DOP 61: How To Use PowerfulSeal To Create Chaos In Your Kubernetes Clusters

Transcript

Darin Pope 0:00
This is episode number 61 of DevOps Paradox with Darin Pope and Viktor Farcic. I am Darin

Viktor Farcic 0:06
and I am Viktor.

Darin Pope 0:07
And today we have a guest that I get to introduce today. This is the first time I've done that. We're not recording video, but Viktor just sort of raised his hands in joy and glee that he doesn't have to do any work today.

Viktor Farcic 0:21
Next episode is yours.

Darin Pope 0:23
Oh, gee, thanks. Today we have. I'm just going to say his first name then I'll let him say his whole name. Today we have Miko. He's the creator of PowerfulSeal. It is a chaos tool. Miko, thanks for joining us today. Go ahead and say your full name. I'm not even going to attempt it.

Mikolaj Pawlikowski 0:43
So my name is Mikołaj Pawlikowski. That's pretty much why everybody calls me Miko.

Darin Pope 0:50
And I'm thankful for that today. That's that's a very good thing. So you're the creator and currently still the primary maintainer of PowerfulSeal.

Mikolaj Pawlikowski 1:01
Yeah, that's correct.

Darin Pope 1:02
Why don't you explain what PowerfulSeal is and does

Mikolaj Pawlikowski 1:08
So, PowerfulSeal is a chaos engineering tool designed for Kubernetes that basically lets you write scenarios, which describe what you want to do to your cluster, and then how to verify that what you think is going to happen is actually going to happen to it. It's a small little tool that's been open source for a few years now. And, yeah, I have the pleasure of maintaining that.

Darin Pope 1:36
Is it really true that you enjoy working on it? Sometimes open source projects can end up being not so fun to work on.

Mikolaj Pawlikowski 1:44
Yeah, I think I've been fortunate enough that there's a few contributors who not only do quality work, but also are actually nice people to work with. So yeah, it's been a pleasant experience.

Darin Pope 1:57
Has your experience been pleasant, Viktor? Nevermind.

Viktor Farcic 2:01
With what? Everything right?

Darin Pope 2:05
Open source projects.

Viktor Farcic 2:07
It is true. It really that there is a bit of everything. Sometimes I started a couple of projects and contributed in a few others. And initially usually it's kind of it's to me it was it was nicest when, in my projects when I was the only contributor because then I do whatever I want. And then it's wonderful when people jump in and you see that kind of people adopting a project and people are contributing, and that's beautiful, right? And then this is my experience, right does not necessarily need to match other people experience. And then if a project grows, let's say hundreds of contributors, then usually starts getting a bit more complicated because, you know, especially if you're the person who started the project, then you might have strong opinions, and then not getting used to strong opinions of others and sometimes complications and clashes, but it's all good usually because you know, you want to fight with people over what is a better way to do stuff or whatever, as long as it doesn't ends up being a silly, silly conversations kind of like, I require that you use tabs instead of spaces for indenting your code and things like that right?

Darin Pope 3:18
tabs and spaces are not a silly thing. That's, that's great religious wars, but that's not why we're here today. Yeah, I sent us down a bad path.

Viktor Farcic 3:28
It's a good discussion because it should be always spaces.

Darin Pope 3:32
Always. Okay, We are ... most of the time we're in agreement, but we're in true agreement there.

Mikolaj Pawlikowski 3:37
You sound like an Emacs person, Darin.

Darin Pope 3:40
Emacs.

Mikolaj Pawlikowski 3:42
Yeah, let's not go there.

Darin Pope 3:44
Okay, let's not go there. So and the other thing that right now, Miko is working on a book that's covering chaos and I'm assuming it's focusing primarily on PowerfulSeal. Is that correct?

Mikolaj Pawlikowski 3:56
No, that's not actually the case. So currently if you're just starting with chaos engineering, there's a bunch Well, a few good books that are available. There is enough theory, I think, to just get you started. There is your book about Kubernetes chaos engineering, that I just started reading yesterday when I learned about it. And then there's a little bit of, I guess, just implementation / administration / tool kind of zone, which I felt wasn't necessarily very well covered. So the book is actually focusing on giving you all the necessary tools, primarily for Linux, on different popular use cases that you might be actually working with when you're implementing your chaos engineering experiments. So it's kind of like instead of going with a particular tool and just teaching you how to use that tool, You know, like there is a book from Russ, who wrote it primarily about ChaosToolkit. I wanted to do something and that kind of goes more to the first principles and gives you all the tools that you might need for rolling out your own. So that's the main focus.

Darin Pope 5:21
Thank you. We'll send you the $5 for the book plug in afterwhile. Now and you brought up Russ, so do you know Russ?

Mikolaj Pawlikowski 5:31
I do. He's a great guy.

Darin Pope 5:35
Oh, boy, that that was that was fairly short. Okay. We're joking, Russ. If you're listening, we're joking.

Viktor Farcic 5:43
I have a quesiton I would like to hear your opinion because I can I give a specific opinion what kind of what do you think is the state of I mean, on I think we all agree that kind of on a theory perspective what is chaos engineering. I think that's well established. But from a tooling perspective, how do you see the state? Generally speaking, not any specific? Right?

Mikolaj Pawlikowski 6:11
Yeah, that's a good question. I think, you know, it's still a bit of a wild west. So, you know, you have the few popular tools that keep popping up. You know, like I mentioned, the ChaosToolkit, everybody seems to know about that. And then you seem to have everybody who knows about Gremlin, they seem to be doing a really good job in communication. If you Google anything remotely to do with chaos engineering, you're likely going to land land on one of their blog posts. And that's great. But I think, you know, one of the like, bigger problems about that is that most people, if you ask them, they're still going to think that it's about you know, randomly smashing things in production. If you Google that, it's probably going to come up with some for you know, breaking things in production on purpose. I think you know, going back to the tools, there's a few of them that are good for the use cases that they cover. There's the awesome chaos engineering list. I'm sure you're aware of that. It has a pretty up to date list of tools. Some of them are dead, some of them are doing pretty good. You know, there is this PowerfulSeal that you should try out. Overall, I think that most of them are still on the kind of stage one, they give you the ability to kind of introduce the failure, and they kind of leave the validation and verification part of it up to you. And I think that's probably where most of the value in the coming years is going to come from, from actually, you know, going past that first stage of introducing the failure which is not that complicated, and producing the value from the automated or half automated verification, and of the SLOs and your assumptions. So that's really what I think.

Viktor Farcic 8:08
That's kind of similar to how I see it. I think that most of the tools are tackling the easy part kind of like how to how to change something, not to say destroy, right? But I'm not even sure where I would start from validation perspective, right? Because, you know, I see it easy kind of okay, let's destroy this pod and see whether the pod is recreated right? And, yes. But kind of theoretically, something completely random somewhere else can be happening as a result of you performing some actions. And I'm just thinking out loud, I wouldn't even know where to start how to validate because ultimately, it's about the effect on the whole system, right? Maybe I'm wrong in in my kind of guess. Right.

Mikolaj Pawlikowski 9:04
Yeah, I think that's true that the real challenge is in kind of extracting that value from knowing how that corresponds to what your expectations are. So most of the tools will be still on the easy part right now I think.

Viktor Farcic 9:22
That's also normal. You always start with easy.

Mikolaj Pawlikowski 9:25
Yeah, definitely

Viktor Farcic 9:27
At least I do.

Mikolaj Pawlikowski 9:29
Yeah.

Darin Pope 9:30
Walk us through. If you're starting with PowerfulSeal, what is the Hello World?

Mikolaj Pawlikowski 9:34
Well, the Hello World would probably be around 10 lines of YAML that would probably pick some random pod. When I say random, probably not random. You'd probably take something like a deployment that's being covered by a service. And then you would select one of the pods you would tell PowerfulSeal to kill it and then you would verify that your service responds to some kind of HTTP probe within a given parameter or parameters. That's probably, that's actually what I've been putting on the on the readme not that long ago as a Hello World. Then you take it from there, like you mentioned, there's a few examples that we can give there. But then at the end of the day, it really depends on the application. Its characteristics, and that's where, where the value comes from.

Darin Pope 10:34
What made you want to write PowerfulSeal? When did it start?

Mikolaj Pawlikowski 10:39
So that's an interesting story. So that would be back in 2016. I ended up working on this new platform, based off Kubernetes and that was fresh off the shelf version 1.2. I'm always going to remember that. We did the evaluation against some other options, we decided that that's probably worth. Well give it a proper shot. And then through some interesting personnel reshuffles, I ended up actually leading that platform. And I started having little nightmares about how to actually verify that something with so many moving parts that I just met. And, you know, starting to use right now is going to be working the way I expect it to. And the answer to that, what we started doing in our team was just automating little experiments. So okay, so if, for example, I restart my kubelet, what happens to the pods that were running on that node, on that node, etc, etc. And for because of the fact that at the time, there wasn't really like a well defined path to setting up all those moving parts, you're getting all these binaries and there are some options of how to set it up. But at the end of the day had to know what you were doing in order to set it up in a way that's resilient. So we started writing more and more of those scripts. And then I turned it into PowerfulSeal that later on we made open source. And the rest is history really.

Darin Pope 12:32
So the project is about three and a half, four years old. Maybe not from the official project, but back to its backstory.

Mikolaj Pawlikowski 12:41
Yeah, that's, that's, that's about that. Yeah.

Viktor Farcic 12:45
That's actually very, very impressive. That's kind of like that would mean when you started basically the very, very early in, in the lifecycle of Kubernetes. Right?

Mikolaj Pawlikowski 12:56
Yeah, I mean, I was I was kind of lucky. You know, it was a kind of greenfield project. And we're able to pick what we needed. We're building microservices platform. So something like Kubernetes was essentially exactly what we needed to make it work. There was another funny coincidence. There was a conference, KubeCon in London that was literally next door. I ended up just seeing it when I was going to work and that's how we ended up going there and meeting some folks and we felt like you know, it had future and it turns out, yes, it did. So that's how it started.

Darin Pope 13:33
So like you said, You started now, correct me if I'm wrong, you said that you started with Kubernetes 1.2.

Mikolaj Pawlikowski 13:42
Yeah, that's the first version that's we're kicking the tires.

Darin Pope 13:46
So in 2016, why did you make the decision for Kubernetes over Mesos? I went there. Sorry. I went there. Because I wouldn't have. Not in 2016.

Mikolaj Pawlikowski 14:04
I think that, you know, one of the reasons was we saw how many people were actually getting invested in that at the time. And we also liked basically the API that it was exposing, you know, the entire idea of the abstractions, you know, the pods, deployments and stuff like that. It felt right. And we did you know, this massive Excel with pros and cons and we just decided to give Kubernetes a shot as the first approach. It played out well.

Darin Pope 14:46
You didn't break out a D20. You broke out an Excel spreadsheet to make the decision. For people who don't know what a D20 is, go Google it. PowerfulSeal...four years old, roughly three, three years, whatever. focus purely on Kubernetes. Is that true?

Viktor Farcic 15:08
It works with VMs as well, right?

Mikolaj Pawlikowski 15:10
Yeah. So it. Yeah, the primary focus is on Kubernetes. But it also has the drivers for various clouds. It has the Google Cloud, AWS, OpenStack. We started with Azure, we started with OpenStack, because that's what we were building on. And the idea was basically to, you know, be able to just take the VMs up and down to simulate going down. So it kind of works on two levels. You work on the Kubernetes abstractions, the Kubernetes type of items, and then you can also do things on the nodes themselves. And also, we can do things over SSH, if you wish so so, you know, one of the reasons why we did that was because we wanted to just not give Kubernetes any headstart when we're deleting a pod. So instead of just asking the kube API server to delete a pod, you can well, you can get PowerfulSeal to issue a SSH command that's going to go on a particular node and then just do docker kill, so that it simulates a crash a little bit better. That was useful at the beginning, when we were really not sure that the entire Kubernetes thing would work out. And we wanted to make sure that it really recovers from this kind of thing.

Darin Pope 16:34
It's 2020. Where do you see Kubernetes in five years, and where do you see PowerfulSeal in that ecosystem in five years?

Mikolaj Pawlikowski 16:46
Okay, so

Darin Pope 16:49
It's easy, right? It's easy. Hey, let me let me make it a lot a lot easier. One year.

Mikolaj Pawlikowski 16:55
One year?.

Darin Pope 16:56
Where do you see it and where do you see it this time next year? Where do you see Kubernetes and where do you see PowerfulSeal?

Mikolaj Pawlikowski 17:02
Honestly, one year from now, I expect that Kubernetes is not going anywhere. If anything, it seems to be still on the upward trajectory. It's defacto becoming the API for scheduling things. Pretty much every cloud provider now has also dedicated Kubernetes offering. I don't know what you think, but it doesn't seem to be going anywhere, anytime soon from what I see. In terms of PowerfulSeal, I think that we've kind of gone through the phase of just a little bit of chaos engineering here and a little bit of chaos engineering there, over the time we've added features. We've done a little bit of an overhaul recently to actually, you know, add the documentation. So that's, that's new, it's shiny, it has buttons and hyperlinks. I think like the biggest missing bit right now, and that's kind of you know, I'm ashamed, ashamed to admit that is that we're using other tools for injecting things like latency and stuff. So it will be nice to actually integrate that directly into PowerfulSeal so that you don't have to hack around to get that. In a year from now, I would like it to cover all the networking stuff and hopefully become a bit more popular too. Do you think Kubernetes is going anywhere in a year's time?

Viktor Farcic 18:27
Oh no. Kubernetes is not going anywhere in 20 years time. And not because not because there will be nothing better. There will be but because the level of investment companies are making, like you mentioned cloud providers, right? And but basically, that's everybody. It's kind of, it's going to be there. We're going to be talking about Kubernetes I don't know 20 years from now 30 years from now as ok like like, modern mainframe, right? Okay, we have cool stuff now, much better than Kubernetes. But we have so much. Exactly.

Darin Pope 19:05
We're still stuck with the stupid Kubernetes thing. Like how we are still stuck with a stupid mainframe thing.

Viktor Farcic 19:10
Exactly. I think that will be my parting words when I retire, I don't know when I'm gonna retire 15 years from now. 20 whatever, right? Oh, finally I can I can actually retire and go on vacation, live on an island and not watch Kubernetes anymore.

Mikolaj Pawlikowski 19:29
Well, that's what you think. Your phone might be running Kubernetes by then. So you might not be able to escape that.

Viktor Farcic 19:37
Exactly, that's true.

Darin Pope 19:41
But if it's an Apple device, you can't have you can't log into it so it won't matter anyway.

Viktor Farcic 19:45
To me, what is interesting about pretty Kubernetes is that the I don't think that we had on that level before the way how, mostly through its API, enables us to do cool things that we couldn't do before, right, and I'm not now referring really to scheduling containers and the whole scheduler story and containers and all this stuff. But you know what we're seeing, like, I mean, to begin with, like with chaos engineering tools, how much easier it is actually to do stuff than it would be before. But that equally applies to the way how we do logging today, how we do metrics collection, and monitoring, and networking. I mean, a lot of things are as I see it as enabler of good things, rather than being amazing itself, and that is really, and that is what actually drives that investment. It's not only about, you know, investment in AWS, investing a lot of money to create EKS, but everybody else investing money to actually build on top of it.

Mikolaj Pawlikowski 20:53
That's very true. And that's a very good point. But there's a dark side to that too, that you know, now when you try to debug something, you need to know how to debug your overlay network, you need to know how logs look like from Calico or Flannel or whatever you're using. It's, you know, the learning curve from the kind of, I want to understand what just went wrong is also pretty steep. So, you know,

Viktor Farcic 21:18
Like a couple of years ago The first time I heard the word CRD, I was like, Yes, this is amazing. Now when I hear when somebody says CRD, I close my ears and play play La la la la la la. I don't want to hear it. I don't want to hear it anymore. I don't know what's going on.

Mikolaj Pawlikowski 21:33
Take your dirty CRD from me.

Darin Pope 21:37
So explain why Viktor CRD is a dirty term to you now.

Viktor Farcic 21:42
I mean, it's amazing and dirty at the same time. It's amazing because the level of what you can accomplish through CRD basically by extending Kubernetes is amazing, right? Like, like Tekton wouldn't be possible without CRDs and Knative would be and Istio and so on as it's not that it wouldn't be possible, it would be much harder. But that level of extension also creates the problems that Miko was mentioning kind of Yes. But now I don't know what's going on. I have a problem. And I don't know where to look anymore. All I know it's not the pod. What is it? I have no idea. It could be networking. It could be one of the 57 CRDs of Istio. Or maybe it's not, I don't know, maybe it's something completely different. Who knows? We're gonna find out one day unless we go bankrupt and then we don't.

Mikolaj Pawlikowski 22:39
Yeah, I laughed pretty hard when I saw that Istio was moving back to well moving back turning into a monolith now. That's kind of like a sign of all of that.

Viktor Farcic 22:50
But, but that's also kind of, you know, to me, that's very I was very surprised when I heard that but not the Not by the news that they're going to monolith but how our perception "what is a monolith" changes. Right? A few years ago, today's version of Istio everybody would be considered more micro service still right? Because it's not really monolith by any standards we had until recently. But yeah, now it's going monolith. It was fragmented too much. It was micro/nano services and now now it's actually a microservice but we cannot call it like that. So monolith.

Darin Pope 23:35
Boy, that went ended up in a really bad place...CRD to Istio monolith, but that's what it is. So Mico's got a book coming out. He's working on it right now. In fact, it's from Manning Publications. It is called Chaos Engineering: Crash Test Your Applications. You can order it on MEAP now, MEAP that is whatever that short acronym, Manning, if you're listening to this I asked for forgiveness now, because I should know it. I bought them before, but you can order it right now the link to the book is in the show notes. But also, But wait, there's more. Also, we have a discount code that you can use to get 40% off. That discount code is podparadox20.

Viktor Farcic 24:24
Okay, look, nobody's gonna remember that, put it in notes, copy and paste. More than three characters is not rememberable anymore.

Darin Pope 24:32
It's okay. That's also in the notes right below where the link to the book is. Finally, if you've gotten this far, hang on a little bit further. So, if you're listening right now via Apple Podcast, please subscribe, leave a rating and review. All of our contact information including Twitter and LinkedIn can be found at https://www.devopsparadox.com/contact and if you'd like like to be notified by email when a new episode is released, you can sign up at https://www.devopsparadox.com/. The signup form is at the top of every page. And there are links to the Slack workspace, the Voxer account and how to leave a review in the description of this episode. Now, if you've listened this far, we have five, count them five codes to where you can get the book for free. Yes, I said, free. This is Miko's book for free. In order to get one of those five codes, you must join the Slack workspace. And there is going to be a link right underneath the show notes where it says use the discount code. There'll be one more link there. follow that link. Go to the podcast channel in the slack workspace. Type in the picture in Viktor's office that's on the bottom right hand corner. We only have five codes. You're laughing. I'm trying to make it up. I'm trying to I'm trying to make people join in the Slack workspace a little bit. Come on,

Viktor Farcic 26:21
I'll let you know later why I'm laughing. It's not what you think.

Darin Pope 26:25
You always laugh at me. It doesn't matter. There will be a picture there'll be a link to the link will say Viktor's picture in fact, you know what, since we're recording this, I better write that down because I will forget it. So Viktor's picture so go look in the show notes, Viktor's picture, click on that type in though. It's a face. So type in the name of the person's face in the bottom right hand corner.

Viktor Farcic 26:53
Okay, you need to write those instructions down. I got lost at the very beginning.

Darin Pope 26:59
That's the thing is they They have to listen to this and understand the understand the thing.

Viktor Farcic 27:05
So it's like my daughter doing homework I always tell her you need to listen to because now it's remote you know they do every remote. You need to you need to listen, listen, listen to it three times. So that you really sure that you understood.

Darin Pope 27:19
So if you still don't understand go back and just hit, hit hit rewind for like, two clicks and that should get you there. So again, Miko's book titled Chaos Engineering, my mouse just died. That's no good. That's chaos right there. my mouse dying actually, it probably didn't die. I just lost it. There it is. That's the problem with having three screens, Chaos Engineering: Crash Test Your Applications, you can order the Early Access Program that's what it is Manning Early Access Program. Manning, if you're listening to this, I remembered. Manning Early Access Program for Chaos Engineering: Crash Test Your Applications by Miko, I'm not going to try your last name. If you want to get it for free, follow the instructions. Go back and listen to them again, because I don't remember what I said. Miko, hopefully you had a good time today. Thanks for hanging out.

Mikolaj Pawlikowski 28:18
It was a pleasure. Thank you guys for having me.

Darin Pope 28:21
You'll have to come back when you get your book done done. It's always interesting to talk to other authors to where you're still in the heavy writing phase. Correct. Is that true or not true?

Mikolaj Pawlikowski 28:41
Sorry, you kind of

Darin Pope 28:42
Oh, sorry. So you're still in the heavy writing phase of the book. Correct?

Mikolaj Pawlikowski 28:48
Yeah, that's correct. I'm about two thirds in right now. And I left myself the Kubernetes chapter for the very end.

Darin Pope 28:57
Oh dear.

Mikolaj Pawlikowski 28:59
Yeah.

Viktor Farcic 29:00
Are you waiting for Kubernetes to mature? Or to stay?

Mikolaj Pawlikowski 29:07
I think I was I had to schedule just after a chapter on Docker and after explaining Docker and how to break it. I kind of needed a holiday I guess. So I took a little break. But yeah, I'm still very much into that. It's not, you know, doing the final polishes yet.

Darin Pope 29:28
After creating books and courses, we totally get it. It's never ending and it usually ends up not exactly where you thought it was going to be when you started day one.

Mikolaj Pawlikowski 29:39
Oh, yeah.

Viktor Farcic 29:41
Hey, is this your first book? Right? Okay, We need to have a session when you finish because I will give a very important question to ask. I mean, I can I can tell you the question now. You think about it until you finish the book. Will you ever write a book again?

Mikolaj Pawlikowski 29:59
Yeah, ask me that in a couple of months. Right now, I don't think I would say yes to that. It's, it's yeah, every time I try to estimate how much something will take, I just need to like double that estimate. And it's still not enough. But it's fun. I mean, that's probably like the best way to properly learn things.

Viktor Farcic 30:21
Yeah. It's amazing. This is I've been saying that all the time and this is the first time I hear from somebody else, kind of like, writing a book is amazing way to actually go deep into something yourself. At least that's my theory.

Mikolaj Pawlikowski 30:38
Yeah, it's like your quote that's been attributed to pretty much every known scientist that you know, if you can't explain something simply you can't you don't understand it well enough. And, you know, when you actually put it on paper, you realize that Oh, look, that's a gaping hole. I should do something about that. So yeah, it's a it's been very humbling this experience and, you know, about halfway through I really didn't think I would actually finish that, but it's coming up. I so so the book you wrote on on Kubernetes chaos engineering. I only scanned through that. It's also pretty recent right?

Viktor Farcic 31:18
Yeah I think I officially finished it we officially finished it. What was it like a couple of months ago? Something like that.

Darin Pope 31:25
The Chaos one you're talking about? Yeah. Miko, is that the one you were talking about the chaos? Yeah. Yeah. That was done in. Was that the last one we just did?

Viktor Farcic 31:36
I think so.

Darin Pope 31:40
Okay, so So right now, this year, this year alone, we've had Canary, we have chaos, and we've got the catalog.

Viktor Farcic 31:48
I have a special formula and I think that Darin is not happy with it. You know when when I heard you saying I need to polish it and I don't know why I just write it down with something else,

Darin Pope 32:02
and that and that's the joys and Manning, please forgive us. That's the joys of self publishing versus going through a publisher. Not nothing personal. But

Viktor Farcic 32:13
Darin makes a pull request that says every every every third word that book is misspelled.

Darin Pope 32:22
You've gotten better. You've started using better tools. So it's only every fifth word now. Okay, Miko, thanks again.

Mikolaj Pawlikowski 32:29
Well, thank you.

Darin Pope 32:30
And if you Yeah, man, thanks. And I said, well, we'll get you back once the book is done. Go ahead.

Viktor Farcic 32:35
Why don't we do some live session and show us PowerfulSeal?

Darin Pope 32:41
That'd be great. Would you be willing to come on live with us on a Friday? Not this week, but we'll we'll wait until this episode comes out. If you're listening to this today, when it first when it comes out. Today is actually June 24, which would be June 24. It's a Wednesday So maybe June 26, which I believe is a Friday should be because the math is right. Now we'll talk about it. Yeah. Maybe you'll come on live with us and show us show us PowerfulSeal live.

Mikolaj Pawlikowski 33:14
Sure.

Darin Pope 33:15
You're willing to, you're willing to do it without a rope. Without a safety harness? Well,

Mikolaj Pawlikowski 33:24
I can try this production, which thing?

Darin Pope 33:26
Now we do it on YouTube. But

Mikolaj Pawlikowski 33:28
yeah, I can't do it on production, as long as it's your production.

Viktor Farcic 33:34
Darin is going to find one of the clusters of his customers. He's not going to tell them anything, and he's going to let you do a live demo on them. Then he's going to ask your customers to guess which one was selected?

Mikolaj Pawlikowski 33:45
Yeah.

Darin Pope 33:47
No, that's not what's gonna happen, but close. Alright. Thanks again for listening to episode number 61 of DevOps Paradox.

DOP 61: How to Use PowerfulSeal to Create Chaos in Your Kubernetes Clusters

Show Notes

Links from the episode

Guests

Mikolaj Pawlikowski

Hosts

Darin Pope

Viktor Farcic

Links

Rate, Review, & Subscribe on Apple Podcasts

Signup to receive an email when new content is released

Transcript