OpenStack Ops Radio Hour - April 24 2026

I think we're on. Um, welcome everyone. This is the OpenStack operators ops radio hour. Um, for anyone who's new, um, think of this as the equivalent of a Zoom meeting where it's an open forum where we get together and talk about issues facing the community running OpenStack. [snorts] Um, I'm Chris Morgan. I work for Bloomberg. Um, I'm the product owner. But, um, this is meant to be a working session, not a presentation. So, feel free to speak up. Uh polite people use the raise your hand feature in um Meepad here or Jity which is based on but um people can also just speak. I I think it's best to do that because I don't always notice the hands up. Um and we do this monthly and we seem to have um a growing um interest in it. Um so this week this is the second meeting on MeetPad where we talk about these issues. Um the first one was um set up by the technical committee OKAC or our opinion for technical committee and there's a a great um set of notes that I've linked um going over related issues because the technical committee you know would like to improve the uh coordination with people who are using the product. Um I can't switch the the document to that pad but I've linked it so if you on your own machine go and look at it. There's a ton of stuff there. I did pull out a couple of talking points here. Um so that's um you know um very common things that we um keep on working on. Um but I the I think the thing that keeps on coming back to me the most is you know um technical committee said to me um how can we get the word out about all the great resources and how can we survey the users um how can we find out what they need, what problems they're having, etc. And um I keep on coming back to this issue. Um I'm hoping everyone can see this. Um the the OpenStack operators aren't a team. It's not a crew. They don't necessarily know each other. And the actual places where people discuss it are all over the place. I put some of the things here. Open stack discuss mailing list, which is official. Reddit, LinkedIn, not official, but uh popular IRC. Um there's an operators list on IRC. [snorts] Um matrix is a suggestion that's come along recently. Um you know it has the properties of being open source and distributed which we like. There's an operators team there. I haven't got the link here but we can put that on. And then some people are using Maston including myself. We used to use Twitter which is now called X. I don't like that anymore for obvious reasons. And then this technology which we're gathered here today is um meat pad ether pad. Um so hopefully people know how this works. I see people put the link. Thank you. Um the matrix link is in the um Jity chat. If you don't see that um at the bottom there's a toolbar and there's a little um icon for a dialogue. Um if you um click on that it says open chat. That's the chat we use in this. Um and then other things um that were mentioned on Monday are things like um are operators trying to contribute when they have fixes? Do they know how? What's it like? And then um a discussion about um the guides available to new OpenStack um operators, things like architecture guide and high availability. um years ago there was a discussion about re getting these rewritten and people were very keen but then it didn't happen because people just change roles so anyway that's the reason I started with that is because that feels like an ongoing conversation from Monday so um if there are things there that strike your interest >> you know let's talk about them during this hour uh but the first I see more people joining the first thing that uh was on the agenda for this week uh is OpenStack observability practices um So is the person who put this on here um able to uh are they here and if so can they um share with us their their questions or their what they are trying to get out of the discussion. [snorts] I'm putting down some questions. I hope everyone can see the ether pad that this video window has an embedded um ether pad which is a shared document and you can type right into it in this video window but it's also um it's linked at the top here um if you want to open it separately um elder code do you know who added that uh item >> observability was that from last week last month I should say that that that's what I that's what I was um explaining at the very beginning that the observability is coming from last week or last week last meeting sorry >> last month yeah um okay so it's possible that the person who um wanted to talk about it hasn't been able to make this meeting um I could talk very briefly about what we do so we have um you know mix of cloud and non-cloud servers in our back end and there's a corporatewide uh metrics portal and you know website with dashboards alerting infrastructure. So we just plug our OpenStack into that. So we send all our metrics over there and it you know will send us alerts based on conditions that we set. Um I think we're using um you know Graphana for um dashboards. Um it's absolutely actually not my area so I really can't share any great info about that. The person who wanted to talk about it isn't here either. We maybe unless other people want to share what they what they're doing. Maybe maybe move on to other topics. Anyone want to comment? As I mentioned, this is meant to be um a discussion. It's not a presentation. It's not the Chris show. I wouldn't do it if it was that. Um anyone want to volunteer what they're doing or what they tried and didn't work? Uh we have a lot of people here. No one wants to speak up. Everyone's muted apart from me. [snorts] Okay. Um then I think we should probably move on. So um one thing we can talk about for sure is um where are you um starting with your you know communication about um working with OpenStack. I put in some um options here. There may be many others as well. Um could I ask everyone on the call to plus one the things that they do use or add more items because um we struggle with actually finding out how to reach the people who might benefit from this and might like to discuss. Um so for instance not sure if anyone's getting what I'm talking about here. So I've plus one the um open sack discuss mailing list because I'm I subscribe to that. I read some topics on that. I can't read all of it. Um and I'm plus one um master funky. Go ahead. >> I was just gonna note that this is important feedback um also from a marketing perspective. I I know we talked about this uh in the the session at the beginning of the week and then I brought it up in the couple of open for foundation marketing sessions that we held um throughout the week. And so we're going to try to find ways to bring these discussion channels and uh information sources to the attention of more operators. Um and knowing where to reach them and where to be able to get their attention is is a big part of of being able to do that successfully. >> Thanks Fungi. Um I uh I agree. um you know the the things that people are using are where we'll try and get the word out about these sessions. Um for instance, the timing is not always set a month in advance. So sometimes we have to work out the date a bit later and then uh you know let the community know and it's hard to know where to send that message. Um it looks like the mailing list is um the winner um from this brief survey. I realize this is this is just a sprinkling of people. No one said Reddit. Only one person said LinkedIn. Um, I was following something on LinkedIn and it just disappeared mysteriously and then it seems to be back or that's that, you know, was um a mystery to me at least. Um, IRC's got a similar number of plus ones. Uh, Pierre, go ahead. >> Yeah, just agreeing with you on LinkedIn. It's very I open it once in a while, but it's very algorithm based. So, sometimes you see the information, sometimes you don't. Uh it's not great for >> it's very accidental discovery let's say. >> Yeah. Um so Reddit's got no um no recommendations. Should we say LinkedIn is a bit funny. IRC [snorts] I'm sure some of you heard me say this. Um my company takes great pains to ban IRC. Not only can I not do IRC, they don't like the IRC cloud website either. It puts my laptop into a lockdown mode. So, it's hilarious that that's part of my job is to use this thing that the company computers fight me on. [snorts] Uh, matrix. Um, someone's put the link there. That's a recent thing that we are trying to standardize on. Um, I'm not very familiar with it myself. I do log on and I can get on and do things. Um, so the subplus was there. I I'm plus one that yes. Um, I said I'll try and do that masttodon. Um, so I like that because it's um, Twitter-like, open source, but it's also federated, so it's resistant to attack and ownership changes. Um, but to be honest, it hasn't really taken off for this kind of thing. I I'm actively campaigning against the use of Twitter both at my company and for the open infra organization um because of politics basically. Um, so I put minus infinity on that. This is just kind of a little joke. Meet pad, ether pad. Um, yeah, that they um they're good when you want to talk about a thing and you're already going to meet, but like it's very hard to swing by in a blank meat pad and find people to discuss topics. Ops radio are lots of plus ones there. Thank you. Um, this is meant to just be a community thing that kind of becomes self- sustaining. Um, I will say that, um, we're low on moderators now. I consider myself a moderator, not a presenter. Um, Eric McCormack was a veteran of the ops meetups thing, but he's recently changed roles and can't really make most of these things. And there's plus one for Blue Sky. I haven't done Blue Sky because although it seems nice now and it's very Twitter-like, it's basically owned by um a corporation. It's not actually federated and it could be taken over just the same exact way as Twitter. Um, so that's just a personal view of course. Um, but there's only one plus one there. Does anyone want to um, you know, advocate for any of the uh, options here more strongly? I mean, you know, this is one of the things that's that's kind of challenging us is to literally maybe maybe agree on a shorter list and use those and and, you know, stop trying on other things. Maybe maybe we can agree LinkedIn isn't great. Maybe we can agree that Twitter isn't great. Um, anyone want to advocate for against any of these things? I really want to emphasize this is meant to be um a discussion not a presentation. If I just speak for an hour I feel like I failed. Go ahead Pierre. >> So what what would be interesting is maybe to get some data I don't know from the foundation maybe they are able to have some visibility on the the reach uh to uh to the community from each of those um those channels. I was just looking at for example uh Masttodon I'm I'm on both Masttodon and and Blue Sky uh Maston for the OpenStack page well page profile whatever it's called it has 452 subscribers and on blue sky it has a bit more than a thousand uh even though I think the the blue sky one is a lot more recent That's fair. >> But I I don't know if OpenStack is still on on X. I've I've not X. >> So, um I gave it up a long time ago. I had a personal profile which I deleted. There was an OpenStack operators profile which I also deleted. >> And uh so I'm on I'm on the board and I said >> yeah I said you should you should not use this now. Quite honestly, some responses were privately that people said, "Thanks for saying that." But other responses were, "I I didn't know there was a problem. What what is what what are you talking about?" So that that surprised me. >> Um >> yeah, well, it looks like the OpenStack profile page is not being updated anymore on X. Anyway, it's still there and it has 146,000 subscribers. Don't know how many of those are bots. Uh yeah, probably most of them, [laughter] >> but it's not posting anything uh anymore. So yeah, that's that's not uh that's not kept to date. >> I heard that um there would be a look from foundation staffers at um the continued presence on Twitter both from you know from first principles, from policy, from morality even, but [clears throat] also just uh engagement. Is it do they get real engagement? There are a lot of bots there. I think there are a lot of bots on boost cloud as well. Um in other areas of discussion away from cloud open sack um I have seen people say that they have fewer followers on masttodon but they seem to be real people >> and not bots. So they actually get more engagement that they value should we say. Um so I think we're I think we are seeing the mailing list um kind of coming back. The the reason I say that is um there was there was a time when we were still trying to organize in-person meetings for you know operators to discuss these things like we used to do for years and um the people who were interested weren't on the mailing list at all. They weren't subscribed. They didn't go and look at the um the mailing list archive. So, um it just seemed like um that mail the mailing list was too much of a fire hose for someone who's sort of dipping a toe into OpenStack to commit to. But, uh there's the the most plus ones uh from this survey on that. So, I think that's a that's a good sign. Um oh yeah, that reminds me. I just see Jeremy or Fungi is his handle has put his name on the attendees list. Go ahead and put your name on the attendees list if um if you wish. It helps us understand who wrote what because the the color coding is you know matched between um your attendance uh record there and then your contributions on the ether pad. So um you know as someone who tries to run meetings to achieve results I think I'll personally say that I will make more of an effort on the mailing list. That seems to be um a troublefree popular item. And Jeremy's actually sharing something that I did see on the other Ether pad that I linked to and I mentioned. Um, you can consume it more as a forum now if you wish. It doesn't have to um, you know, fill your inbox of your personal email or any email. Um, do you want to speak a little bit more about that? I'm not sure I quite understand how that works, Jeremy. >> Uh, yeah, sure. So, um, essentially, um, if if you create, uh, a login on the list.openstack.org website, um, and you're looking at the web archive of the mailing list, um, there's just a little button, uh, link, whatever, underneath each post that you're looking at where you can reply, and it'll open up an editor in the in the browser window there where you can post a a reply on that thread from, or you can start a new thread. um from the from the main archive view of the site. Um so basically like posting on other like web browser oriented discussion forums. That's really cool. I I never actually done that. I get it into my Gmail and it's a lot um so of course I filter it into a different tab and then I tend not to see it too often. Um so that actually might work better for me. So that's that's really cool. Um I wonder if we could actually capture that. I can literally go here do this. Jeremy, correct me when I get this wrong. Is that reasonable? The the best practice in these things is to actually try and get um cool information that people share written down and then these ether pads become valuable as a reference in future. Um, >> yeah, that that looks good. Um, I I included the link I included the link to the archive as well. So, but basically, if you're logged in there, there's a button at the top right that says start a new thread. If you want to like start an entirely new discussion on the list, or if you're looking at an existing thread, you can there's a little reply link below each post, you can click to to follow up to that specifically. Okay. Um, so that's um that's a pretty strong vote I think for good old mail or use consuming it via the web. Um, that does work for me and on my corporate communication laptop. [snorts] Um, IRC still popular, matrix popular. Um, I think the other things are, you know, losing the race. Um, any any more comments on Oh, there's actually even more. >> I want on the on the main list. >> Go ahead. >> Please, if you're posting on it, please make sure to follow the the standard >> uh of of using tags to um so a word a keyword with or multiple keywords between brackets. square brackets. Um, and for example, I I personally use that to um move or keep rather some of the important stuff uh at least for me in my inbox and then other things are going into a folder which I browse on a less less recon basis. So, I've got a a mail filter that will keep um the the messages about the projects I'm involved with. >> That's a great point, Pierre. I've tried to capture that down. Um so, for if those are not familiar, it is a mailing list, but and it's just email, but if you put tags um as Pierre has set up, you can then basically only see certain topics right into your primary mailbox. Um, so you can kind of filter pre-filter it to the things you're interested in and then filter out things that you're not interested in. Um, so, um, you know, if everyone does that, it makes the experience of the mailing list being an overwhelming fire hose less so. Um, and since we have seemingly, um, good consensus to agree to kind of get back to the mailing list, um, that's definitely a good reminder. Okay. Um at about the same time Pierre said that I mentioned that um there are some subcommunities of the OpenStack operators who use Slack and uh we did um discuss that last year on Ops Radio and other places you know because it has um good software generally I mean it's it's it's rich you know it's it's it works on my phone and my laptop and my iPad and these kind of things. Um, but it turns out that the free teams kind of suck now because uh the content gets deleted quite quickly and if you don't want to experience that to pay for a team on Slack is non-trivial money and there are some um very big um open- source foundations finding that a bit too much. So, um I'm aware that uh some people would just like the Slack experience, but um it goes against what we're trying to do to have, you know, great discussions and then 3 months later it says, you know, you can't see this unless you pay us. Um so that was that was a bit of a shame. Anything else or any other outlets that people find useful for discussing um you know, operating oversee? you you're welcome to just put them, you know, just type into this ether pad yourself. Um, I see someone is making a comment about matrix. I think that's Jeremy. Okay. So, he's making he's making a comment there. Um, yes, this is it's Matrix is Slack like. So, there's a recommendation for you. If you're used to Slack, you'll probably get going easily with Matrix. Also, there's um a wealth of open source clients for Matrix. There's really many many ones, you know. Um literally, if you want one that has um you know, cute cat icons, I think there's one for you. Um so, um yeah, that's that's something I'll also keep trying to remember to open on my laptop and be available and maybe we can get that to take off. Um okay so another thing that was um brought up on Monday at the technical committee operators uh session was uh contributing um some operators say that they their operate their contributions get ignored. Uh I shared some experience that we have um with that. Um I think we worked out why ours were being ignored. Um the trouble is that if you just turn up and no one knows your name and you go straight from never having contributed to talking about a very top very technical issue, nobody knows who you are. Um and you kind of haven't given the community anything to justify um asking for attention. Um so I and excuse the language I call this the who the [ __ ] are you problem and we realize that we have that problem and we um we have been actually joining meetings um trying to reach out to the community and actually build consensus around change rather than just hi you don't know me here's my change proposal um so we do have some things landing uh in OpenStack and in Calico which is our neutron plugin um but we had to actually, you know, uh, get our head out of our asses, basically, excuse my coarse language, and actually reach out and do the do the thing properly. Um, but what do people think? Like, I'm I'm really trying to get you all to speak. Has anyone tried contributing and found it um off-putting, unproductive, or they don't even know where to start? Is everyone just uh finding the software perfect or are you just doing patches locally or what's what's going on there? [snorts] I'll I'll be frank, we have um some things that we're trying to upstream um that are being welcomed and other things that we like to upstream that are not being welcomed and so we're actually having to uh carry some patches oursel and then carry on the conversation. Um in particular we're trying to um introduce um unifor uniform quality of service management um which basically means limiting the resource usage of VMs >> to a fair slice to give everybody um predictable managed you know behavior performance um and we've actually found um Nova community less welcoming to that than um the other subsystems like um Cinder Neutron. Um, but we're still working on it because um it's critical for us to not have uh greedy applications um draining uh neighboring VMs of performance. Um so it it is it is difficult. Um I see Jeremy uh sharing some updates from inside the foundation. Um, another thing that was difficult and I think was changed was there used to be really excessive levels of pedentry. Um, you know, as an example, many years ago, we tried to submit a fix for a crash on a null pointer. And, um, the request was, well, you need to add it to the documentation and provide a bunch of unit tests. And the person who'd found a fix for null pointer was just like, never mind then. Um, and I think that there was at one of the foundation meetings I went to, not foundation, one of the summits I went to, there was somebody saying that, yeah, we maybe let that go a bit too much back when there was a bazillion contributors. Um, that was common and now, you know, we actually have to value new contributors. So, we maybe have to be um a little bit more flexible, should we say. Um so um you do need to kind of raise your game and join um the relevant communities and become known and have some um proof of contribution before you can expect people to care. But I think that there's also a little bit of movement from the insiders understanding that a new new contributor maybe needs some time to settle. >> Uh G Coper, go right ahead. >> Yeah, thanks. I think it's ties into the first first question a little bit on the you know communication >> [clears throat] >> uh tools that are being used. I think it's a big uh slowdown for for a new contributor to have to join like IRC to meet people to be even considered uh uh you know uh enough contributor to patch a oneliner here and there. There's also it's going to be depend on projects of course and there are more active projects and uh with bigger communities that uh work in a different way and OpenStack is very diverse but there's also an issue with a lot of untriage bags in less than popular uh OpenStack projects. I've I know I've opened some and they haven't been touched or seen by a human for uh multiple months. So unless somebody's actively joining already an active member of community can join the IRC and poke the right people uh he's not uh getting in easily. >> Uh it's you know it's it's great feedback. Um I will say that um my colleagues who are upstreaming um are definitely not joining IRC directly. Um maybe they're getting maybe the website is working for them. Um so I don't I don't know um if that's the the mandatory part, but definitely you have to find out those teams preferred ways and join them somewhere, right? The mailing list, IRC. Um I think some of them maybe do use Meepad these days or something else similar. Um and of course the the the events themselves like um uh open stack days open in for summits um some other regional things um to actually first of all just meet the people who are are active I think is a great icebreaker um I I talked about not having you know reputation um but literally even if someone is just someone that you've met I think it makes it um more likely that you'll consider their their patch or their request. Um so um I do recommend the inerson events if you can. Um a lot of them are no longer in North America for obvious reasons so I get to much fewer of them. [snorts] But maybe we can do something maybe we can you know invent something new um a bit like ops radio. I mean, this isn't, you know, one hour a month isn't going to do that all that much, but um I mean, maybe there should be a, you know, a a Nova team uh icebreaker on MeatPad regularly and people can maybe talk about something that they're thinking about because one of the things that we find is it's good to talk with the maintainers before doing a lot of work on a change proposal because sometimes they just say, you know, we we're against change in that direction because we don't I think it's um the responsibility of our subsystem to be doing that. It should be done in so and so layer and that can really avoid a lot of wasted work. Um I'm just thinking aloud here, you know, maybe because it seems like anyone who's interested can can can make a I keep on calling it a Zoom meeting. It's not technically Zoom, but you know what I mean. uh a video meeting with with chat and um u you know reactions and and those kinds of things. So I'm just um thinking aloud that maybe that would help. I mean I don't know what the person G G Cobra I don't know what your first name is. Do you think that might help if you could actually just maybe spend even just half an hour meet the team and talk about it before going to electronic written communication like IRC or mailing list? I'm sure that that that can be one thing to that can help. I was kind of thinking of different forms of engaging with the community. The the IRC is I mean I I' I'm using it myself and don't get me wrong, I'm not talking about my personal issues here. I've been a contributor for a while, but it's a it's more of a those are kind of a big steps to go through for somebody completely new uh to install additional software to know which projects are meeting where uh to to then join the the respective place and try to engage with the community. I think some something easier either it needs to be better documented in the contributing guide per project or it needs to be a little or or some some other software needs to be used there to uh to facilitate the conversations because it's daunting for most of the uh you know >> uh engineers to to meet up in person and start talking about their issues. >> I agree. Um so um there is some great information from uh Jeremy and you on my screen it's in black. I don't know if it's the same for everybody. He says community managers are trying to identify and smooth over contribution challenges under the name bridging the gap. Um there's some links there. Um I think at the in the old days when we had two summits a year there was a there was a track or at least sessions about um mastering those those challenges of the software required. you know, it's not um GitHub, it's um some other thing. And they're not PRs, there's other thing. And uh you know, if you want to um build the docs, you have to have certain tool chain. Um was that called upstream academy? My name my memory is failing me this morning. Only had one coffee. It's terrible. Um but of course the found the um summits are um few and far between now. Um >> we are not we are not running upstream institute uh or the community hasn't been running the upstream institute um at the summits for for some time now but the um the the contributor guide is available and if people want to run the training again >> um that would also be um a possibility maybe but we need to always look into space and things and uh Jeremy your hand is up. Go ahead. >> Uh yeah, just related to that while we haven't been doing upstream institute um there have been sponsored sessions at most of the summits um for basic Git and Garrett introductions. Um Amy ran one at the Paris Summit. Um I think she's trying to decide if it's going to be worthwhile to try to get one incorporated into the Shanghai summit or not. So if this is something that folks want to do or want to have um let us know added to the ether bed here or um reach out on on OpenStack discuss. That would be um that would be a great uh great way to uh to let Amy and other folks know that people are interested. You can also hop into the the Ops matrix channel. So just share it anywhere when it come comes to mind that that it would be useful. Uh, so to the person who raised this, G. Coper, I'm sorry I don't know your name. Um, I see some comments here from not just Jeremy but also Gam who's on the technical committee or leads the technical committee. They're going to reconsider restarting Upstream Institute. So Upstream Institute was um um, you know, base camp training to get up to speed with everything you need to know so that you could contribute a patch to OpenStack. Um obviously super important because the project has had a lot of churn. People who mastered all that have moved on. Other people coming in their organizations would like to get going with OpenStack because of you know uh more recent pressures. There's um digital sovereignty. There's um proprietary virtualization vendor price rises. I try not to speak vendor names. I'm not really supposed to. And then um another thing that we do um or we talk about is um just just basic economics um believe it or not we have enough scale that it's actually cheaper for us to run private cloud for certain things than to use a AWS. Um you do need to be at a certain scale to to make sense but you know um I have heard conversations about repatriation which means apps need to come off private cloud and into sorry off public cloud into private cloud because their spend is enormous. Um so um maybe we need to um focus on this to actually um make it easier for the complaints that um have been raised to go away and to people to feel that it's easy to get started. Uh there's some there's some great uh commentary here from from Jeremy as well. Um it's actually not easy to get your code landed anywhere I think is what he's what he's saying. um the person who um worked for me at the time who couldn't get the most trivial patch land on OpenStack did have things landed on other projects but probably not as big and busy as OpenStack at the time but so I I kind of agree with Jeremy's point but um I definitely saw someone um give a quick try and give up and they were an able developer um and I'd love to fix that. Uh Doug, you got your hand up? Go ahead. I think you're on mute. >> Yep, sure was. Um well I I was I was also going to say that um part of the part of the problem is also um some of the project attitudes and some of the maintainer attitudes um towards contributors and and I don't I mean I don't know how to fix that but I mean that is a that is a big problem in some projects. So, some projects will tell you that the cores don't have enough cycles to review things and um to consider things that aren't on the priority list. Um but it's impossible to get new cores into the project because nobody can contribute anything and be engaged. So, they kind of are death spiraling themselves. Yeah. In a way >> that's an interesting point. Um I have heard that um there's a universal pressure on you know people active contributors. Um but I think it's not often mentioned as um you know that's maybe the source of the problem as well as you know um the consequence of the problem. Um as you say death spiral it feeds on itself. Um, so it does seem like really important to uh make it easier to get uh, you know, new recruits, should we say? >> Go ahead. >> And I mean, I I can I can name names as far as projects go, not not people, but as far as projects go, um, that that we've got. And um I mean I've got I've got people that work for me at at Rackspace that are like ready, willing, and able to get engaged in some of these projects, but I can't get traction from the existing course and I'm a TC member, right? Like to to make it happen. So, I can't imagine how daunting it must be for a a one-time contributor or a new person um to encounter that. >> Yeah, that's great feedback. Um I think it's probably more for your fellow TC members than for >> Yeah, sure. >> certainly myself. Um but I think it's great to talk about. One of the things that we always did in the um operator meetups in person was trying to say, hey, it's a safe space. you can actually share your frustrations, share your challenges. Um, frustration with having the patch to a to a bug and not be able to land it would be very valid. Another classic one was I'm nine releases of OpenStack behind and I feel kind of embarrassed and I don't want to even talk about it and we would compare notes and we'd find the person who had you know a deployment that was 14 releases. Um, but this but this is serious business. This is sort of like ex existential for the project. if uh if the people you know owning plus one rights are kind of done uh then the project's done right so um this is kind of a a grim story but um as I mentioned with this ops radio thing this is a this is a ground zero rewrite like um we had this whole plan of two two meetups a year in person two days long and then just couldn't sustain it anymore maybe we can do something as cheap and easy and cheerful as as this thing for um some ice breaking and some um consensus building or otherwise. I mean, one of the problems, as I mentioned, is I, you know, let's say I have a patch. If the developers just don't like the direction, they maybe don't even respond, right? I I have heard that some people say like, >> "We've been over this a million times. Nova is not going to be in the business of of that." Um, and I think it would it would think even getting an answer like, "We don't like your direction would probably seem a little bit less unfriendly than radio science." Somebody's It sounds like someone's boiling a kettle. I don't know what that is exactly, but uh >> um >> my bad that did not I thought noise cancelling was working. >> I am. >> It's okay. >> Heating up water. I'll mute myself again. [laughter] [snorts] Anyway, um I think this is something to think about and maybe if people have suggestions we could put it on the agenda for next time, which is uh and I think Gam, you have your answer that there's there's tremendous challenges to get going and it's not just I don't know where the link is to tell me how to set up my tools, but a people problem. Um the uh the core devs are overwhelmed and not even keeping up with enough sort of welcoming feedback to even keep their communities from shrinking further. Uh Gam's got a comment about the noise cancelling. Yeah. Um I I'll be frank, I'm a little deaf, so uh background noise is difficult for me. It's why I have to have the fancy headphones. Um, anyone else want to talk about their challenges or successes with contributing upstream and where they are and you know, are you somebody who has patches that you'd love to contribute but you're you're feeling burned or is it something that you're not even at yet? Not not at the level u Okay, I think the room is pretty dead on that. Um again I'm going to emphasize that this is meant to be a discussion not a presentation but I will since I seem to be running this meeting or as a presentation. The next thing is that I pulled out of the thing on Monday um the getting started guide. So there are a couple of um operator targeted documents that were going to be rewritten by people with eminent qualifications. there's um an architecture guide and a a high availability guide for OpenStack. Um and then those people just no fault of their own just found themselves doing something else. Uh so my question is um are the people who are coming in and keen to kind of do OpenStack fairly freshly in need of such a thing or is are they finding that they can just get the information from other places? In other words, are those documents still in need of rewrites or is it kind of missed the opportunity? And if people don't know what they are, the link the link to the other ether pad is on today's agenda and the links to those docs or at least the document org within OpenStack are there. I know that's very indirect. Yeah, operations guide. Thank you. Um because uh it's terribly dated material at this point. Thank you. Thank you for the for the links. That's great. [snorts] Um, if people even know about these things, would they would they use them if they were somewhat fresh, >> Graeme? Go ahead. >> So, I'll be honest. This is the first time I've actually even looked at those documentations. um four years using using OpenStack and never really even looked at those uh essentially. So my my comments in in that is that yeah it it's probably not well known to a lot of people coming as as a starter guide uh where to find or or even start looking for those. So just just as a my experience >> so um thank you for that. Um, one of the things I didn't pull out of Monday's meeting into these meeting minutes, but was um, conveyed very well um, was that um, there are actually resources uh, that it would be great for new OpenStack operators that are hard to find. And um, one of the things I volunteered for, I haven't had a chance to do it this week, was to actually um, try and make a landing page that links to more of those. So there's an OpenStack operators page on the OpenStack wiki which I happen to have edit rights to and I'm going to do that and I'll try and pull these things together and then I'll try and put them on here and then I think um every time we do one of these and people may come along for the first time we'll say hey here's a single page that links to resources for getting started with building your OpenStack um with contributing to the OpenStack software project um and someone else's linked security guide I think the first thing to do is to actually have people find those docs and have some opinion about them um before we you know make any commitment to try and rewrite them um again. Um Jeremy's now sharing um more info here. Um, as I mentioned, I can't actually switch the shared document that we are looking at to something else, but the OpenStack operators page on the wiki is um easy to find and um um probably should be expanded to link to more of these things and um we can maybe get people to say, you know, I'd love to do OpenStack, but I have no idea about the architecture and what people are doing. Maybe we can start there. Or maybe like I've got it running but I have no idea how to operate it. You know, day two problems, upgrades, machine maintenance. Um so um on that particular thing there is a bit of a plan. Um I think someone running OpenSack for four years and not knowing those documents exist is kind of telling. Uh so comments some comments in the in the MeatPad chat here. Um Pierre says he looks at the project docs rather than these guides. Um I think the project docs are are very good on specific things but like um your overall architecture of your cloud is kind of a big decision and it's hard to know where to start. Um like in some cases for instance which flavor of a subsystem you should use is very hard to determine. And I'm thinking Neutron for instance. Neutron has uh plugins. It has competing things. There's OBS, OVN, Linux Bridge, Calico. Um I don't know if there's a page that says which of those things are alive today and should be considered and which things are dead. There are Neutron implementations that were basically from vendors that have come and gone. Um yet more context from uh >> Jeremy. There used to be a team that did this. Yes, >> I should say I um I look at the project docs mostly because I when I need to configure some niche features and things like that and for the overall architecture I mean I I'm a developer on the the collab project. So we have some conventions there on what what's supported. We we tend not to support every single um historical back end of each project but more go with the the things that we know work and have have been well tested. So I use that in combination with the the cola docs actually have some general guidelines and and yeah general documentation about how to set up a a cloud. It kind of overlaps with the operations guide that is it's cola specific. >> Understood. Um, I mean, one option I think I'm certainly considering is, uh, instead of having an operations guide and an architecture guide as docs, they should probably bow out and just say, "Here are the best, um, documents in this space. Go look at these." Um, I think they are so old that they were written when there was nothing. Um, so someone put together a document and said, "This is how you can actually make this OpenStack software do something in your data center." And of course, we're we're well past that now. I think someone else had their hand up and >> yeah, we have two folks. Go ahead. >> I'm not sure who was first. >> Um, I'll just go top to bottom. G Coper, please go ahead. You're on mute. Go ahead. >> Uh, cheers. Uh, it's Greg, by the way. Uh, uh, yeah. What I was wanting to, it's kind of an off topic, but we're still on the docs subject. So have you all of you noticed when you try to Google for OpenStack documentation you usually get uh I don't know wabi release docs as first results instead of the latest one. Is anyone actively looking at positioning those things in Google or archiving older documentation because there's a lot of change stuff and confusion that can come out of it. I certainly do um have that experience that I Google and I get some random old release that's no longer maintained and I don't know what the fixes. I will say that um retiring old docs has to be considered very carefully because um there are many um deployments that are on old open stack versions and still need those docs to be there. Um we were on um an version of OpenStack where the documents were deleted. Um so um I think that that's a problem about um search optimization partially which I I don't do. Um what I do is um I Google for my topic and then I land on some page as you say walabe and then I kind of have a rough idea where it might be in the up-to-date documentation or at least the documentation for my current version of OpenStack and I drill down until I find the the sort of corresponding documentation to the software I'll be running which is a very cumbersome process. So you make a very good point Greg. Thank you. Uh another hand up is Graeme who's been extremely patient. Um, >> um, I just want to get back to the um, on on the doc doc side and the the operational guide documentation is is that because there's so many different ways of deploying OpenStack, it's really hard to say here's one general documentation that describes the environment, the different ways of deployment. And every every deployment has a guide of this is how you operate it. This is how you use it. This is how you maintain, upgrade and do your process in that way. So it kind of feels like it it is an old document that was was created a long time ago when there weren't that many deployment tools. And now you've got cola, you've got triple O, you've got all these other other different deployment styles. um they've written their own documentations around how how these deployments work. And then on an architectural side of it is you you kind of still need to have your your basics of how everything should fit together in in the data center world and how networking works and all of that that sort of thing. you you should be coming in with that sort of knowledge to understand that from the deployment documentation on the operational side these are the things that you need as an architectural side of things I know that's probably not a uh everyone would agree with that but it it is something like you need that background knowledge to be able to to come into open stack with >> great points Graeme um I I think that what you just said supported in my mind um a suggestion I'm kind of forming which is that the operations guide maybe shouldn't shouldn't be a thing that we're trying to maintain shouldn't exist anymore. I I said just a minute ago maybe it should should stand down and it should actually instead you know we should say there's no such thing as a single bite to operate in your cloud. It depends on how you build it. You know here are the uh the live the live competing ways of building it. you know, Kala, triple O, KOB, uh, OpenStack, Anible, many more. Um, you know, go there because those projects maintain their documents and, um, you know, if you come across, um, something saying how to maintain your cloud using some other thing, it may have been dead for 10 years um or five years. I mean, um, at one point air airship was very exciting, right? And then it just stopped dead. Um, I think it was, and if I'm offending anyone, please I'm sorry, but I think it was an AT&T project and then they they just decided to go public cloud and um the end, right? So, the airship stuff is still around, but >> the team >> they they basically outsourced it to Microsoft, moved the team to Microsoft, and then Microsoft wound it down. >> There you go. >> I still use it, but >> yeah, >> it's like maintenance mode. That's the problem and that's what's hoping holding back OpenStack Helm is that um they don't want to break the behavior of airship. >> There you go. That's some great insidefo. I I always wondered about that. I looked at air airship and then I was at a summit and they said, "Oh, well, airship 2 will be, you know, um all different but much better. Watch the space." and then Airship 2 started and then it just slammed into a wall uh without much public information about hey what's going on here. Um so there you have it. Um so anyway um if you think that the operations guide as a concept is is dead uh plus you plus one the thing I put here saying it be retired um if we have strong consensus that it would be much more useful to just point to things in that space that are specific to technologies that are still alive. Um I could actually make that happen. Just delete the doc. I see funky left the meeting. We're actually out of time folks. Wow, that that the last bit went quick. Um, so I'll say thanks very much and um we'll work out a time and notify on the popular >> [clears throat] >> um outlets for next next month. >> The next last Friday in May is May 29. Is anyone violently opposing to that one? >> Uh I don't have my calendar on my screen, not logged into my thing. So, uh, May 29, unless we say differently, and we will make every effort to get it out to people if that's the case. So, I'll pencil that in. Dildo. >> Okay. I'll I'll sync up with you afterward to make sure that it works. Thank you all. >> Okay. So, end of meeting, we actually ran over. I always try and end more smoothly, but I guess we were busy. So, thanks everyone. See you next time.