 Okay, I think we should probably start. Do you want to kick off, Jane? Sure. Can you see the screen? We can. Thank you. Okay. Okay, so here's the survey. So it just goes through the questions. The first question just to ask people to identify themselves. So here we're saying that if you identify yourself with multiple of those personas, then you fill this out multiple times. Do you think this is necessary to separate out to fill this out multiple times? I think if the audience is going to be the end user group, I think that's possibly overkill to do it that way. Okay, so maybe... Because most of these... End users, right? Yeah, okay. Yeah, we also included vendors, but yeah, but I probably it's more for end users. So just remove this. Just say multi-select then. Otherwise it would be too... Yeah, good idea. Too complicated otherwise. You know, I'm sure there are a bunch of the end users who are actually, you know, project contributors to projects and things like that. So I think those things are still valid. Okay. Okay, so yeah. And what project... What storage project you are using or contributing to? That's the next question. So here we include storage projects in CNCF. And then the other, if there is a project that is not... Yep. You can specify here, anything here. So first, the first question. Anyone has any comments on this category? So this looks okay. Do we need to add more or is it too much? Or just keep it this way. You can always comment later. Yeah, I think that's fine. Okay. Do you know what would be actually, sorry, just something that occurred to me. It might be useful to have a question at the top to kind of say, have you seen or are you aware of the of the white paper that we did? Because at least... Oh yeah, sure. Might make them look at it. Yeah, sure, sure. That's the question just to say... Oh, that's on here. Yeah. Enscape. Okay, I include the link here. And then what is your experience with the cognitive storage? And the next one. How is your storage solution deployed? So hardware solution, data centers, software components, commodity hardware, public cloud providers. So Quinten has a comment here. Maybe we should add another category. Storage as a service providers that are not primary cloud providers. If they are offering like the database as a service on public cloud or something. So... Yeah, that makes sense. Anything else? Let me just add another. I'll just add another. And what data access interface layer is consumed? So there's a lot of questions. That's cool. Then next is the blog section. So here are the questions if you have blog storage. Here are all the questions. So the first question, basically what is the blog storage system you're using? So Quinten has a comment here that if it is a well-known storage system, then user can just skip all the other remaining questions. I mean, I don't know what is... If it's AWS, EPS for this well-known. So should we give any examples here or just let the person to the judgment? Yeah, that's a good point. I think if they're using a well-known vendor or a well-known cloud provider, then it's probably right. I guess a lot of the rest of the questions kind of become... We sort of within the storage working group anyway, we should be able to determine the rest of the stuff. Do other people have any comments on that? Do we want to list some names? I think I would struggle if we were to... Maybe just leave it to the person whoever is filling this out as you decide. And then if they give a name, then if we don't know, then we can always do some research and we reach out to the person who answered the question. Okay. And then the protocols and also whether it's a local remote or distributed. A lot of those can probably be skipped if it's a well-known. And also, most of these questions, we also added a don't know at the end if it's probably sometimes people don't really know. And the storage of topology, centralized, just beauty-shotted, hyperconverged. Is that something specific to block or should we...? So basically, the problem is if we just... I have this question in every category. So for file, we have this one as well. So if you're only using block, then you just answer this one. You skip the whole file section. But if we just make this one question, then I just... I mean, how do we differentiate? That's why I kind of... So kind of those questions are repeated actually for block and file. So we can go through this and maybe decide do we want to take this out or because they could be using distributed block, distributed file, right? So this could be different. Those are in both sections. I think that we would probably know the answer to that based on... Yeah, so basically, a lot of those can be answered already. They just say what is the storage system you're using, right? Yeah. But probably not all of them. There are so many storage systems. I mean, yeah, if it's just like AWS, and those we probably know, but I just can't say that for every sort of system. So here they say, okay, if you already... That's the name. Maybe you can just skip all of those questions if you want to. Yeah. I mean, let's leave them in for now. Okay. Okay, yeah. It's probably... Can you see there are too many questions? Basically, those questions are repeated for each blog and file. So we can, yeah, go through this and then maybe we can talk at the end. I think question eight and question nine are probably sort of, in my mind, they're question marks. I think whatever products they say they're using will probably understand the answers to those questions. Okay. Yeah. So they can probably just skip all of those if we already know the correct name. I think question 10 and question 11, those are interesting to understand because if lots of... We might actually end up having some use case scenario or maybe some focus on perhaps, I don't know, snapshots or encryption or erasure coding or something. So those are probably words... Those are probably words asking because different systems can support multiple things generally. Mm-hmm. Okay. Okay. And then the fiscal layer. Yeah. I agree with Quentin. That's probably not that important. Mm-hmm. Okay. So just leave it to know or they don't have to fill out this one? Sure. Okay. And, okay, the next file system is pretty much very similar to the blog section basically. So if you answered what is the file system you're using, then you can probably skip everything below. Mm-hmm. The same, you know, the... Okay. I don't have this one because we have a table. I don't know if this is maybe too complicated. So I just added, don't know here. What is the underlying storage layer for the file system? Mm-hmm. This is the... Yeah, we have this table in the white paper, right? So... Yeah. It's kind of complicated. Maybe people wouldn't pay that much attention to all the details, but we don't know here. Clint, I think that table was something you had put in. Do you think that's worth collecting from the end users? Sorry, one more time, Alex. The question 14, I think that table was something you might have put together. I was wondering, do you think we need to collect that information from the end users? I think that this came after describing that there is essentially the complexity under the file system, like there's different ways of running the underlying storage. And I don't actually... I think the audience is probably ready to answer this more generally. I think it's a pretty... It's a more complicated question. Okay. What does everybody else think? Yeah. For what is right. We can remove this one then. I'll just remove this. I'll remove this. Okay. So the protocols. I don't know if it's worth asking this. Clint says maybe all of this are hidden. So I do add it. I don't know here. Sure. And then the storage topology. And so all of those are the same as before. Data protection. Yeah. I think similar to block. I think question 16 is probably not so important, but question 17 and 18 are probably useful. 17 and 18. Okay. Okay. We'll move after we're done with this. Should we remove some of this or still keep them? Okay. And then still have this last one. The same. Yeah. I think question 19 probably isn't needed again. Probably should just remove this one. Yeah. Okay. Unless anybody thinks it's important. Okay. So for object store. Basically it's just. Mainly just answer what object store you're using. And then you can skip this one. But I'm thinking so that's kind of the thing. But if you are setting this up yourself. It's like a local object store. Or swift. Then you may be still. We'll be looking at this layer. So. So here's okay. If you are using up just for our public operators keep. This one. Otherwise they can still answer this. I don't know. What do you guys think? This is not, not that important. We should just to get rid of this question. Okay. So, so I think. If we're looking at this in terms of. What's useful to us in terms of what to focus on next. And which projects to look at on that kind of thing. I'm not sure that. I'm just trying to think. I don't know that that actually gives us information to. Okay. I mean, it's, it might be an interesting to this thing to kind of say. Okay. 50% of our end users are doing so this day versus. But I don't know that it's actually something we can use. Okay. So probably maybe just need to just remove this one for every, because we also have this one. For the blog and just remove this one for. Yeah. Okay. All right. I also have this one foot. This one basically it just listed all the key battle stores. This is far in the white papers. Those are our reference in the white paper basically. So I don't know if there are other well known ones that we want to list here. Okay. So I'll just remove this question here. And. So, so the database, do we want to list any. Well known databases or just let the user answer this. That's a good question. It was some of this actually I was. So like some of those, right? Some overlapping assets like Cassandra. I thought I was thinking it. You should belong here, but it's actually here. But we can list a few. You know, I mean, that's a good point because Cassandra, for example, is often. And conqueror should be, for example. Yeah. Yeah, this one too. Yeah, those, both of those I thought it would belong here, but it's actually, because it also has the key value APIs. That's why it's actually listed in the white paper. So that's why I put them here, but they could be here too. So. You can be in both places. Then those maybe. Cassandra and. Okay. So do we want to list anything here or no? It's. What do the other people think? Do we want to list a set of databases or just have them? I think the other option would capture them. Thank you. Somebody wants to specify them as key value stores. I'm sorry. The other option would capture Cassandra and. You know, other databases if they're using them as key value stores. Right. Understood. Okay. So we can do it with the purpose of this, just to round out our knowledge of what the most popular ones are. Or what's the. Like are we, are we asking in context to like, who's running it on Kubernetes or. And what's. What do we have. Okay. Okay. So maybe we should add some questions. In here. So, so we do have, we do have this oxygen section, but Kubernetes or what content orchestration system do you use? And we do have those sections here. And then the plugins. Some questions about plugins. So we do have it here. But, but on 21, 24, we. When we're trying to understand where people are, I mean, I think we can find out the popularity of databases. In a ways. We're trying to understand if they're being used in cloud native ways or what. Well, so the. So we're going to send this question. They are to the, the end, the CNCF and the user. Group, right? So. We assume they are using them in the quality way or how do we, how do you phrase those questions? So. If I'm trying to, if I try and capture those, those tools for a second. So I think. We want to get a little bit of an understanding. So. We're going to send this question. They are to the end the CNCF and the user group. Right. So. We. Assume they are using them in the quality way or how do we. How do you phrase those questions. So. We want to get a little bit of an understanding. As to. What. The end users are actually using. Because. We might see, oh, I don't know. 60% of them are using cockroach to be in there for maybe a. Use case on cockroach to be beautiful. So. We might also find that they're using some. Projects which. Maybe as a candidate for. Becoming a CNCF project that isn't one yet, for example. So, so finding out. What, what those end users are focusing on might be an interesting might provide us with some interesting data, but I do also agree with Clint. I think it's probably like having a question, a more generic question. Maybe in the orchestrator section where we say, okay, I don't know. I don't know. I don't know. I don't know. I don't know. I don't know. I don't know what orchestrates are using. But then with what's thankful workloads. Oh yeah, that's a good point. Okay. Okay. Add this. Like what workload. What workload. What. Workload. What type. See you running. Okay. And I think it's okay to keep that to narrow. But I think there might be a way to make it a little more useful. By setting context that. You know, out of these, out of these database technologies. Which ones are you challenged by? Like, which ones are you thinking about? Operating differently. Like, you know, which ones are you interested in? Yeah. Maybe, maybe we can do that as, as, as questions. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. So if we say, what's the full workload? So you're running today. And then the next question could be like in three months time. What do you want to do in six months time or something? Does that make sense? Yeah. Yeah. I think in other issues, people may use these databases or key value stores. And they may use Kubernetes, but they may not use them together. Do we want to. So, yeah. That's kind of tricky. How, how do we separate those? Because yeah, so they, they may be thinking about, okay, they will be today they're running this, maybe not on Kubernetes, but they are maybe wanting to move them to Kubernetes. In a few months or something, right? So how do we. So maybe do we want to phrase the question as. What types of workloads are you running on the container registration system and then have another question to say what workloads are you consuming from the container reconstruction system or something like that. So, so we sort of differentiate what they're consuming externally to Kubernetes. Yeah. That would be interesting. If you've bought actually. Because I kind of suspect there must be quite a lot of people who are consuming stateful work. They might be running applications that are using states, but the thing that's storing the state might be outside of Kubernetes. I guess that would be a very common use case. So what type of workloads are you running that this will be include both or not on containers or should we just say. Not on containers. Or. That this include both. And then, and then specifically, what are you running on containers and what do you want to run in the future or something. Or perhaps we freeze the question along the legs of. Do you currently run. Database and key value stores inside Kubernetes or outside Kubernetes. And then do you plan to run databases and key value stores. In the next three months or six months or something. Outside of. Like. What is. We can say. No. Not something like it's not not on containers or something. Or. Maybe you should come up with a list of application categories and then reuse the same list for. On containers off containers. I don't know how you want to get in that list either. I mean, the. Detailed would be stateful or stateless and that's all the description, but you could drill down into specific things like. I don't know Kafka. MongoDB. If you wanted to get that detailed. I think at the very least to be useful, it might be say transactional databases. Key value stores and just put some broader categories. But maybe we want to get more specific. Yeah. I think the problem. With most of these services. If they're very, very long. You know, we might struggle to get. Yeah. Not only that, if you get too detailed on specific implementations like you list Mongo, then you. Arguably you'll leave somebody out and it just gets interminable. So I think maybe putting it, but just the two categories of stateful and stateless. Aren't. Enough detailed to be enlightening. So something in between. I mean, what, what, what, what, what if we say, what if we say. Do you run. Stateful workloads. In your orchestrated environment. And in brackets. Put it like example. Databases and message buses. And I don't know, you know, just, just, just give maybe a couple of examples of classes of stuff. That could be considered stateful. Maybe you want to add in production because. If you just want to know what people are doing, I think you can just go to Docker hub, which sorts by popularity and download. And you get a pretty good clue of what people. Are at least playing around with. Doesn't mean they're putting them in production though. So. Yeah. Okay. Okay. I mean, for the examples, you might as well pick them off. You know, go get the popularity sword order on Docker hub and choose whatever bubbles to the top as your examples, because obviously those are the ones people are familiar. Okay. So could do that. Just to get some. From there. Okay. Should we cover the rest of the questions then. Okay. So this is actually the, the end basically just. Yeah. Extraction. You know, the plugins. What type of plugins are using what sort of system. So the last question, maybe we can just, just to remove it. This is what are the frameworks and tools in addition to the plugins. But they probably already answered that in, in the beginning. Say, what are the projects that you are using or something. Yeah. Okay. Okay. So, okay. So. Okay. So if, if we can, if we can tidy this up. And get it to, and get it to Cheryl. She'll have a quick review and then forward it on to the, to the end users. And, and then she's also going to show you a session where we can, where we can attend. One of their meetings probably because I believe it's, it might be in two weeks time. So we can, we can then sort of prep over email for that session. Okay. Yeah. Sure. Okay. So I think we, okay. Let me just go through this one again because we talked about, okay. So I'll remove this question for all. And so we'll keep 10 and the 11. So for the rest of them, you think we can just remove them. Because it's, maybe obviously. Okay. And this one too, since you feel. Select it. If you give an answer and then we should, we can figure out. And. Yeah, it's probably this can go to. Okay. So that will make this simpler. Yeah. And then they can still skip if they want to. So we still ask, ask them if they want to skip this too. We just want to keep. Then only two questions. The body can say, which is a little bit there. Yeah. Yeah. It's only not that many anymore. Okay. Okay. So we'll do the same thing for the, for the file systems, which is, which is, you know, you know, you can get rid of all of this. Just to keep data production and data services. Okay. Okay. And. Okay. And then, okay. So I need to do some. Search on the doc hop to get a list of. Things people are using. Okay. So when you're, when you're done, if you can send it to the. To the mailing list, then I guess, you know, anybody can present any final comments. Okay. Those are the sort by popularities. Postgres is number one. Interesting. Yeah. Even though it says database, there's key value stores, like Redis is in there as numbers. Okay. So it's partly okay. We don't have to really differentiate. I would just put those names there just say, and then people can. Yeah. Okay. I don't know. Great. Okay. Okay. Good. Okay. Just to use this one, I think it's simple then. Those are okay. So this is database. They do, they probably, okay. So they, they kind of put everything together. So this includes the key values. Okay. They didn't really differentiate those two. It makes sense. I think the people post these images self-classify. So it's just whoever published it gets to. Pick their own category. You can see a category list over on the left. So. Okay. Some of these have gone in under. Arguably bad categories. I don't know. I don't see. There's no key value. I think they probably included key value in as part of database, which is fine. I think on this list. So that's, that's, that's clearly a key value store. Right. I suspect we're overthinking it. We just need a few examples. Really. Just to get going and then they can free form type the rest. Okay. All right. From my point of view, I had a quick update. I was prepping the white paper to publish it onto the, the storage working group could help page similar to what the serverless working group did. It was a bit difficult to convert the Google doc and to mark them. I think I'm just about done now. So I'll get that PR into the year tomorrow. And then we can throw a line under that. Great. Thank you. Thank you. Did, did anyone else want to raise any other things? Or in that case, I think we're done for today. Okay. Thanks everyone. Thanks. Bye. Thanks. Bye guys.