 Hi and welcome to the SIG usability conference talk for this coupon. Why does anyone use Kubernetes anyway? Happy to see you all. I'm Tasha Drew. I am director of product incubation in the office of the CTO at VMware Gabby Hi, my name is Gabby Moreno. I work at IBM. I'm a senior product designer Working on Kubernetes Sometimes OpenShift and just generally open hybrid cloud technologies Carl. I'm Carl Pearson I worked at Red Hat once upon a time working on OpenShift mainly and I'm still involved in SIG usability and working on Kubernetes And I'm Josie Pinata. I'm at VMware. I'm a senior product designer and I work on developer experience on Kubernetes Cool. So a quick introduction to our special interest group So we are part of the upstream community in Kubernetes SIG usability And we are tasked with the mission to look at core end user usability of the Kubernetes project So this ends up covering topics like user experience Accessibility and the goals of our SIG are to ensure that the Kubernetes project is maximally usable by users of a variety of different backgrounds and abilities I'm one of the co-chairs. Gabby is the other co-chair Currently the main project we've been driving as a SIG is a jobs to be done study that we've presented on at previous KubeCons And are continuing to drive gathered data for With the goal to eventually be able to release all of this data and use it as a foundational open source block That we can then drive additional studies Into user behavior and user needs from we do have monthly meetings of our upstream group They are Tuesday at 10 a.m. Pacific if you want to engage with us Some of the easiest ways are via these links at the bottom of the slides. You can join our Slack group We're in the Kubernetes Slack and then we have a channel called SIG usability that we all are in We have a GitHub page where we store project information And then finally that we have a Google group And if you join our Google group you will automatically get the calendar Invitations including the zoom information so you can attend our meetings And you we also have an agenda doc that is linked to that Google group So if you have a topic you want to talk about you can propose it either via the mailing list or via slack And then we can add that to our next meeting agenda as time permits So today we are going to talk to have more of a panel sort of interactive format for this presentation and We're going to kind of leverage the wealth of information that Gabby and Josie and Carl have all been Gathering as we've been going through this jobs to be done study So to kind of kick that off we're going to give you a quick overview of just how are we studying users? So Gabby would you like to tell people how we're currently studying users? Yeah, sure Tasha. Um, so there's kind of like a Term in the user research field called jobs to be done research that is really cool Because it takes kind of like the best of qualitative research which we think of in terms of you know, I'm gonna go interview users to takes qualitative and then the best of quantitative because What you do is when you're interviewing people sometimes they say things and that you're like, oh, that's super interesting But I wonder how many people like is it a real Market need so like, you know, it's really hard to be like what one person says applies to many people and so I think one of the really cool things about Chops to be done is that through the interviews that we're doing We're kind of trying to get through like all the opinions that people have of Kubernetes because you know as we all know Kubernetes users can be super opinionated And so we're kind of trying to see exactly like what they Specifically do with Kubernetes their behaviors and ultimately Kind of like distill that from the interviews and turn it into a survey This is kind of very typical of jobs to be done. The survey is sent out and You use kind of like a typical like for each Job or behavior that someone could do with Kubernetes It's rated on how important it is and then also how satisfied people are today with how they're with the Experience of doing it. And so at the end you kind of get this nice nice Insight into what are the most basically underserved Jobs that or tasks that people are trying to do with Kubernetes that that as an open source community We kind of go rally around and tackle Awesome And Josie, what kind of users have we spoken to so far? We've spoken to a bunch of folks from a very like a very variety of different places in the world We've spoken to a couple of people in Southeast Asia as well as Europe and in Canada These range from IT consultants Dev managers SREs, we've even spoken to a couple of people who teach courses on Kubernetes So a real variety of folks Awesome And so how is the work that we're doing here different from typical work on Kubernetes? Carl, what do you think? So what's interesting about this is user research in a more formal process isn't Often applied in the open source space that people get tons of user feedback But applying this more kind of structured user research approach I think it's kind of a new thing that we see in the developing the open source world especially so One thing that we've had to figure out There's a little different than the rest of the Kubernetes contributors is something around security Of course, security is really big in Kubernetes overall That's obviously a huge priority across a bunch of different working group areas But for us, the the domain of security looks a little different So we're like Gabby mentioned actually interviewing people So we have videos and recordings of them talking about the way their organizations work One way to get the best findings from people is to make sure they feel comfortable to share anything that they might want So we have to make sure that those findings are anonymized and their identities are Protected from their actual findings so you know they can speak freely So that's one thing we've had to figure out is you know, how do we protect our users and Their identities and keep that information secure and in the open source world You know, we have a lot of really interested and excited people that are kind of jumping into the project So we had to come up with some more rigorous processes to figure out You know, where are we going to store these videos who has access to them? How are we going to keep these videos over time and you know, discard them eventually? Where are we keeping the findings who has access to them? So we've had to come up with some more operational components to figure out You know how we protect the users that are ultimately serving this research project So a little bit of a different take on security than I think most people and I think it about Kubernetes working groups Yeah, absolutely I know that that's come up Several times as we think about this because people have been so open with us And when someone sits down for a jobs to be done study like it's like an hour plus, right? Like it's not, you know, kind of like a 30-second You know survey that you're taking on SurveyMonkey or something And so they're really giving us a gift of a significant amount of their time so that we can really Understand from them what jobs they are trying to do with Kubernetes and how they're approaching that to sort of inform our research And when someone gives you that much time, you really want to make sure that you protect their data You know that you're thoughtful about it But at the same time we're balancing that with the needs to be open and share things with upstream so others can leverage them And so that's kind of the line that we've been walking and really kind of carving out for ourselves Since it's a different use case than many of the other SIGs and working groups in upstream are dealing with So yeah, it's cool So for people to kind of follow up on that who do want to start contributing What is a good way for them to get started? Gabby? What do you think? Well usually, you know, we're pretty friendly group so I you know coming to the monthly meetings are usually kind of like the best way To get started and you know, we'll usually be giving an update on where we are and jobs to be done at that time Or if you have any project ideas as well but some of the ways that people have contributed in the past and I can also talk about some of the Opportunities that we have now When we were doing the one-on-one interviews We were really leaning on contributors You know like Josie like Carl like myself that have kind of like Done user research before or maybe Carl you you came in with like a really solid jobs to be done Kind of like a project under your under your belt. And so, you know, we had kind of like the this Opportunity of you know, if you're a user researcher or designer that has done this before like Come do it in an open source type of environment But if you haven't done it before One of the opportunities that we offered was for people to shadow these interviews just to kind of learn How they're conducted The opportunities that we have available now are more around now that we have actually finished our phase of interviews Now we're going through and analyzing them. So we're looking for things like what are people specifically doing and so You know that can be a little bit time-intensive, but one of the things that we are offering is, you know, just Training up people and or just giving tons of examples on how do you go through a transcript of an interview and highlight or identify behaviors? and so Again, like the the tagging contributor role and ultimately that will be reviewed by By my Josie Carl myself Just to kind of like have some type of quality review to the process. So So, yeah, so those are kind of like some of the examples of how people can and have gotten involved Awesome. And is there a good way for them to kind of reach out to get involved? Like what would that look like? Yeah, um ours. I would say either through Slack You can hit up myself or you know any any of the people here on the call So Slack Or or one of the monthly meetings are probably the best place Awesome. And how is the how are the findings going to be used? One as once we're in a more complete state with this effort So, I can take this one. Yeah, I think that's an interesting big question mark Because like I said, no one's really there aren't a lot of examples of this formal research process and open source Contributions, I think there have been more research contributions in other Projects that are more usability focused or it's like, you know, fix this button. This is very concrete What we're doing is a little bit higher level. Hopefully a little bit more influential where you can really reshape strategies and Make sure that you're kind of pointing the ship in the right direction So I think that's one thing that remains to be seen. I think a big thing will be Socializing the findings make sure that we really get to share them widely and make People understand just how it came to the conclusion that you know, we don't just think these are fun They're like good that we actually have a pretty rigorous empirical process to get there So I think in a best-case scenario, you know, it might shape some possible Like new working groups that are tackling problems that like Gabby mentioned are part of this kind of underserved quadrant Where they're really needed but really hard to do right now And that would kind of be the best case and we can get some specific work streams coming out of the high-level findings that we get Awesome. Yeah, so now we're gonna transition into talking more about what we have been learning as we've been going through this Research study. So Josie, I'm gonna give you a nice spicy question first So why are companies moving to Kubernetes? So keep in mind, we're still synthesizing this research and this is like What's on top like at top of mind right now? Maybe when we do the full research The findings might be a little bit different But a number of folks mentioned hype They've spoken about the fact that the decision was sort of made because the market is moving in that direction And they didn't want to be left behind And so even before like the technological opportunities There was a really Big consideration of like, oh everyone's moving to Kubernetes. We need to as well Yeah, I mean there is like a huge part where just the ubiquity of a platform ends up driving the ubiquity of the platform Carl, did you did you have any sort of takeaways from this when as you're speaking to users about why they're moving to Kubernetes? Um, some people I think yeah I get pulled in by the hype and then they find that the community there is is really supportive and they have a good chance to contribute And I think there's this kind of collaborative nature That brings people into it And I think that that's probably one of the biggest drivers I've seen beyond just the hype and then some people also have had experience at past companies now as it gets more mature and Get into a new place and say I think we could set this up in a way that you know Might serve XYZ need a little bit better. So there's a little bit more kind of Institutional knowledge now that's moving across companies. Although I would say that's more rare now I'd be curious to see if it gets more common later down the road Awesome. I gap you anything to add there Um, I would just kind of like echo What would Josie and Carl have already said? Um, I think the the Carl what you said about like the community that Kubernetes has I think is kind of like a huge part of it, you know, even being here at KubeCon, um, you know, just having that community and also just generally You know for the people that Do kind of like start trying to put together like proof of concepts for Kubernetes and they just kind of like discover The api and just kind of like how well formed it is. Um, I think that can be really Exciting for a company that has done all of this kind of like manually And now they're discovering that they can just think about it in terms of software. Um, so I think that that's something cool, but then you also kind of hear things like When you open this up to like the Maybe like a very product development focus team that doesn't want to focus on infrastructure Then you start seeing all like the articles about like why Kubernetes is hard and why Um, why it's not worth moving over to Kubernetes because it's really like an infrastructure oriented project So as we're speaking to these users, um, when they do move to Kubernetes Do they find it valuable? And I think this probably ties into these different layers of concern like the infrastructure user Or administrator versus the development development teams Josie, have you found anything about this? So those companies that we're talking about the hype around Kubernetes um, a lot of the challenges that these organizations deal with is the fact that Um an upfront cost has been made in order to transition to kubernetes But then there's no runway for onboarding developers or rewriting applications so that they can best utilize kubernetes um, and so what you have is uh teams that are trying to product like provide value but The application isn't set up in a way or hasn't been rewritten In a way that they can really make those changes uh, so Like the hype is very real And there is a lot of opportunity But it's just time and money needs to be invested so that developers are properly onboarded Um, and applications are rewritten so that they take full advantage of the features and capabilities of kates And then when we do speak to Users who do manage to make that full evolution into writing kubernetes native applications versus perhaps cloud or container native applications What then do they or do they not find beneficial about kubernetes at that point when they're really trying to adopt it? You know holistically um So a lot of the the common key points of scale cost speed And the fact that like kubernetes is written in a way that makes sense to developers Often like devs don't really want to interact with the infrastructure at all But kubernetes and infrastructures code sort of makes it more accessible Awesome So have we learned anything about just the getting started new user experience? So far carl Yeah, I think we've learned a few things I mean one we've explored the kind of concrete pathways that people take to learn about kubernetes and You know people definitely spend a lot of time in docs in tutorials just a lot of kind of like getting their hands dirty and Just finding the resources that are out there and luckily it seems like there are a lot of resources that are available to the infrastructure type person who Can really get their hands dirty and you know get to testing things really quickly Like the getting started experience part of it is just like You know to spin something up and see how it breaks and figure it out from there And they're really quick to be able to do that um I think the other side of that is the kind of developer kind of like what the end user of the kubernetes might be is It's kind of a different story. Um, it can be a bit more more challenging for someone that isn't as infrastructure oriented to To get on board and learn kind of the unique ways that kubernetes will interact with their applications So that's something that It seems like a big job of the more administrator role is is is helping developers on board So, you know, they can get their hands dirty really quickly But then they have to bring those developers up to speed which can be a big challenge There aren't as many formal resources for that partially because a lot of it's very idiosyncratic for how their organization works But that is kind of there's there's two sides of the coin for who finds it easier to get started with kubernetes Yeah, one of the fun facts that I remember from when we first started doing this study was that we were learning from a lot of people that Unless they had to learn how to use kubernetes for their job like their boss was basically like learn this or you're fired We the kubernetes community tends to churn people who don't have that really driving need to like figure it out And so I think that really indicates to us that there's a lot of better onboarding We can do to make this a project that people can grok faster without the big push to learn like for their job immediately And so I think that's something that I really hope to see our team start to kind of crystallize is like what could we Invest in to make this easier And I think that there's a debate going on too about Do developers need to care that they're running on kubernetes, right? Like where does that line have to live for the infrastructure team to be able to use it incredibly Impactfully, but the devs to just be like here's a docker container. I'm moving on with my day, you know, so yeah What do people love about kubernetes gabby? What have you found? Yeah, I mean I would kind of like go back to like the I think what we hear again and again and again Is people saying that they love the api like They see the api and they see these abstractions uh like infrastructure abstractions and They're just kind of amazed that they wrote This themselves before and that now it's like You know, I don't ever have to write this or even maintain this again Unless you want to contribute to like the open source project But then it's like you write it once and now everybody can use it um So I would say that you know, there's many things that people love about kubernetes But but I would say that that's kind of the most resounding one that I've heard Awesome, uh, carl. What do people not love about kubernetes? Um, one of the big things people don't love is upgrading Uh, it can be a big kind of stick in this the smooth process of of letting things run automatically in the background It opens you up to, uh, you know The idea is to you know patch more security vulnerabilities But there's a lot of little things that can kind of break along the way and to keep your eyes peeled for You know changes to the api or anything like that and You just people are never quite sure how it's going to interact with everything they've built on top of it so It's still kind of a careful careful process for people and They are very strategic and when they choose to do it some people are faster some are slower but everyone's definitely thought about the kind of The the trade-offs between upgrading faster and slower and and try to make the right choice there Carlin even to to add to that, you know, not everybody uses like managed kubernetes You know, like if you're running Because I mean the environment is also a thing like if you're if you have something that's maybe like latency sensitive So you might be doing something on prem Um versus something like and like the nearest like cloud region. So I don't know I think it's also like Upgrading your own version of kubernetes has its own kind of like hoops and hurdles to run through as opposed to upgrading Something that's like managed definitely uh, and then so Just to kind of wrap up the session on a bang What based on the data that we've been collecting What has been our user's core underlying goal for using kubernetes? Carl, do you want to kick this one off? Yeah, this might be a hot take that could change as well But uh, ultimately I think kubernetes really Gets infrastructure people super excited because you see this is such a cutting edge and efficient way to do things when you get it set up Right, but it's kind of funny because I think ultimately this is in service to the applications that You know are running on top of kubernetes So it's like this is all for the developers like that and goal is just to make the developer process For their applications really really smooth. So it's kind of like get out of the way as much as possible for these developers and You know So I think that's kind of an interesting dichotomy because we're mostly talking to infrastructure people But and the very end it's kind of in service to these developers and they're like business critical applications that Are are running on top of the platform Yeah, absolutely So how do we optimize an experience for a tool set that And a user base that basically doesn't want to know that it exists, right? Like it's like, yeah Kind of fascinating cool. Uh, does anyone have any uh last thoughts you'd like to um put on the board? No Cool. Okay. Well, thank you everybody so much For participating in this panel and we'd like to just really welcome the community to join sig usability We will be working continuing to do these jobs to be done studies With a future goal of being able to wrap this up with enough data for us to Make some really nice conclusions and then share that anonymized data that we're going to go through a really Strong sort of sanitation process on before we serve it back up to everybody So you can do your own reporting and learn from it as well. Um, so thanks and we'll be chatting with you soon