 All right. Thanks everybody for staying for the very end. Again, no laughs guaranteed, but this is a total experiment. So you are guinea pigs. This is the first ever GitOps comedy show. So welcome. Thanks for staying. Take your waitress. Yes. All right. So my name is Tomo Nakahar. I run developer experience at WeWorks. I have my trusted sidekick Sebastian here who's also at WeWorks. And I was really lucky that I invited Christian Hernandez from Red Hat, as well as Chris Short. But Chris Short is now short on time and off to record a radio show in three minutes. So the good news for you is that Chris Short was the funny one. Yeah, he was the funny one. So that's why we need to find out. Yes. So first quiz question is, Christian, are you doubly as funny, half as funny, or 0.233 as funny? I am. I'm almost fractally funny. Yes. He impersonates funny. Not as, I don't have as nice his voice as he does, but I try. No one has the radio voice of Chris Short. Okay. So quick plug, though. Hopefully you guys saw the cards for us. It was very, very exciting. Please join us next week on October 20th on this GitOps one-stop shop event where you will see all these amazing companies show how they offer GitOps to their enterprise customers. So if any of you are thinking to start working with GitOps and want to use one of these platforms, please join us. The website is gitOpsdays.com. Same one that May was talking about is our various talks on GitOps days. So we just put it on the same site. So please join us for that next Wednesday. All right. So as I mentioned, this is all participatory. So hopefully if you use Twitter, check out our brand new handle, which is at GitOps comedy. So this is all experiment. We'll see if this works. So as I interactive, by the way, Sergeant Rupp interactive. So we really do want you to go and pull up the account. Yes. So for those of you online, for those of you here, hopefully you can see the polls that our trusted assistant Scott over there will be tweeting out at the right time. So at that time, like put in your votes and then we'll see the results and see if this works. We're not kidding. These are the tweets people. They seem to be distrusting that people will tweet, but I understand not everybody's on Twitter, but we'll try this out. All right. So we're going to do an easy one. And maybe Christian, do you want to read off? So we're going to do easy one. What is GitOps? What is GitOps? So first choice, a compliant way to do operations that uses Git for version control with an audit history. All right. Number two, a way to minimize downtime because a Git repo is a single source of truth for changes and rollbacks. Number three, a way to automate manual tasks that increase velocity or for all of the above. Yeah, sometimes that's that was my favorite answer in college. It got me through college. Yes, it's all of the above. So to give you some time. Fairly. First of all, does everybody see the tweet? Did Scott get the tweet out? Yes. Excellent. Everybody's voting rapidly. All right. Voting. All right. And Christian and Sebastian, any thoughts? Oh, why do we not have a choice for like a declarative source of pain? Yes. Well, the fourth one was all of the above and much, much more. With that, we're going to see if this actually works and we'll slide over. Is more really an option? Yeah. More is an option. Always an option. I can't find my mouse. Oh my God. I'm going to need the tech help because my mouse is invisible. You guys just got to shake the finger. That's what I do when I get nervous. Seriously, it is not there. Yes. It's like a little... Okay. Well, we're going to have to trust... Even to shake the finger? Okay. Scott, yell out. What are the poll results? So we have a mic. Found it. I don't need to shout it. The poll shout out metaphorically. Okay. So the poll results are ops with git version control, 9.1%. Well, just tell us the top one. What's the top one? Oh, the top one is all the above. Okay. All right. Yeah. 90.9%. I was going to build up some math, but okay. There you go. See? All right. That's how many votes I did. Congratulations on that. So one thing we definitely want to mention, as has been mentioned many times today, we are really committed to GitOps. Obviously, that's why we're here. And we think it's really, really important that there is a CNCF official entity out there that governs those definitions and those principles. So we're so excited that they've been published and official. So thanks to the hard work of the many members and the chairs, especially including people like Dan and Scott working in the GitOps working group. So please check that out. All right. I'm here because they're paying me. Sorry. I'm here because they're paying me. And we got one laugh. The first one I've ever intentionally gotten. I'm mostly unintentionally funny. Yes. All right. So hopefully you got the hang of this. So we are very lucky that many of you might know, Javeri Khan, who is definitely one of the tops GitOps practitioners out there, and currently works at Snowflake and has done GitOps at previous companies. So Javeri very kindly has done these pre-recorded bits that will be part of our quiz. So I'll do a quick bio. And I'll let you know that before we got up here, timeout, who's a Star Trek fan, wanted me to do the con. The con. I refuse because the GitOps con. And Javeri pronounces the name hot. Hi, everyone. I'm Javeri Khan. I'm a senior platform engineer here at Snowflake. I've been in this infrastructure domain for the past seven years, and specifically with Kubernetes for the past four. I've been a part of different infrastructure teams that have spanned over startups and enterprise on-prem and also multi-cloud at the moment. I mainly have experienced managing high-volume deployments and supporting infrastructure buildup with a special focus on reliability and automating everything related to GitOps to aid our developer productivity. In my current position right now, I support the building and maintaining of our container platform here at Snowflake, which comprises a multi-cloud Kubernetes cluster environment. Excellent. So we are at very good hands. So thank you to Javeri for providing some of these actual real-world experiences and thoughts that come from that. All right. So our next real question is, what are ideal environments for GitOps? Christian. So everywhere, every environment, everything. You can use GitOps for everything. Just everything. I've GitOps my dog. GitOps my bathroom. It's true, as May was actually saying, some people say they want to start GitOpsing all the things. Yeah, GitOpsing all the things, exactly. GitOps is only for production. Take time to set up and use it. Number three would be, GitOps is only for test, dev, staging. You don't want to scale it. And then number four, GitOps is for home tinkering. I can choose any one of these, but I definitely use it for home tinkering. Yes, or I guess it's for only. All right. So hopefully you've started voting to see the poll from Scott. I think of all the missing options that we could have had. GitOps is for fun. All right. Those of you online and here. All right. I think that's a good enough time. So Scott, what's the top vote? Okay. Well, right now we have, hang on, hang on. Let me just refresh this. We have nine votes. Let's let people catch up a little bit last time. Oh, my bad, not bad. Okay, okay. 23 votes. You forgot to refresh. That's what it is. Eventual consistency. Okay. That's right. Yeah. So you had some drift. It's 250 votes, right? Right. Come on, everybody. Okay. It's, we got this. We got this. Okay. So top vote right now is everywhere. 91.3% and then home tinkering only 8.7%. You guys are so smart. The answers are not obvious at all. We've got, you know. All right. So yes, hopefully you saw that again. And as we go a little bit more into these details, it's not that there's a right or wrong answer. These are sort of the experiences that Javeria has. So don't feel bad or feel like it's focused in a particular way. But yes, for this one, still fairly broad and GitOps, all the things. All right. So, oh, yes. And Javeria is going to share a little bit from actual experience. So in all of my experiences, we've used GitOps for staging and dev and for production environments and certain tools like Flux. They're not only easy to set up, so they are ideal for test or frequently updated environments, but they are also scalable enough for production use cases. For example, Argus CD and Flux have first class to perform multi-tenancy and they also provide a host of security features for production environments. So in a nutshell, GitOps can be used everywhere. Excellent. I don't know which Javeria I'm supposed to be looking at. Yeah, I know. There's a little duality there. Well, Javeria is everywhere. That's right. Just like GitOps. In all things. I'm more scared by that than by the GitOps. We're going to dig a little bit deeper now. So if you need to test clusters that are spun up and down regularly and need frequent updates for application changes, what GitOps tools might you use? So this is a little bit trick question because we're going to have to try to guess what Javeria did. So one, Christian. Flux notification server to update PRs with deployment results? Or? Argo CD and YAML files in a single repo? Or? Or you need Flux and Helm customized single repos? Or? Or again, my favorite answer, got me through college, all of the above. Which is possible, but we'll see. It is possible, yes, right. I really did write an answer to this question that you didn't put in there. I just read bash. You see bash? That's right, yes. Until the right. So while we're waiting for the votes to come in, there are people who do bash. What would be some... Me? I'm not joking. What would be some things that might not work out maybe at scale or whatever? I don't know. Up until recently, there was a bunch of bash in Kubernetes, right, in those static pods. What's wrong with that? Yeah, let's not look at every go wrong with a bunch of bash scripts. Again, it's all use cases, right? So people have different. All right, Scott, how is it looking? Don't say the number because it's in the hundreds. I will not say the number. Also, I was about 10 seconds late, so that was my bad. So let's give everyone just another moment. Wait, you're saying you didn't wait? Wait, 19 seconds? Oh, 19 votes. No, okay, we got 19 votes, we got 19 votes. So you weren't paying attention? I was basically... I'm not interesting enough. I feel like I'm in middle school again. Sebastian. Okay, they wouldn't let me on stage in middle school. Yeah, okay. Very good reasons. Various reasons. All right, well then, we have 19 solid votes. Let's say 20 votes. Okay, top vote. Ooh, okay, divergence. Top vote is, it is all of the above, but not by a 90 plus percent margin. Only by 52 percent. All right, what was the second? Second, second vote is, a second highest vote is 21 percent, and that is Argo, CD, and YAML files. Yeah, interesting. Well, again, this was just what Javeria has done, so why don't we hear what Javeria is to say? Who actually uses all kinds of technologies and tools? Not just one. So in my specific experience, for these kinds of environments, I've used flux with a combination of helmet customized and with small single repositories because they're obviously test clusters. The reason being that as Kubernetes is getting more and more widely adopted, we're seeing many companies adopting a development life cycle that allows developers to spin up these test or criminal clusters on demand and use them for testing their applications right from the start. This is great because developers get to test how their applications will behave in a containerized environment that is as close as possible to the production one. But in my opinion, the biggest benefit get-offs tools offer here are automating the update cycle and cutting down the time it takes to get new changes on these test clusters. Now, since these are frequently deployed and destroyed clusters, you want something that is quick and easy to set up every time. And I think flux offers this simplicity and which is just a few steps. You can have a GitOps-based workflow set up ready for your test environment. This installation is as easy as deploying any other typical pod in your cluster. But then generally, I would say the most useful feature for these kinds of environments is the automatic image pool feature that Flux has. So basically, it not only watches the manifest in your repo, but the automatic image pool feature will deploy new application changes as soon as it sees these new images in your registry. So you don't have to even do any changes or any pushes for just that specific change set as well. I think by now you guys are all probably wondering why Javeria is not up here trying to tell the jokes. She did write out a bunch of jokes for us. She's a planner, but I forgot them all because I have no working memory. Yes. So we prepared plenty of extra ones. If you're interested in following up with us, we're happy to chat with you about more best practices and such. But given the time constraints, we'll skip ahead and just thank you for coming. We really appreciate it. Again, come to our October 20th event. And Christian and Chris are always on every Thursday, right? Every other Thursday. Every other Thursday. I do Twitch. We're on Twitch, primarily, but you can also catch us on YouTube. Get Ops Guide to the Galaxy. I love this show. I made my youngest child Zafod. And I'm a huge Hitchhiker's Guide fan. And I consider it equal to the Bible in many respects. My next question was, why are you guys still here? Yeah, that's it. So realize that there's alcohol. So with that, we will end. Thank you for coming to our Get Ops Comedy Show. Thank you very much. Thank you.