeah, normally in September, you will have the events and the SRE,

first episode of an SRE and open telemetry available for the community. Just because it's fresh on my mind, open telemetry, you should talk with Bernd Faker. He's one of our colleagues because obviously we've just talked about technology typical in the cloud native space and why all of this came out in cloud native. Bernd is currently working on instrumenting Jenkins pipelines with OpenTelemetry. They are contributing back to some open plugins that are basically creating spans and traces every time you're executing a Jenkins pipeline.

So they actually see now through open telemetry where pipelines spend their time. Is it waiting for an agent? Is it checking out source code? Is it building which task? And so just as a reminder for me today, when we talk about open telemetry, we talk about cloud-native technology, but some of these things you're covering is not just purely focused on

Kubernetes, right? No, no, no, no, no. Clearly not. For example, I will have an episode with some DevRel from Google on 4Keys. And in fact, what you just mentioned gets directly in the Google 4Keys topic, where you not only monitor your applications, but you also monitor the tools that you have around you. So Jenkins is definitely a way of, you can collect some metrics and some traces

to figure out how efficient you are in your releases, your pipelines and so on. So I think there's a lot of great information and things that we can collect to improve our process, internal process. It's incredible that people are now going to be monitoring Jenkins and the pipeline tools because Andy, like way back when we started talking about so much more of the automated pipelines, the concept came in as your pipeline is your new code, right? Because it's getting more and more complex. There's more and more lines of code to write all this. And now people are getting to the point where it's gotten so big, we need to monitor it to find some efficiencies. But not only efficiencies probably in the programming of the code and the way it's being used, but also, I guess, timing certain areas.

You can take a look at how long certain processes are taking to find ways to improve them organizationally. Not saying like, oh, not in the point that Jenkins slowed down slowed down but what's the um what's the term about the uh how long each little phase takes there's that uh i can't i'm blanking on terminology when we look at um when you look at a pipeline uh management solutions no that's whatever i can't think of the term but it's it's interesting because you can now look at your pipeline and see what sections are taking long and then focus on, hey, how do we improve? How do we find efficiencies for this team

so that we can speed that part up? And you're right, because maybe Jenkins is taking long on it because there's something that they're doing. Yeah, or I think what Bernd, what he showed me today, just finding out when builds are kicked off and which environments these

builds run on and where there's a lack of resources so as you know right we internally when we build our agents we build them on all sorts of different technologies um including solaris i mean all different environments and if they're all kicked off at the same time they all wait for the same resources and we only have a certain pool of resources so that means this information now allows them to better schedule their pipelines because they can then schedule them so that individual pipelines don't have to wait too long or you know at certain let's say times during the day maybe developers some are tending to do more pull requests and maybe at this particular point in time we need to scale up the Jenkins agents

to just have more resources available. So it's interesting, exactly. It's very interesting what we can do with this technology that kind of emerged out of cloud-native, but now also benefits the non-cloud-native, the more traditional world we used to live in. So when we start observing the observability tools, what do we call that?

The watchman? It is a good one, yeah. They had a customer the other day saying, hey,

do you expose your metrics in Prometheus format? Do you expose the way you collect your metrics in

Prometheus format? They say, that's interesting. But yeah, basically they are asking, can I monitor Dynatrace? Or can I observe Dynatrace?

Well, we do self-monitoring, so we definitely have a lot of metrics that we give to our customers, and we also use this internally. So I think next time, hopefully you answered this, you can say, yes, we got the metrics. We have a lot of self-monitoring. The question is what they're interested in. Henrik, so is it observable, the YouTube channel? We definitely put links to the summary of the podcast.

There's the GitHub, is it Observable organization with a couple of repositories for each episode. Is there anything else? Do you have a Twitter account already? Do you have a, I don't know, I think a domain? I heard something about that you're trying to get a domain. What else is there?

How can people follow you? So at the moment, I didn't create a dedicated Twitter account for Easy to Observable. So you could basically reach out to me on at H Rexed. That's where my Twitter account. Or you can reach out to me also on LinkedIn,

where you also have my profile. So if you find a weird guy with a Lego face, don't worry, that's myself. So don't be shy. It's not a fake account. It's a real person behind this plastic. There is a real heart beating inside. But otherwise, I'm planning to open a website.

So it will be easier for those who follow me to search and filter for for various episodes and uh and also i'm planning to probably do a forum where you could suggest your topics uh so uh maybe uh we we started with the first promotion about Mission Impossible. So maybe the next future episode, the followers could post and record their voices. I can use it too for the intro. That will be fun.

But I will offer your idea. Technically, if everything is possible, once I start to build the website. So I think people can also reach you if on the third full moon of the year, if they go to a clearing in the woods and shout your by opening it, I will hear it from long distance and I will join you. You just got nothing to add. No, I got nothing to add. I just know now that I need to get some Belgian beer in September because I know Henrik is going to visit the place, the only place in the world where they have beer and barbecue, as he said in the beginning, Linz, Austria. So I need to get some Belgian beer.

No, I got nothing else to say. I just want to say thanks again, Henrik, for doing this. You do us all a big favor because not every one of us has the time to look into these technologies and you're making it very easily consumable i got already some great feedback from people um that i've kind of you know sent the link to and they said yeah this is great you have um you have a nice set of humor the way you do your videos and it's just educational and yeah that's all i can say. Thank you so much. You're welcome. You will thank you with a beer.

Okay. Okay. Any closing thoughts, Henrik? I wouldn't say. Do you think we would be able to observe the quality of the videos that is produced? So how can I observe myself doing a good job? That would be a question for all of you. So let's bring observability in. Is it observable?

See, the DevOps practices, which is really from the Toyota factory practices, those can be applied everywhere. Observability can be applied everywhere. It doesn't have to just be metrics. You know, are you what was that a dog or a cat behind you that was walking through? That was the daughter. I know that we all have a lot of hair at home, but it was my daughter. I just saw a black shape. So you can ask her, are you being a good dad? That's observability there.

It depends on if you're giving her candy. I'm too afraid to ask those kind of questions to my kids. I don't know if the source data... Anyhow, yes, observability can be applied everywhere. Thank you, Henrik, for being on and I look really forward to seeing where you take your show.

If anybody has ideas for Henrik, we'll have links in the places to put it on. If you think you might make a good guest for him, reach out to him. If you just want to share a Belgian beer with him, you know, you can do that as well. Same with us. You have any ideas or want to be a guest on the show, reach out to us, pure underscore DT on Twitter, or you can send us an email at pureperformanceatdynatrace.com. And thank you, Henrik. Thank you, Andy, for always making this possible.

And I guess thank you, our listeners. We love you. And stay safe, everybody. We're not out of this yet. So, thanks, everyone. Bye-bye. Thanks. Bye.