From d4b3f39a377012993817a6c1dd5dbb2b425567f8 Mon Sep 17 00:00:00 2001 From: Drinkwater Date: Tue, 28 Nov 2023 16:22:18 -0500 Subject: [PATCH] sig-support Create 022-2023-11-22.md (#390) * sig-support Create 022-2023-11-22.md Signed-off-by: Drinkwater * Update 022-2023-11-22.md Signed-off-by: Drinkwater * Update 022-2023-11-22.md Signed-off-by: Drinkwater --------- Signed-off-by: Drinkwater --- sig-support/meetings/022-2023-11-22.md | 437 +++++++++++++++++++++++++ 1 file changed, 437 insertions(+) create mode 100644 sig-support/meetings/022-2023-11-22.md diff --git a/sig-support/meetings/022-2023-11-22.md b/sig-support/meetings/022-2023-11-22.md new file mode 100644 index 00000000..cde0753d --- /dev/null +++ b/sig-support/meetings/022-2023-11-22.md @@ -0,0 +1,437 @@ +# Akash Network - Support Special Interest Group (SIG) - Meeting #21 + +## Agenda + +- Participants will look at issues in [Akash Support Repo](https://github.com/akash-network/support/issues). +- Artur will lead triaging issues, and discussion of implementation for issue fixes. +- Artur and Anil will lead discussion putting severity on issues. +- Scott will discuss Support from the community and ask for any community feedback. +- Participants can have an open discussion of any other issues or support related items on Akash Network. + +## Meeting Details + +- Date: Wednesday, November 22nd, 2023 +- Time: 07:00 AM PT (Pacific Time) +- [Recording](https://l5yeilkhvwmpqetchbk2pax4nxmu6miv23cxy7iszl7nvgbk4lka.arweave.net/X3BELUetmPgSYjhVp4L8bdlPMRXWxXx9Esr-2pgq4tQ) +- [Transcript](#transcript) + +# Participants + +- Andrey Arapov +- Artur Troian +- Benjamin B +- Deval Patel +- Max +- Rodrigo Rochin +- Scott Carruthers +- Tyler Wright +- Zeke E + + +# Notes + +- Artur talked about how busy the last couple of weeks have been. There was a mishap with 1 small edge case after the network upgrade. +- New Mainnet 9 coming next week, November 28th, 2023. This should be a quick upgrade. No action is needed for any clients. +- Artur talked about AMD support. This is something that has been thought about it since GPUs were brought onto Akash. +- Testing of AMD provider will be tested on the sandbox after the network upgrade. +- Provider will need to be upgraded when AMD is implented. +- Scott has been working on Feature Discovery. This is a "p1" issue. +- The group talked about issue 153. All issues need to be in proper formatting. Benjamin has talked to this user on discord. +- From experience, any feedback or recommendations i've given to the user has not been followed. there's no reason to waste time on someone asking questions but not listening to the answers. +- The group talked about issue 144. Artur mentioned that there is no way to fix it right now. +- Issue 148 was discussed. Andrey mentioned that this might be fixed with v0.30.0. We will monitor it. This issue was assigned to Andrey. +- Sig-chain mainnet 8 is closed. + + +## Support + +- Not all providers have upgraded to mainnet 8. +- Praetor team is working with users who had trouble during the upgrade, and did not use 1 click upgrade. +- Rodri noticed that there are old docs from old network upgrades. +- Rodri talked about his provider issue. He installed some extra stuff on his provider during the upggrade. +- Deval asked about how to decrease development time. Artur asked Scott to share link to development documentation. +- Scott shared documentation with Deval https://github.com/chainzero/akash-engineering-notes/tree/main/akash-development-environment. +- Artur gave a demo as apart of a new series with developer office hours during sig-support biweekly. +- The demo was around Akash development environment setup. + + +## Action Items + +- Track issue 148. ThiS may be solved. +- Discuss issue 153 after network upgrade to Mainnet 9. +- Max will be working on endpoints for Praetor team. +- Make a new issue for sig-chain Mainnet 9. + + +# **Transcript** + +_This editable transcript was computer generated and might contain errors. People can also change the text after it was created._ + +Tyler Wright: All right, welcome everybody. Who seek support monthly meeting I believe this is number 22 or student biweekly meeting number 22 during these six support biweekly meetings. The group goes over any open issues in the support repo anything that's a waiting triage and Artur and sometimes Scott lead the group through triassing those issues last week or excuse me, two weeks ago Artur was out of office of Scott. Let us through triaging and looking through a number of issues in the support repo. + +Tyler Wright: One other thing that we're looking to add as a part of the sixth support biweekly is after we triage issues and if anybody has any questions about any issues and those questions are answered we're gonna hand it over like we normally do to talk to Andrey who's been very hands on and support and some of the other members of the Insiders that are on this call to talk about any support issues or any issues that might need further conversation when it comes to new and existing users using We find in Discord and then after that if any time remains and usually we do have time to remains after that, I want to start opening it up to technical questions. So I know that Archer and Scott members of the core team along with Andre + +Tyler Wright: Are often at these meetings? And so again if there's any time left after triaging any issues and support repo and talking about General support in Discord and Beyond I do want to give some time for folks that have specific questions. I know deval's here today from the pray to our team. We've been working on content moderation and we'll hear a little bit more about that in Sig providers, but I know he has some specific questions related to the code base and so again, When the opportunity presents itself, I want to use some of this time for members of the community to meet with members of the core team to help with anything very technical. + +Tyler Wright: So without further Ado, I'm going to hand it over to Archer to get us going with triaging any issues initiating and the support retail. + +Artur Troian: Thanks, Hello, Alright, so last couple weeks were quite busy for us. So on this last Sunday we formed. successful Network upgrade There was a little mishap in one of the things, like the new feature it's Be attributes. + +Artur Troian: one of the functions Was implemented slightly right? So we are on the way to make one more Network Tuesday, November 28. It should be really quick and does not change doesn't alter any state. It's just + +Artur Troian: the behavior called function does cross validation between the order so hopefully it will be really quick and there's no actions required from any clients. It's purely the blockchain conversation. Moving on. Yeah, we have I think. Not that many new issues right now. So I want to talk about this one. + +Artur Troian: We have been thinking about MD my sense In southern our table and we have pretty much everything to support that. Right after Network upgrades next Tuesday. We will be a bread updating provider that allows users that allows provider to beat on MD. And gpus so right now this part is disabled intentionally. As soon as we do that we will be doing the testing on the sandbox. Say that you don't have any major issues and if everything goes it will be publishing provider. Vs. MD support. So once again from that perspective, there will be no updates needed by the clients. It's really The Wider part. of this feature + +Artur Troian: The next one is future Discovery. This is rather big Improvement and Scott is more money working on that. So we don't have any levels here one wider idea is that we want to improve visibility on Water Resources provider how as well as a remove manual lightly of the nodes for example, writing for the persistent storage and the gpus as well as this service will be updating provider attributes on chain. So the completely rules are user error. I mean users to provide owners still will be able to do that. They need to but we try to encourage to that's a provider features of our service is out. + +00:05:00 + +Artur Troian: Providers to just use automated thing as well as we still thinking that to add quite a few same into the future Discovery. It will be merged with the inventory service. and the gonna drastically improve visibility of the dashboards like Cloud boss. To see what actual amount of gpus and what exact model of the gpus for example available and given a Time. + +Artur Troian: Okay, and this one, what is that? my God. I mean Yeah, this is. + +Artur Troian: Let me see. + +Artur Troian: Let me just answer. + +Andrey Arapov: I think someone is just running a SSH service and it's getting killed. I think actually this one of the deployments that was causing all those zombies + +Artur Troian: and yeah, let's go start with promoting the sdl. + +Andrey Arapov: Yeah. + +Artur Troian: + +Tyler Wright: While archers are typing response go ahead Benjamin. + +Benjamin B: those wanted to say that this user has written in the Discord server as well and some of the suspected reasons to why crashes is that he's trying to deploy something with only one gigabyte of storage and then when it tries to install something that's larger than one gigabyte. Of course, it will crush because the STL and the lease is more limited than what he's trying to do. I wrote to him in this group that is trying to rent the persistent storage as 100 gigabytes or whatever depending on my stylus using but unless + +Artur Troian: Yeah, yeah, let's see. I've encourage everyone if you see such poorly formatted requests or the issues, let's start first of all with awesome them to promote it properly because it's just absolutely + +Artur Troian: Unreadable, right and why should anyone Vice the time to try to realize what's going on there? so yeah, I just made the + +Artur Troian: reply there. moving on. Let me see. + +Artur Troian: the first of all lost + +Artur Troian: Good. Yeah. + +Artur Troian: Yeah for this one. I don't think + +00:10:00 + +Artur Troian: I still in trying to figure out what's going on here. + +Andrey Arapov: Do you want me to tell you? + +Artur Troian: The problem the wider is going to take each other GPU it has available. so + +Artur Troian: let's yeah, we will review that after the upgrade I think Or maybe would fix this thing. If there is an issue is the MD support. + +Artur Troian: and yeah, so this is the sdlery design and there is no way we can improve it right now, unfortunately. + +Artur Troian: Okay, so that's the service count here. that you see. + +Artur Troian: All right, that one is done. So what happens with this issues is the wider statistics. it's purely the statistics of point. So it doesn't account the replicas for the cheap resource image On this objective marketing to fix that. + +Andrey Arapov: I think I've tested this. + +Artur Troian: that is + +Andrey Arapov: Sorry Artur for previous one. I think it's already solved but we'll see in all 29 or c0 that I tried setting one GPU encount to and it's seen us to request it. So, I mean, it's possible that your latest commit fixes this in all 30 Network. + +Artur Troian: Okay. Thank you. Okay, so I'm not sure what we can do here that is all already supported and I believe you talk about that. So the ram be + +Artur Troian: it does so as soon as label. Let's say not + +Artur Troian: this attribute is not quite actually correct. It has to be like that. + +Artur Troian: please + +Artur Troian: And so it has to be this way. mmm + +Artur Troian: Okay. + +Artur Troian: Yeah, so as soon as not is labeled and the + +Artur Troian: and why they're attributes labeled correctly into work so that I believe it all has invested in the you test not so. + +Artur Troian: Yeah, I don't think anything to do here. Is that okay? + +Artur Troian: double check + +Artur Troian: Okay, so you are here? + +Andrey Arapov: And I send it to me, I will go of his Foundry. I will try this attribute. + +Artur Troian: presentation All right,… + +Andrey Arapov: Thanks. + +Artur Troian: I think yeah, this one is done. + +00:15:00 + +Artur Troian: So that was when you don't date on 19. + +Artur Troian: okay, I think that covers any questions + +Tyler Wright: Thank you As Archer just asking anyone have any questions about any open issues before we move on to General support. + +Tyler Wright: After we talk about again any issues that need triage and if anybody doesn't have any questions or after we talked about the question related to any open issues and support o. The next part of the agenda is around just general support and Discord and in other channels with members of the community new and existing users that are interacting with the crash. I know that Scott has previously been giving some updates but now that he's more an engineering side and Andrey been helping a lot with providers. I'm going to ask Andre if there's anything he's seen in Discord on the provider side and then there's a number of insiders here that I'm hoping will continue to come to this meeting and give any updates that they see for many users either deploying or becoming a provider so Andre, I know that you're quite active in Discord helping again providers. + +Tyler Wright: Helping anybody that needs support with the cash. Is there anything that you're seeing over the last couple of weeks that might be worth note to the group here today? + +Andrey Arapov: August the main concerns that not all providers have upgraded to maintain or they may be have but they need to bounce the cash provider pod. Basically, they providers need to always check that they can beat after they upgrade probably added to the next procedure for great. So the ensure that The quality is there so I guess the only point and don't have anything else to add. I guess nothing comes to my mind as of right now. + +Tyler Wright: I said go ahead Deval. + +Deval Patel: Eight of our providers so far. I'll create that's what I know. I don't know how the objective later right? That was the number last night. I'll double check. today morning and find out. Yeah, that's right who is not being able to update using one? Click button or something? some people actually tried doing, the official document and one click great those people having issues somewhere, right? So I'm trying to help + +Tyler Wright: Executive all that's a good call out. One thing that will make sure that we do is clear as possible is because I did see this in a channel when myself or Andre are making announcements. what we tried to do is as early as soon as we can we try to share information with the page to our team. Sometimes it's running up against the clock and they don't have much time to be able to do one click upgrades. But again, we try to share that information with the parade or team as quickly that the provider upgrade had been tested and are ready to go out when we make announcements. We'll make sure moving forward to make it clear that if anybody's as upgraded or has used parade or path for their build to have Specific Instructions that will link to the ecosystem pray to our page and then Deval into gar will get the instructions on when things are ready to go there. + +Tyler Wright: just so we can try to avoid any extra confusion for folks again, even though both are using Helm charts. It seems like there may have been some issues for a couple of folks. So we'll make sure that the messaging is very clear and then in the provider announcement channels and the providers Channel moving forward for those that are again upgrading or using pretor for their bill versus using CLI at home + +Tyler Wright: Does any insiders have any other things that they've been seeing recently from the community that might be worth either potentially improving to documentation or moving to the process in general that might be worth talking about now. + +00:20:00 + +Tyler Wright: I'll take that as enough. All right. I appreciate go ahead Patrick. + +Rodrigo Rochin: Is I couldn't find the antique. Yeah, I had the little confusion the other day with a document an upgrade. I noticed that there are instructions in the docs from previous upgrade part of their necessary necessary still if not, maybe get rid of some. + +Tyler Wright: Thank you I think that's a great idea. comb through the site and see if there's any older documentation. This is a beautiful segue into there is a Sig documentation meeting happening. Next Tuesday as has been previously discussed. We are going to be relying on the community more in 2024 to help with documentation. There will be some bounties and rewards around those that are consistently contributing to adding new editing documentation. I know the onus has been on Scott for a long time and again with him moving into more of an engineering capacity. We just want to take that off of his plate Andrey and I will be there to help. + +Tyler Wright: Help with submissions and help with RS. Obviously I could help with a copy standpoint and making sure things are structure properly will have very specific instructions on how the structure of the facts should look how the quality of images Etc. So, there will be a framework in some specific rules that we'll discuss next Tuesday, but I know there's a number of people on this call that have shown interest in documentation and Andre will be there somebody that can look at the code. And yes, this makes sense and then, be a test and then again, we'll certainly be relying on the community as well because Andrea the number of other things that he's actively working on for overclock, so It'll be like a team effort. But we just want to make sure that we're getting consistent contribution from folks in the community that can contribute because certainly will need it. + +Rodrigo Rochin: Keep looking through doctors and I'll share the ones that I think are not relevant anymore. Will you do whatever you think of them? + +Tyler Wright: Excellent I much appreciate that Roger. + +Tyler Wright: cool, All right, so I'll go ahead. + +Rodrigo Rochin: through the upgrade recently, and I don't have for instance persistent storage enabled and there's this step that I was supposed to skip but I ran the command. I don't know if I messed up anything or not, I guess not because I think it's working. + +Tyler Wright: Are you saying that there was a command that didn't need to be in the documentation or are you saying that you're having issues with your provider right now? I might have missed that question. Okay, perfect. + +Rodrigo Rochin: from that and that document I ran the command on 2.5 step 2.5. Which I shouldn't have because I don't have presentation search but I didn't read about skipping that section after until afterwards. + +Rodrigo Rochin: I guess nothing happens if I run that. Or should I do something? + +Tyler Wright: So I think Roger's question is if he ran the command to ask for precision stores,… + +Deval Patel: yeah, I Know none,… + +Tyler Wright: but doesn't have persistent stores set up. Yeah. + +Deval Patel: nothing will happen if that command view run with the install then it might have installed. But other than that we should be fine. + +Tyler Wright: Again, we've gone through the part of the agenda where Artur leads us through triaging any issues that need triaging in the support repo. I know that the cloud most team does a great job of any issues that come through Discord or anywhere else that get put in the issues inside the cloud most o. They do a great job of responding to those issues and fixing them in real time so much. Appreciate the Max's Real quick is there anything Cloud most related that anybody wants to talk about as it partake to open issues in the cloud most repo? + +00:25:00 + +Tyler Wright: Yeah I know that they're doing a great job Max that you go out mute. I'm sorry. Did you want to say something? + +Max: Yeah, so when I say to develop that I'll be working on the endpoints that you guys requested to be built today. + +Tyler Wright: Okay. + +Max: So all the + +Deval Patel: Okay, sounds good. Thank you so much, Mike. + +Tyler Wright: Thank you Max. So again, if anybody has any issues related to Cloud please throw those in the cloud most repo and then those issues will be worked out by the Max's again everything continue to talk about those issues here during the six support meeting or we via the specific issues in that report. + +Tyler Wright: Cool, we've also talked about just general support and Discord and other channels. Thank you. Rodri Ben Max many people on this call Andrey that are just active and Discord and making sure that again users are getting the kind of support that they need. I know that it's a number of people that have written to me talking about how quickly they give responses and how it just an improve their user experience so much appreciate the folks that are active on Discord all the time. + +Tyler Wright: This part of the agenda again is a little bit new where we want to have folks that maybe have technical questions get to ask those technical questions Who members of the core team? I've asked of all to join the call. I know that he's been working on content moderation and as a result is been looking at the provider and provider code-based code a significant amount has been working on content moderation Deval. I know that you had specific questions that maybe Artur Others might be able to help with so I want to maybe give you the floor to be able to ask those questions. If anybody else want to drop off feel free. But again, this will be recorded and we want to use this time as a place for insiders vanguards and just Dental members of the community to ask any technical questions, whether it be around products that they're working on that are + +Tyler Wright: built on top of the cost or just the core code base. And again, we'll just try to do this every two weeks and just try to keep it to the hour time slot that we have. The Deval pass it over to you because I know that you had some specific questions at the pertains to content moderation implementation. + +Deval Patel: Thank you So a couple of things I want to give update and mention before we jump into any cool business stuff. So I'm trying to do the coding, in the provider Service as you guys already know what I do is that basically I fought the Ripple and tried as well as the provider and whenever I need to make some changes or taste it right. I push those changes in my frog triple in provider and help and then I tasted becoming provider and things like that, but is there any other way which I can use and do the development faster than I would like to learn that That's number one thing and that's most challenging because then I need to wait for I plan to be past and the talker Bill to be around and + +Deval Patel: things like that, So that will reduce my development time in terms of many hours Yeah, that's one thing. I want to learn. + +Artur Troian: Yeah, so there is a way to do that Scott. Can you please share link to the development environment documentation? So if you look into run Cube directory. That's the way it works. So there's a bunch of my commands and there's actually with me file in it if you go into. + +Deval Patel: I actually tried that right and… + +Artur Troian: Let me stay home and second if you go to run. + +Deval Patel: I was struggling to + +Artur Troian: Yeah, if you go to run read me runcube read me and… + +Deval Patel: okay. + +Artur Troian: just goes slowly step by step like how to initiate it. we take this out quite a few times and there's a documentation Scott you people point you. So the biggest challenge issues you have to use multiple terminal windows. So the node is running one and then you have a provider running and then you just have a third terminal. That you execute your commands. So once again, you check if you go through the readme file that step by step. It says I dedicated, separate or new terminal window for each command. If you just go step by step it will help you to start using it. + +00:30:00 + +Artur Troian: as well as once again,… + +Deval Patel: Okay. + +Artur Troian: yeah, there's a documentation on I believe that thoroughly goes through that and we have been testing it for quite a while and that's actually what we use internally to develop teachers and best things. + +Deval Patel: Okay, that would be very helpful, right? + +Artur Troian: Yeah, so yeah, just to summarize on that. We've been thinking on doing developer office hours. + +Artur Troian: Then gonna show people how to be more involved in the code bias. And so if I can actually try to demonstrate it right now if there is a lot of questions, I'm okay to do that. Okay. + +Deval Patel: which yes + +Deval Patel: Yeah, that would be helpful. so few things to go through that and the developer office hours is a great idea because that way we can learn about the s. There are So let's say the bead engine module the order module and things like that that would be a helpful for us to understand, this module does this stuff? Right? And this is the direction of the code or framework of the court, and that way we can be more productively involved in coding the provider service if needed to be right. So yeah, I don't have much because CID part I already kind of figure out right so I don't have question on part, that particles is all on my side. The next question I have is how to to close all these or at least closing trans. + +Deval Patel: Works and I wanted to understand the flow of the lease closed transaction. And because now we want to allow the providers to be able to close the lease as well rather than developers on deployers only So that part needs to be tricked in the provider Service as well as them if there is a node involved and yes then that has to be done. Right? But I want to understand that part as well. + +Artur Troian: Yes, so the easiest part from that is the least can be closed by multiple ways. So you either do deployment so from client perspective from Talent perspectively other than by deployment calls or is from provider side the provider discloses be and that's So as soon as provider issues transaction be close. The list will be closed. So yeah and that provider actually doesn't. + +Artur Troian: For let's say if the lease Not speeding up for whatever reason with image is wrong. Button doesn't exist. So what writer will do in five minutes if they lease is not healthy. It will actually close it by issues because + +Deval Patel: + +Deval Patel: Okay, another question. So when you say the least is not healthy right does it count as well that if the Manifest is not Properly deployed or it does not so let's say I want the leaves right? + +Artur Troian: you + +Deval Patel: But I do not send the Manifest or the Manifest has some problems like what happens at that point of time. + +Artur Troian: if the Manifest it has some problems that's the different issues. So by the time manifest is created. It's been validated that it's actually correct. So if manifest is incorrect, that means that the code base is wrong is having issues. yeah, but anyways to answer your question that counts as well. So let's say something wrong even the money fast. The provider will close the degrees by issues be closed. + +00:35:00 + +Deval Patel: Okay, okay, and we want to expose that using a transaction that will be initiated by provider through let's say command line or the UI from printer. + +Artur Troian: It is done automatically by the provider the issue is right now. They provider is the main user of the blockchain key and there's a thing called sequence number. So if you're gonna issues that command manually that will break sequence number and provider will click out. So there is a way but it's just not going to break things. So the beat close command must only come from the Provider Services in this stage. + +Deval Patel: Yeah, that's what I'm Okay providers service or… + +Artur Troian: but It's already implement. + +Deval Patel: okay, the very close command has to come from. + +Deval Patel: Okay. + +Artur Troian: Yeah, but that's already Implement. + +Artur Troian: I'm trying to understand what issues you're trying to solve here. + +Deval Patel: So we want to give the access to Providers where they will be able to close the lease if they find that lease is not appropriate for that. + +Artur Troian: I see. So the best way to do it right now is they have to stop providers Services first. Then they can be close. And then restart the Provider Services, that's unfortunately on the way to do that at this stage. Yes, we fully aware of this issues and we've been thinking require actually quite a while. On the ways to fix it. So this will be once we finish Provider Services split. And that's planned for some time next year. so right now the best way to do this is to stop the kill the lease by doing Because we start the Provider Services and then Provider Services will see that list has been closed. + +Artur Troian: And it will clean up the namespace for this particular deployment. So that's the way it has been done right now. + +Deval Patel: Okay, so that gives me clear idea what I need to do. And we really Implement that for now, right and you say we gonna change that in a cleaner way of doing things right at that point. We will make the new requests or make our modification as well right on the content conversation side. That's Okay with that. + +Deval Patel: Yeah, thank You are talking about the office hours. If I am done with all my questions and… + +Artur Troian: yeah, do you still want to we can do this kick on the Small version of office hours today and… + +Deval Patel: yeah. + +Artur Troian: show how to run the deployment how to debug things. + +Deval Patel: Yeah, but that sounds Really good. + +Artur Troian: Okay. + +Tyler Wright: as long as our arthritis + +Artur Troian: Yeah, so yeah. + +Artur Troian: just yeah. + +Artur Troian: it's actually probably. Let me just spare the screen. Is that on the second? + +Artur Troian: All right. let me know if you say my screen right now. Alright, so I'll just + +Artur Troian: so I'm gonna go to provider. + +Deval Patel: I'm not doing the commands with you because the recording will be available. So I'll be able to do that, on my own screen,… + +Artur Troian: Yeah, you don't need to do it. I'm gonna be showing it all… + +Deval Patel: right? Yeah. Okay. Yeah. + +Artur Troian: how to do that and… + +Deval Patel: Thank you. + +Artur Troian: okay, so, Wonder + +Artur Troian: build the example is runcube let's open. we need pretty much free internal Windows, right the top one will be. So let me just make sure everything is clean. + +00:40:00 + +Artur Troian: Up Foster. the + +Artur Troian: my Joker is dark little second. Let me Learn all that. okay. I'd like to sit down the previous cluster Starting from completely clean stage. The first command to issue is made you poster setup and this command takes some time maybe five minutes actually, maybe less what it does is it deals all of the binaries and Docker images required the local testing so for other services it builds nodes + +Artur Troian: Then what it does is it sets up, generates certificate all the keys and blah blah blah pretty much to spin up local version of the blockchain this one radiator. in Paradise Genesis and then + +Artur Troian: Right now so yeah back up on that. So first of all it builds Akash binary over here compiling a car you can see it builds it from vocal sources. Then what it does is it generates Keys then it creates the validators that then it adds the Genesis what it is. The Genesis is alright. and after that at some tokens then generates the For provider and the clients I mean until as keys. Once the account that we're going to use to deploy. When there's Genesis one more time. Then it builds the provider. + +Artur Troian: And then after that it spins up actual kubernetes cluster. Right now, so it uses time to do that. It starts control deploys whatever needed for full development process embody diets that the required deployments like engine X and other things are up. + +Artur Troian: it spins up the persistent storage if you need that and yeah, this process the last step of making sure that the nginx controller is updating sometimes Minter tool. and + +Artur Troian: Yeah, let's say now we are done. so what it does now it Applause the are images that we just build to their kind cluster and the cluster is fully up. So next command you to start the blockchain mode. We do that by doing that make node run. And it starts making blocks. So the next step will be to create the provider. So you use for that obviously separate terminal window do create + +Artur Troian: second. Yeah, right. + +Artur Troian: And it just creates the provider. until the transaction is successful so we can do make I believe very provider you can see that we have provider on the blockchain now, can this actually starts Provider Services. It was the inventory and it actually ready to accept this. So what we do right now is my deployment rate So this command runs if we take a look here? there is I believe one second. Let me double check file which as the past I believe it's gonna use this That one is yellow. + +00:45:00 + +Artur Troian: This is the deployment that you'll use to create to test. So if you do make deployment create by default, the sequence is one. So the sequence can be set as the environment as a variable. So you do like that. So this is the default one always and if you create my multiple deployments you just set to three. Four and so on so forth. and if the sequence is other than the zero, you have to include that in please create and deployment goals. I will show it in a second. So you do makeup on and You see it deployed deployment yellow and the provider you can watch here that provider actually code that. Feel it was a ration and submitted beat. So if you do make + +Artur Troian: query did You can see the Obviously the resources offer this is the recent feature from minute 8 and we're ready to create list. So what we do is make these And Chris leaves and starts to the coil. so we can see that it received manifest. + +Artur Troian: the lease is up if you look sitcl + +Artur Troian: Get an ass. To be here and it's not why? yeah, I didn't lose that manifest. That's so we do mates and manifest. + +Artur Troian: It's almost pass and you can see there manifest received on the provider Windows names and completely deploy. So if you do you get them spaces use our lease. In space and you can do Cube Ctrl and that ointment. so it has two Services as you can see one nine new and one train web and you can stay ports and if you have and point there, you can do the proxy and access it with the browser or see URL to close deployments that we do is make the tournaments close and it will shut it down. + +Artur Troian: Sort of completely closely, then you can see the provider actually immediately reacted there on that. So it unreservation and completed turn it down. You can check the spaces. Once again, you can see the reasoning space is gone. So let's say you want to create new deployment this because the sequence one already been used by this key. We need to specify different one. So it doesn't have to be consecutive. It has to be unique just all the time. For this sake of example, we're gonna use the sequence number two. Let's say the sequence to my deployment. Okay, I'm gonna use the same. manifest for that so the sequence tool it created deployment you can see for provider immediately reacted and submitted for humans and submitted bid as soon as we have beat do. + +Artur Troian: sequence to make a list rate that has been done provided is happy with that. And now we do once again the sequence to make money requests and manifest is done. And as you can see the provider completely deployed it and just double checking out. you see how business spaces using your namespace and you can + +Artur Troian: get services Yep, everything is hungry. And once again to pause it the sequence might want to close. + +Artur Troian: That's pretty much it. And to stop the provider into this Ctrl C that control see that and everything is stopped. Yes can go ahead. + +Scott Carruthers: Yeah, I just want to reiterate what you mentioned earlier Archer. So the link that I shared. In the chapter this meeting are detailed notes for this exact process. This is a + +00:50:00 + +Scott Carruthers: Documentation that I link to is currently I'm linking to my own GitHub repository. eventually I'll live on the cost documentation site, but we're awaiting they migration to docs 2.0 what did make sense to go through starting to bring this live on the current platform months or way until it's migrated to Not only this but all the developer documentation that created at the beginning of the year will be there. So Deval or anyone else is you're going through that documentation if you have any suggestions, it's definitely documentation that isn't vetted I think it's very clean and very thorough it just has by many people so clarific Answer additions that we can make please. Let me know. everything that Archer just went through from the creation of the local blockchain to spending up a local validator to creating the + +Scott Carruthers: provider and then also the steps make amends to create the deployment and send the Manifest and all that is detailed in that documentation. yeah. I mean as I let me know if you recommendations on where there could be improvements but should be pretty straightforward to use. + +Deval Patel: Yeah. Thank you Scott. And thank you Arthur for the demo. Thank you very much. + +Tyler Wright: Deval I'm sure you. + +Artur Troian: No problem. Yeah, all right. yeah. Thanks. + +Tyler Wright: Yeah, I was just gonna ask the ball. I know that you may have other questions. I really appreciate that demo and I'm sure we'll do more of those over time if anybody here has any specific topics that they want. Archer Scott Andre or others to do more of a deep dive in and pertains to either the code base or intricacies of using the class Network. Please feel free to message me or throw something in six support so we can add those as potential agenda items. Otherwise again, we do want to leave some time for folks to ask specific questions that are more Technical and give people the opportunity to do that with regularity. developers any other questions that you may have at this point. I know we only have five minutes left before some of us have to run over to Sig providers. + +Deval Patel: No, I'm good. Thank you. + +Tyler Wright: awesome again Archer much appreciate walking through the support repo and any issues that needed triashing. Thank you again to all the Insiders that give support around the clock to new and existing users that are interacting with the cash. I know it improves The Experience so much appreciate All again. Sig providers is coming up in about four minutes where we'll get an update on all the work to pray to our team has done this year. + +Tyler Wright: The kind of a look ahead on some of the things that they're thinking about for 2024 as well as I know that Benjamin is going to be talking more about his pre-draft discussion around an auditor. looking forward to those discussions in Sig providers. But again, thank you very much for joining. We'll meet up again in two weeks. I think after that we might take a couple of weeks off for the holiday. So if anybody has any specific things that need covering let's look to cover those in two weeks, but again much appreciate everyone's time today, and hope everyone has a great rest of the day. We'll talk soon. + +Scott Carruthers: Thanks, everyone. + +Tyler Wright: Thank you. + +Deval Patel: Thank you. + +Meeting ended after 00:53:48 👋 +