From a263af7519252fb7106c9d112f67ca537609f4fd Mon Sep 17 00:00:00 2001 From: George Appiah <48031636+iamGeorgePro@users.noreply.github.com> Date: Wed, 22 Jan 2025 18:51:42 +0000 Subject: [PATCH 1/3] Create meeting notes for sig-support #40 Signed-off-by: George Appiah <48031636+iamGeorgePro@users.noreply.github.com> --- sig-support/meetings/040-2024-1-08.md | 544 ++++++++++++++++++++++++++ 1 file changed, 544 insertions(+) create mode 100644 sig-support/meetings/040-2024-1-08.md diff --git a/sig-support/meetings/040-2024-1-08.md b/sig-support/meetings/040-2024-1-08.md new file mode 100644 index 00000000..e0418af4 --- /dev/null +++ b/sig-support/meetings/040-2024-1-08.md @@ -0,0 +1,544 @@ +# Akash Network - Support Special Interest Group (SIG) - Meeting #40 + +## Agenda +- Issues Awaiting Triage in the Support Repository +- Improvements to SDL Documentation and Automation +- Community Moderation and Discord Support +- Open Floor for Additional Feedback and Discussion + +## Meeting Details + +- Date: January 8, 2025 +- Time: 07:00 GMT-7 +- Recording: Coming Soon +- [Transcript](#transcript) + +## Participants +- Tyler Wright +- Scott Carruthers +- Amanda Keay +- Andrew Gnatyuk +- Andrey Arapov +- Benjamin B +- Damir Simpovic +- Deathless +- Derrick Hammer +- Dimokus +- Joao Luna +- Pavlo Dereniuk +- Rodri R +- Zeke Ezagui + + +## Meeting Notes + +### Opening Remarks +- Tyler Wright welcomed participants and provided an overview of the meeting’s purpose: discussing current support repository issues, community feedback, and ongoing improvements within the Akash ecosystem. +- Emphasized collaboration between core engineering, community members, and Akash Insiders for resolving user issues and enhancing the platform's usability. + +### Issues Awaiting Triage in the Support Repository +- Scott Carruthers led a review of support repository issues. + - Discovered a default assignment of the "testnet 5" label to new issues. Tyler Wright and Derrick Hammer confirmed this behavior, and Scott committed to removing incorrect labels in real-time. + - Reviewed a longstanding issue involving NVIDIA container runtime and secret management for K3S providers. Andrey Arapov highlighted internal documentation updates but noted the need for more comprehensive public documentation. + - Discussed a bug causing misrouting of custom hostnames and ports within pods. Derrick Hammer described workarounds and emphasized the need for fixes to Kubernetes ingress mapping. + - Addressed a manifest update error that led to pod crashes. Derrick Hammer shared his temporary workaround of introducing redundancy into his scripts, and Scott committed to investigating a long-term solution. + +### Improvements to SDL Documentation and Automation +- Derrick Hammer raised concerns about the lack of comprehensive SDL documentation and inconsistent API specifications. + - Suggested creating a single JSON schema as the source of truth for SDL specifications, enabling automated updates to API documentation and minimizing ambiguities. + - Scott Carruthers acknowledged the need for improvements and noted this feedback for internal discussion. + +### Community Moderation and Discord Support +- Participants discussed enhancing moderation and support in community channels: + - Derrick Hammer emphasized the importance of clear roles for moderators, suggesting a "Moderator" role alongside or instead of "Vanguard" for better recognition by new users. + - Benjamin B cautioned against overly broad permissions but supported the idea of a taggable, invisible moderator role for handling spam and abuse. + - Tyler Wright committed to discussing these suggestions with Adam W., the community lead, and ensuring actionable outcomes. + +### Open Floor for Additional Feedback and Discussion +- Derrick Hammer shared challenges with scaling and modifying existing leases without creating new deployments, calling for more flexibility in the Akash platform to align with traditional cloud services. +- Benjamin B summarized these concerns, highlighting the need for lease renegotiation and dynamic resource allocation. +- Derrick Hammer recommended reordering meeting notes in the support repository to display the most recent meetings first for better accessibility and clarity. + +### Wrap-up +- Tyler Wright thanked participants for their contributions and outlined next steps to address the discussed issues and suggestions. +- Encouraged ongoing collaboration via Discord and GitHub for feedback and support. + +## Action Items +- **Scott Carruthers:** Resolve the default "testnet 5" label assignment and follow up on ingress mapping and pod crash issues. +- **Andrey Arapov:** Enhance public documentation for K3S-based providers and NVIDIA container runtime settings. +- **Core Team:** Evaluate and implement improvements to SDL documentation and automation. +- **Tyler Wright:** Discuss community moderation enhancements with Adam W. and ensure improved clarity for new users in Discord. +- **Community Team:** Reorganize the support repository to display the latest meeting notes first for improved accessibility. + + +## **Transcript** + +Tyler Wright: Hey everybody, welcome to the first SIG monthly meeting of 2025. It's January 8th, 2025. The special interest group for support meets monthly to discuss all things related to the inside of the Akash network organization. the support repo is where all issues related to the core codebase live. Again, there are a number of Akash insiders and Akash me community members inside of Discord that give almost 24 by7 responses to those that are looking to use the Akash network. We have a group called the Vanguards that leads those efforts. + +Tyler Wright: If there are any issues that are not being or not solved by insiders, Vanguards or other members of the community, there isn't a escalation path for all issues that may be related to the the console, etc. there are again separate repos for issues related to the core codebase that go in the support repo inside the AOS network organization as well as again another repo for the AOS console where all issues related to console live. Usually during this six support meeting, we have a specific agenda that we go over monthly. Members of the core engineering team mostly led by Scott go through any issues that are awaiting triage. Anybody that wants to discuss any open issues can discuss those at those time. + +Tyler Wright: If anybody wants to get involved in understanding an issue in more detail and contribute, this is a meeting to come to on a monthly basis and then again we can discuss those specific issues inside of GitHub in the comments between meetings or inside the SIG support discord channel. After that, we see if there's any members of the community, Vanguards, Insiders, or otherwise that want to talk about specifically support and Discord, Telegram, or any other channels. If anybody has any other talking points that they want to go over this week as it pertains to feel free to drop those in the chat and we can discuss those. I think people are already starting to put items in there. but without further ado, I will hand it over to Scott to get us going with any issues that may be awaiting triage. + +Scott Carruthers: All right. Thanks, I have a few urgencies going on, so I'm going to try to get through this and I don't think it's going to take too long. So, let me just try to get into this real quick. Okay, so I'm sharing my screen. I think it might still be coming up. Is everyone able to see my screen? + +Tyler Wright: Yes. + +Benjamin B: Yes. + +Scott Carruthers: So as we can see, there's only three issues awaiting triage. It's kind of interesting. I'm wondering why these are all, have the label of test net 5 on them. so, Ben, maybe you could, make comment on this because I know that you just opened this one issue this morning. So I'm wondering if it's just defaulting to assignment of the test net 5 label. could you comment on that? Did you intentionally label this as test net 5 or did it do that automatically? + +Benjamin B: I didn't do it intentionally, but I didn't look that much on the labels. + +Scott Carruthers: So Tyler, that might be something we want to note. And I have a feeling that because it doesn't seem like any of these are really test net 5 related. so maybe we could do a check and see if this does the signing by default. Okay, I see Tyler said and Derek, confirm that it's defaulting. so as we go through these, I'll remove the test net 5 label here in real time and then we'll adjust that so it's no longer the default. So two of these I think we have members of the community that open the issue on the call. + +Scott Carruthers: So we could go through these and get comments from the individuals that opened them. So first up, Andy, you created this issue. so I'm just going to take time to get rid of test net 5. and also just while I'm doing this, I'll take it out of the waiting triage and put it into the provider repo. So Andy, as prep for this call, I was reading through this issue that you opened in late November and we go through the history of this, but you did some testing and I I not the final update, but you said if the workaround or the configuration that you were working on and validating if it works out that we wanted to advise K3's providers to disable + + +### 00:05:00 + +Scott Carruthers: secret management and Nvidia container runtime. Do you have a latest update on this? And it looks like possibly you validated this across providers. Was this added to the K3 provider build documentation? + +Andrey Arapov: Yeah, that's correct. Thank Scott, we just need to addit this to documentation. we have made documentation docs on the darkash network for cube spray based providers, but this is only for K3s and… + +Scott Carruthers: Yeah. Yeah. + +Andrey Arapov: actually I'm not sure I think we had some documentation for K3S, but it could be very tiny and not detailed. So I'm not sure if we need to maybe first put something for K3S because we do it internally for some of the providers and the script se that we are using it's still work in progress so that's why I'm keeping this kind of open for not to forget to basically update it should we add some documentation for K3sbased providers because majority are + +Andrey Arapov: using cube spray. Hence, it's still open. + +Scott Carruthers: Is this now part of the internal build process or added to the scripts or has that just for our own use? + +Andrey Arapov: Maybe it's worse to add linking it to the script. But what I did actually internally I suggested we have the internal doc for the provider builds. I added this step there. + +Scott Carruthers: Yeah. Yeah. + +Andrey Arapov: So basically whenever we build something new I refer it to this ticket basically. but yeah, that I just don't want it to be closed because we can forget. + +Scott Carruthers: So, completely makes sense. I don't think there's any current do items. but when we write documentation for K3S builds in cos.network we can add this. So, understood that you just want to keep this open as a reminder. so yeah with that I don't think that anything is necessary. maybe I'll just assign yourself also just so you see it in your issues. but I don't think there's anything that we need to do at this point. sounds good on that. so now if we look a new at awaiting triage. + +Scott Carruthers: So Ben fiberdub I think you actually made reference to this in the chat for this meet as so first of all I will take this out of five. and we could read through this in real time. I know this has to do with the assep field and AOS STL's. so we could read through this, but do you just want to in your own words kind of summarize this issue and either Andy or myself or others on the call might have some thoughts. + +Benjamin B: I just wanted to say that the links I placed in chat are more like bumping previous issues that has not gotten any responses on the posts since a month or two back. the issue here is just something completely + +Scott Carruthers: Okay. I don't know. if you don't want to. We could obviously read through the issue so I'll put this into the provider repo as a label. do you want to go through describing… + +Scott Carruthers: what the issue is or if not we can just read through the issue async. + +Benjamin B: I think the best thing is… + +Benjamin B: if you just read through it because I don't think I'm going to make it more clear by just speaking of my mind directly. + +Scott Carruthers: So, I read through this quickly before, So, I'm just looking at SDL. so we're using Okay, I'm not sure that I was entirely clear what the issue was But it may be just because I was reading through it quickly before the meeting. and again you don't have to review this in real time but maybe if you just had some thoughts on so you use this in the asept field and I believe you were so what is the issue at that point is are you able to use this asept URL to reach it looks like this react app does it route to that react app properly + + +### 00:10:00 + +Scott Carruthers: All right. + +Benjamin B: Yeah, that roots correctly. + +Derrick Hammer: Can you hear me? + +Benjamin B: But I can hear you. + +Derrick Hammer: Can you? + +Benjamin B: But Do you want to take this over? + +Derrick Hammer: Yeah. yeah. no. Yeah, I'll handle it since I have Yeah, Scott, this is my issue. I was going to let him handle it because I'm not the best communicator and my hand I had this issue recently. and I literally just had to destroy my deployment and do a lot of BS because of it. apparently if you're having two deployments I think custom host name being involved two ports with custom names involved I had port 80 and 8080 and I was expecting them to both work on custom host names somehow I routed both of them to the same port on the same deployment and in my case I was actually running a CS service and S3 bucket service and somehow I was routing the S3 to web + +Derrick Hammer: And so I was getting problems with errors because my S3 service wasn't working. It turned out it was routing to custom host names the exact same port when I was using two and even though it had the two ports going over ingress, it was still routing them to the same port internally when they should have been to two different places. So it's like you can't run two web services all of a sudden on the same system with ingress and yeah So yeah the overall problem is that if it's going to have those two ingress it should have been to two ingresses and it did it correctly but the port binding was wrong and I actually had to manipulate Q8 ingress controller to control the routing. + +Derrick Hammer: Apparently that didn't stick because apparently I think the provider automatically repoles and resets configuration but it worked for at least five minutes for testing. So yeah, it's a bug somewhere with the K ingress mapping that it's sending everything to the same port even though it's obviously different ports. + +Scott Carruthers: Okay. + +Derrick Hammer: And my original expectation was I was actually surprised because I didn't think you actually go through ingress or at least I forgot. But yeah. Yeah, that seems to be a case. It seems like you can't have two web apps on the same thing now to do this. That's for the same application because I expecting one to one be a web service and all of a sudden when I was doing curls or my work I was getting bad responses. I was wondering why + +Scott Carruthers: All right. So yeah, all makes sense. So Andy, I just assigned both of us to this. maybe just depending on who can get to this to test first, so I understand the high level issue is if you're using two different web pods in the same STL that the accept fields. + +Derrick Hammer: No, that No, It's not even two different web pods. It's the same pod itself. It's all one image. + +Scott Carruthers: Yeah. Okay. + +Derrick Hammer: It's all one image. I personally can't use multiple pods because of a lot of other things I've actually talked with Piper about. I've given him feedback and improvements on what needs to be done in that cache. all my deployments are single servers pods so to speak they're publicly exposed because I can't really scale or anything without rebitting so that's a problem so all my stuff is running it's multiple services out of the same image + +Scott Carruthers: Yeah, I see that I was just skimming through the STL, but not multiple pods, but two different services within one pod. So, understood. + +Derrick Hammer: Yeah, it's services within the single container. Yeah, not services abraction,… + +Scott Carruthers: Yeah. Yeah. + +Derrick Hammer: but yeah,… + +Scott Carruthers: Okay. + +Derrick Hammer: that's it. + +Scott Carruthers: So, yeah, I understand the problem. So, Andy, I assigned both you and myself to this so that we can do some testing of this, whoever gets to it first. So, that's fine. so I think somebody had their hand up. + +Benjamin B: Yeah, I have my hand up. I just wanted to mention that if you want me to comment the replication of this issue, the one I created the issue for, just let me know and I can explain more in detail. Sure. + +Scott Carruthers: No, I think with that additional explanation in this example, STL gives us enough to go with. So, yeah, I think it should be fine for So, we'll do some testing of this and get back to you as soon as possible. Any other thoughts on this before we move to additional waiting triage items? so, I believe that leaves one other and this is actually one that I didn't review prior to the call. So, first of all, obviously I will take this out of + + +### 00:15:00 + +Scott Carruthers: test net 5. so this is another one that Derek PC freak you opened. this actually opened some time ago. + +Scott Carruthers: So it looks like this hasn't got any attention yet. So sending a manifest update error crashing pod should force deletion of the pod. Okay. + +Derrick Hammer: Yeah, this one I've had some time fun times. + +Derrick Hammer: I've actually gone so far where I actually have to put my entry point scripts of a while loop. So if I get exit one or error exit, I autoatally back off so it doesn't crash because if I crash because if it crashes, I can't update the image. Yeah, it's kind of problematic. it's like if you're on a remote server, it's like if you're on a remote VPS and you suddenly do something wrong and… + +Derrick Hammer: it goes down, the only option is delete and recreate it because you can't fix the image without and now I can obviously fix it if I'm local if I am the provider, but it gets into an error state. unreoverable without deleting it. + +Scott Carruthers: Okay. + +Derrick Hammer: Yeah. … + +Scott Carruthers: Okay,… + +Derrick Hammer: I've had to put my own redundancy in for that error cuz because it can happen if you're not paying attention. + +Scott Carruthers: I understood what the request is. So, I assigned both core. So, yeah, I'll follow up on both of these issues, the one that we discussed previously and this issue and do some testing and discuss it internally and get back to you so I'll get an update on both of the issues that we just discussed as soon as possible. + +Scott Carruthers: I would hope either later today or possibly tomorrow with the late Okay,… + +Derrick Hammer: So yeah, if you need to ping me, I'm pretty much here full time. + +Scott Carruthers: sounds good. any thoughts on so that now gets everything out of awaiting triage with proper assignments. Also fix the issue of things being in test net 5. so a couple of items that we just discussed I'll follow up on very soon. + +Derrick Hammer: There's one thing I'd like to bring up. + +Scott Carruthers: any awesome. + +Derrick Hammer: That I gave the but I'm not sure how they're rating triage but if you can go and search for issues and search them by me you'll find one that's a recommendation that I found that a while back that I think is very flipping important. I'm not sure how the triage works, but if you filter the offer by me, you'll find it. yeah,… + +Scott Carruthers: Yes. Okay. + +Derrick Hammer: it's actually in the middle schema for Yeah. + +Scott Carruthers: So that's interesting. + +Derrick Hammer: Yeah. I have no idea. + +Scott Carruthers: Was this intentional that it's got a bounty? + +Derrick Hammer: I don't know who did that. I never put that on there. + +Scott Carruthers: I wonder Tyler, is this something that when do you have any recollection of this? + +Scott Carruthers: Was this + +Tyler Wright: No, there's no bounty behind this. + +Tyler Wright: I think we should take that label off. + +Derrick Hammer: I don't know + +Derrick Hammer: I don't know how that happened, but yeah, I was talking to Piper about this. I was talking to Piper about this 30 minutes ago and I'll read some of the comments I was telling him in just because it's easier. basically you can reference it anywhere to be sure you're using them in Val which is why it should be a dedicated book. The spec should be source of truth. The go proto books should be based off of and then create the API docs from the JSON schema because right now the SL sticks are kind of dog s*** docs wise a lot of this can be automated CI sorry I tend to ramble so I just am referencing what I was telling him the bottom line is my experience is that everything been all over the place with SEO and really the SEO documentation is kind of bad and there realistically specwise there is no formal reference for the spec that I could find and so automatically if you can and I + +Derrick Hammer: actually did some very rough stuff to create something just as experiment but if you need to have something that's a source of truth for the STL so you can update it and dedicate repo and then everything can waterfall from there and then pretty much everything can automatically understand that and there's something I believe I put in the link but there's a website schema store where a lot of ID + +Derrick Hammer: and editors can automatically pull from it if not via plugin and that means that every single editor especially like jet brains and via storm atom can automatically parse and validate in the editor just by having an adjacent schema and then my opinion is you should take the protobuff based off the schema to keep them in sync and then you can use CI to automate generation of everything else including API documentation and so you can basically fix all the schema questions and problems just by having a single JSON spec. And this is a real problem because I'd actually had to go into the provider code reverse engineer to figure out how some things work for my own development in this past November and December because… + + +### 00:20:00 + +Derrick Hammer: because of some ambiguity I wasn't sure what was going on and I was doing a lot of deep edge case stuff. + +Scott Carruthers: Tyra,… + +Scott Carruthers: I noticed that you just removed Bounty. as I recall, so if you actually look at the history of it, and it's been almost a month and… + +Scott Carruthers: a half to close to two months since this was open. I think we actually intentionally assigned a bounty for a community member to look at this. So, I don't know if you remember + +Derrick Hammer: I personally think this is something that core should be doing… + +Derrick Hammer: this is because I think this something should be considered frankly a priority given the nature of it and especially if you want people to understand because I'm used to dealing with hard problem it comes to the territory what I do but I had a very rough on boarding and took me almost for the month of November between November and December to get aash and outside of getting provider setup it which was a fun five days of working with Seth and reinstalling the operating system up 50 times it was figuring out the SCOs and understanding it all because most of things because half the time you have to talk to a community member to get help and I've made rapport with several community members who have helped including Pyber and + +Derrick Hammer: and Norse. But yeah, but the SCOs's inbuying needs to be solved if you want to get adoption because it's not clear at all past a few probably out the examples. + +Derrick Hammer: You need to have a spec and automation from that spec. Right now it's kind of like chaos. + +Scott Carruthers: All right. + +Scott Carruthers: So, power, why don't we, have a discussion about this? + +Scott Carruthers: And so, if we do believe that, so it seems like there's a number of things here improved STDL documentation. I think that's definitely something that a community member could help with. + +Derrick Hammer: The SEO documentation can be approved,… + +Derrick Hammer: but SEO documentation can literally just be also API docs with comments from the STL spec. If you comment everything, then you can basically treat it basically swagger API generation for a JSON spec and then basically have everything be automated. Then you don't have to worry so much. You can use tutorials and examples, but you can just literally have everything be automated API most rest services and… + +Scott Carruthers: Yeah. Yeah. + +Derrick Hammer: then you don't have to worry about it being out of sync. + +Scott Carruthers: Yeah. Completely understand. So we'll discuss this further internally and determine next step. So yeah, entirely understand. It's obviously not the first time that we've got feedback that the SDL documentation needs to improve and… + +Derrick Hammer: + +Derrick Hammer: Yeah. + +Scott Carruthers: automate. + +Derrick Hammer: My hope is I can give you a concrete more like standards accepted way of doing this cuz kind of really it feels like that ecosystem is decentralized enough to… + +Scott Carruthers: Okay. Yep. + +Derrick Hammer: where there's a lot of chaos and a lot of things are not consolidated enough still for people coming in and… + +Scott Carruthers: Completely understood. Okay. Okay. + +Derrick Hammer: And I'm giving Pyiver a lot of feedback as well so he can try to route it a little better cuz I'm just like seeing a lot of rough edges still that I'm having to deal with this having to deal with while trying to work in prod and… + +Derrick Hammer: doing things + +Scott Carruthers: Yep, understood. + +Scott Carruthers: So, yeah, we will definitely discuss the STL documentation and obviously the automation behind it would be certainly an improved and preferred direction. So, yeah, we'll definitely take a look. + +Scott Carruthers: Any other issues that anyone wants to discuss? + +Derrick Hammer: I'm not sure this will be applied in support,… + +Derrick Hammer: but I have brought something up that I would appreciate if core can deal with in the provider repo that I left with the draft. something I've been dealing with that I've had to deal with and I'd like to get some attention on is I listed the draft but it's been sitting there. I'm not really sure procedures for anybody reviewing that stuff. + +Scott Carruthers: opened or you just want. + +Derrick Hammer: It's a draft PR that I'm meaning feedback on and and I think I've talked to both Norris and Wuna about it and so yeah, but it's not an issue to really close, but it's something I do want to get Co's input on because I need to know how this should work because I need to get some merge upstream. Yeah, I think. + +Scott Carruthers: Okay. So,… + + +### 00:25:00 + +Derrick Hammer: Yeah, I think. Yeah, It's a draft regarding environmental variables. + +Scott Carruthers: and have you submitted that PR. + +Derrick Hammer: Yes, it's been saying yeah,… + +Tyler Wright: Yeah. Sounds good. + +Derrick Hammer: it's been sitting there for a while. + +Scott Carruthers: Okay. + +Derrick Hammer: Not a while, definitely more than a week now. + +Scott Carruthers: So Tyler, maybe we want to follow up and have Arer review that PR. + +Derrick Hammer: And I honestly don't expect you to take it as It was more of me just kind of hacking things because I'm figuring out a code base. So, it's more of just how I did it and however you get the result is better than how I did it probably. + +Scott Carruthers: Okay. + +Tyler Wright: Yeah, I have that poll. + +Scott Carruthers: Yeah. Yeah. + +Tyler Wright: I think it's provider poll 259 and I'll follow up with internal core team and make sure there's a comment in the next day or so. + +Scott Carruthers: Okay. Yeah. Sounds good. + +Derrick Hammer: No problem. Yeah, thanks. yeah, sorry. I'm not sure what the governance for teacher on this, but I just wanted to get that I had to get that one handled. + +Scott Carruthers: Yep. Yeah, absolutely. Yeah, we'll take a look very soon. Okay, open table. + +Tyler Wright: No, I do. + +Tyler Wright: Go ahead, Scott. Sorry. + +Scott Carruthers: So, anything else anyone wants to discuss either in open issues or free forum for anything else on anyone's mind? Okay. Yeah. + +Derrick Hammer: + +Derrick Hammer: Don't think so. Unless there's something I opened up and I forgot about I've been a bit busy the past month or so deep inside the system. + +Scott Carruthers: Definitely appreciate all the efforts and yeah, if no one else has any, so we've now taken all issues out of awaiting triage. There's a couple issues that we are going to follow up with. if there's no other questions or any open issues or… + +Scott Carruthers: discussions, I'll pass it back to you, Tyler. + +Tyler Wright: in the provider special interest group, but it's also great to bring it up here. So, Derek, much appreciate you bringing up these things and bring attention to them and we'll make sure they follow up. + +Derrick Hammer: … + +Derrick Hammer: one thing I do have a question about that I like to figure out why I'm here. because I actually did a lot experimentation on this and I had to stop because I was under time constraints for my own priorities. I actually went and I spent a lot of time on this but it has some barriers. I said I literally kind of rebased Luna's probably year old SSL experimentation for Vhost and I was wondering if Core has any plans in the road map actually handed it themselves. + +Derrick Hammer: I'm not sure what Luna is planning on doing with their project, but I think it's like one or two PRs they had sitting out and I actually went in just and rewrote half the PR to because the code they old was a year old and I was wondering what's the current position on getting SSL support in via host without needing a decip address using the vhost routing. Why? Yeah, for basically let's encrypt and stuff. + +Tyler Wright: I don't know where that fits in on the road map. I would say that if you have any specific questions like that. using the SIG support channel is a great place to start because I know me many members of the core engineering team look there. and then again we can make sure conversation You're also looking in the right places with members like Pyber and… + +Tyler Wright: Luna who all have easy access to many members of the core team and can kind of push things along. + +Derrick Hammer: Yeah, FYI. + +Derrick Hammer: Yeah, I'm waiting on Adam. I'm probably going to be insiders within a week, assuming that he doesn't get sidetracked. + +Tyler Wright: Adam W. + +Derrick Hammer: Your community lead. + +Tyler Wright: Okay. Gotcha. + +Derrick Hammer: Yeah, he's basically dragging me into the insiders. + +Tyler Wright: All right. + +Tyler Wright: no, but I would say that if you could drop that specific issue that you've been working in speak + +Derrick Hammer: I don't have the PR. I more just kind of rebased and worked into my own for branch because I need to get done and I just ran out of time and I als and I assume that Luna has more resources and able to do it better than me cuz he started it but it almost feels like something that course should probably take on and so whatever I might have done is probably just hacks. I'm far from an expert in the codebase. been working in act about two months. So I'm more of trying to figure out if core is going to take it on not cuz it is something that's kind of a real annoyance. and I've had to forgo doing SS on a lot of my deployments and… + + +### 00:30:00 + +Tyler Wright: Cool. No,… + +Derrick Hammer: I've actually have day IP piece for my ISP because of that. So yeah, I figured I'd bring it up because it seems it kind of has just failed you got fill in the noise. + +Tyler Wright: I've noted it and I'll make sure a member of the core team puts a message somewhere. I'm probably going to talk to Luna and then put a message in SIG support. again there is a updated roadmap that's pretty comprehensive that has been talked about in steering committees and that will be shared to the wider community what about things that the core engineering team is going to be focused on over the next year. + +Tyler Wright: and then again, I'll get you some feedback here. + +Derrick Hammer: Yeah. + +Derrick Hammer: I guess if there's an updated room, I figured I might be able to voice this because basically I just kind of brain dumped on Pyiver about 30 minutes ago. However, my other suggestions about the limitations I was having about resizing and resources, would you think you'd be able to communicate that better than I could? + +Benjamin B: Yeah, I don't think I'll do it better than you, but I can give it a shot. + +Derrick Hammer: Yeah. I'm a rambler and I can be technical… + +Derrick Hammer: but yeah I'm not communication voice isn't + +Benjamin B: Yeah, just give me one second to bring the conversation back up and… + +Benjamin B: so I have it even more + +Tyler Wright: might be if you are again you have been a fine communicator to this point but if you feel more comfortable I would just ask you can drop these messages the same thing that you're kind of dropping to Pyber you can also drop in a public channel and then get more people across the community that might have similar thoughts or… + +Derrick Hammer: Yeah, I Trust me. Yeah. Yeah. + +Tyler Wright: opinions Thanks. + +Derrick Hammer: Trust I already ramble in public chat. it's just a question of making sure that the right people are seeing it. Which is why I'm telling you here because I know when I call is very different than AC chat when you're having to work in five projects once. + +Tyler Wright: Yeah. No, that makes sense. + +Derrick Hammer: So I'm just trying to make sure the cord team. + +Tyler Wright: Go ahead, Benjamin. + +Derrick Hammer: So yeah, I've already given a lot of feedback in a lot of different areas,… + +Derrick Hammer: but having different members respond when everybody scatter brain on their own stuff is very different than having a call to make sure it's awareness. + +Tyler Wright: No, I appreciate that. + +Tyler Wright: Go ahead, Benju. + +Benjamin B: Thank you, I think the best way to summarize this is basically that is suggesting that we open up more control to existing leases so that we're able to scale them up and down with we used to have this road map item that was renegotiation and we also have a lot of users supporting some other stuff like being able to restart the pods, being able to change the ports, being able to change all these details that are not possible right now without creating a new deployment. I think everything relates to that just being able to change the details in some way, even if it requires a new monthly price or anything. + +Derrick Hammer: Yeah, I think that sums it up pretty good. Bottom line is that is I'm working more on traditional a I'm not using AWS and full type stuff but I am working such that I'm very limited. I've been able to work around it but it's like if I want to resize stuff or if I want to deal with ports, memory, RAM or anything, the answer should not be open a new deployment, migrate everything over, close and close a new that's kind of a non starter when you compare to any other traditional cloud service like the atomacy of if you want to change anything that requires that about the money that you have to start + +Derrick Hammer: especially and I'm not sure who L if it was Luna or I think it was North who's also said he was pushing to have a dedicated persistent sources its own separate resource which also makes sense that if you're having to recreate a resource and has persistent disc you're kind of effed and so versus where at least with VMs like you can just attach and reattach hatch it. For example, digital ocean, if I want move just between servers, it's not that hard. I might have to shut them down some cases, but I can just reattach them. The limitation right now is that if I like what I just had to do due to that recent port bug, I had to go and destroy the deployment and have my Terraform recreate it. + + +### 00:35:00 + +Derrick Hammer: Then I had to all shut it down with another pod and sync over K8 with this copy and some other stuff a consensus database because I had another node to reync the database because I'd had to reync the whole thing together because the only solution was destroying everything. I couldn't fix it without destroying it. And if I didn't have the database, I'd be sitting here for 3 days reyncing. And so I've had to do a lot of workarounds because the auto atomy of the deployments and I could not and it was very very very rigid. I get what aash now works for the current use cases and that you have something working now but this is criticism feedback if you want to grow to be your long-term go to replacing AWS or digital ocean or vulture + +Tyler Wright: Again, I think that's very valuable feedback that we can share with the core team. This meeting is recorded and transcribed. So again, we can follow up on a number of these items, both the ones that are issues or… + +Derrick Hammer: … + +Derrick Hammer: one oh, one oh,… + +Tyler Wright: ones that are general concepts. + +Derrick Hammer: one other thing I noticed. I'm not sure why I'm going to your community repo and at least this and at least the support repo. I've not seen any of the comment I'm not sure I'm looking in the right place, but any of the notes for the meetings at all. I'm not sure they're being put somewhere else, but if I go to the six support and the community repo, I don't see any latest meetings at all. I'm not sure where they're at. I'm not sure and I'm not looking in the wrong place. Something that's giving the impression that meeting that a lot of the meetings are out of date because I've looked over the past year and checked back in to see some of the means what might have been going on and I couldn't tell if some active groups were active or not still. + +Tyler Wright: Sorry, I'm on mute. I was saying I dropped in a link to the six support repo. We changed the cadence of this meeting from weekly to monthly. and I think in November we didn't have anything, but I can work with a member of our community that helps out with this and make sure all the notes are They should all be again up to date. there's a member of the community that updates the notes. I add the videos and the transcripts and then on the add they read me all inside the community referral. So if there's a meeting that you haven't seen in a while it not exists anymore. There's also a calendar that I can link you to. but again anything related to the structure of the SIGs and… + +Tyler Wright: working groups feel free to reach out to me via Discord. + +Derrick Hammer: Yeah, I guess for some reason I think I can't tell you why,… + +Derrick Hammer: but for some reason when I was looking impression of reading them that it felt like things were out of that like you I guess that something wasn't being updated. yeah, I'm not exact Hold on, let me see. Yeah, I was actually sure. One thing I might suggest is that you structure it so the latest is first because if I have to scroll all the way down to the bottom that I just glance at the number and… + +Derrick Hammer: it's like I'm looking I think they should be descending not ascending. + +Tyler Wright: That's great feedback that I can take. + +Tyler Wright: So much appreciate it. + +Derrick Hammer: Yeah. It's just like the point is that my opinion is if you're looking at this, you're want to see what was happening the latest. If you're just seeing and the first thing you're seeing is the fact that the first meeting was over 3 years ago and some reason I got the impression that something wasn't happening. So I think that some just might review another review just to make sure that everything's in line on the calls least at one point I've had an impression that some of these meetings were either dead or I don't know what. yeah, I think that a review of this and probably restructure it so that make sure that if someone's looking at this, they get an impression of what's happening the latest,… + +Derrick Hammer: not what's happening first. + +Tyler Wright: Gotcha. I think that's great feedback. + +Derrick Hammer: But especially if you want people to get involved and know and know what was done the latest that probably makes sense for them to see that first and then go back historically. + + +### 00:40:00 + +Tyler Wright: All I know we have about 18 minutes left and we want to get to the second part of the agenda where we talk a little bit about just general support in Discord, Telegram, and some other channels. I know there's some insiders here. I think what Derek just brought up about restructuring the community repo to make it easier for folks to see, what's happened most recently and kind of restructuring the readme files. It's a great feedback coming out of support. + +Tyler Wright: Just want to see if there's anything else that anybody has seen over the last month or so. I think we haven't had this meeting in December because of scheduling. but just want to see if there's anything else over the last month or so that anybody wants to call out in terms of just general support or stuff that you've seen in Discord, Telegram, or elsewhere. + +Derrick Hammer: I was waiting to see if he was what else was going to go because I've given some of the feedback to Adam, but one thing I think there's one in let's just say I'm probably going to be nagging Adam about this anyways because there's a lot of annoyances. because when I came in and I gave him a lot of feedback regarding preventing anti-abuse because there was a lot of dumb simple lowhanging fruit that could be fixed and he did fix a good amount of them. but I guess two things are and probably want to talk to him about this is one's kind of dumb. I think in the process of trying to block Discord links, you can't link to messages in the Discord itself. I've been blocked so many times. not too many times, but maybe two, three times now. Or if I just link to a previous message, I get timed out not even an hour, for a whole day because it's a Discord link. And I would probably bring that up to Adam, but it's kind of dumb. + +Derrick Hammer: And the other is that I think that I brought this happen too. I'm not sure if he's decided against it, but one thing I found is that for people coming in and I had to learn this being told it is I was not aware your vanguards are basically your moderation team. And anyone coming in is going to look for a mod group. And the fact is that you don't have a mod role. And if anybody wants to tag a mod, they don't know. it feels like there's outside the core team, there's no authority presence to deal with the clear spam you have. And that's kind of bad. + +Derrick Hammer: So I think something so I think you should be looking at definitely how to improve further making sure of moderation of people coming in that they can know how to deal with someone because even if you have put in tutorial when you come into the server you can be honest that most people don't read stuff the amount of time I see people even side come ask question they don't read anything so you have to go with the common experience of what discord roles people expect and so if you can even ali some of that's possible but I do think a minimum you probably have modded man or something like that basically make it so that people can know who to turn to easier to be able to be sure that somebody can get some attention because vanguard is something cultural to hash and pretty much nobody outside of the tribe is going to know what the hell that means. + +Tyler Wright: Again, I think that's good feedback. I can make sure that that gets brought up again at insider meeting. and I'll talk to certainly Adam about that. much appreciate that. + +Derrick Hammer: Yeah. Yeah. I gave him some of the feedback. I'm not sure what he has or hasn't done, but I've deemed him a lot of random stuff over the last two months. And so I've already given him some of his feedback in DM, but I figured it'd be best to repeat here. And in Saiya I come from a community and where the community manager as a compliment kind of rules with iron fist and nothing really gets past them and it's very very clear who is who. + +Derrick Hammer: + +Derrick Hammer: So yeah, they don't have as way as many problems as a cash and visual they actually told me that they get 10 to 100 bots a day but you don't see any of them because I assume it's in private moderation channels but yeah I'm seeing other moderations being done and I'm just seeing what happened when I came in and given that fact I know I'm with a minority in what I do and my skill set. It means that things need to be more obvious and simplified. + + +### 00:45:00 + +Rodri R: Yeah. So, maybe just adding mods besides Vanguards. usually people tag us when they're scammers and all this stuff like this guy said, like Derek says. U, maybe that will help a bit. + +Derrick Hammer: I'm not really sure what the limitations are, but I think there should be some way that you literally have at least a mod role that you can just mod uppercase or lowerase and just be able to quickly flag stuff. that's kind of the biggest point is if you see a spammer and such and you like type and type mod then it should route correctly because right now nobody if you're a week in Aash even day west and you see something going on and you see Vanguard but if you see a mod or a min it's kind of common sense. + +Benjamin B: I just want to say that I'm against giving a lot of people the roles to ban people and time out people because you never know who's going to get their account be hacked or fished or anything. But it could be possible to just keep everything the same and add an a role that isn't visible on the wallet list for moderators and give them to all the vanguards. So, they're taggable. I'm not sure what is the value if they're not on the wall, but just wanted to say that is a possibility. + +Derrick Hammer: Yeah, I think it might benefit to be both see them visible and for them to be untagable. And and I'm not saying by saying who has a power, just the fact that if I go into I'm seeing a suit of bot string spam, I did not know before somebody literally had to tell me in private that a Vanguard is a moderator. I did not know Put that way. + +Benjamin B: Yeah, maybe it would make sense instead of calling them vanguards to just call them moderators at least on the + +Rodri R: Yeah, that's kind of what I was saying. Maybe just either change vanguards to mods or just that besides parenthesis. + +Derrick Hammer: Yeah, it's not really about what their role name is. I guess it's more making sure that whatever happens that new users who do not know insider culture aash doesn't know knows how to get help. + +Rodri R: Yeah, makes sense. + +Derrick Hammer: And right now if you're using tribe specific terminology that ain't and you're trying to abort people that's going to make culture clash. + +Tyler Wright: I do and again I'll make sure that I follow up with Adam on that and… + +Rodri R: Yeah, I did too. + +Tyler Wright: we could discuss this further both internally and then probably an insider meeting but that makes clear sense to me. just want to see if there was anything else that anybody wanted to discuss as it pertains to support. If not, then again, I'd much appreciate some of the new voices in this meeting, some of the conversation. + +Tyler Wright: I know there's a number of action items that I have as pertains to the community repo and following up on some of the items just to give awareness to the core engineering team and again we can get feedback on some of those issues PRs etc. again if anybody has any comments or questions related to any issues in the support repo please drop a comment in there and tag somebody that is working on it. I think that just by being assigned to it, you're going to get notified. if anybody wants to have any follow-up conversations around support, please feel free to use the six support channel. It's a public channel, so again, more people insiders, others can get involved in the conversation. but again, much appreciate this conversation. + +Tyler Wright: I'll make sure that I talk to community members about just updating the structure of the notes so that this recording transcript notes are just as easily as accessible as possible. So thank you everybody for participation. Scott, thank you for leading us through issues awaiting triage. I hope everyone has a great rest of their day. and I'm sure that we will talk soon, but thank you all. Goodbye. + +Benjamin B: Thanks guys. Bye. + +Rodri R: and grace to you. + + +### Meeting ended after 00:50:11 👋 + + From 355b5b22a05848e80645572fa1f113831e2b39ce Mon Sep 17 00:00:00 2001 From: George Appiah <48031636+iamGeorgePro@users.noreply.github.com> Date: Wed, 22 Jan 2025 19:30:57 +0000 Subject: [PATCH 2/3] Rename 040-2024-1-08.md to 040-2025-1-08.md Signed-off-by: George Appiah <48031636+iamGeorgePro@users.noreply.github.com> --- sig-support/meetings/{040-2024-1-08.md => 040-2025-1-08.md} | 0 1 file changed, 0 insertions(+), 0 deletions(-) rename sig-support/meetings/{040-2024-1-08.md => 040-2025-1-08.md} (100%) diff --git a/sig-support/meetings/040-2024-1-08.md b/sig-support/meetings/040-2025-1-08.md similarity index 100% rename from sig-support/meetings/040-2024-1-08.md rename to sig-support/meetings/040-2025-1-08.md From e20e675ea4c640b74a9364d2552b4c6f83ca3e5b Mon Sep 17 00:00:00 2001 From: Drinkwater Date: Sun, 26 Jan 2025 07:32:19 -0500 Subject: [PATCH 3/3] sig-support Update 040-2025-1-08.md added recording link for video Signed-off-by: Drinkwater --- sig-support/meetings/040-2025-1-08.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sig-support/meetings/040-2025-1-08.md b/sig-support/meetings/040-2025-1-08.md index e0418af4..f2ee4e19 100644 --- a/sig-support/meetings/040-2025-1-08.md +++ b/sig-support/meetings/040-2025-1-08.md @@ -10,7 +10,7 @@ - Date: January 8, 2025 - Time: 07:00 GMT-7 -- Recording: Coming Soon +- [Recording](https://jhlxjfafh7tviorcuhxwgow26eafh6jjyje6pazlsxzsn5yjq6dq.arweave.net/Sdd0lAU_51Q6IqHvYzra8QBT-SnCSeeDK5XzJvcJh4c) - [Transcript](#transcript) ## Participants