Skip to content

Commit

Permalink
Update 019-2023-10-11.md
Browse files Browse the repository at this point in the history
Signed-off-by: Drinkwater <[email protected]>
  • Loading branch information
brewsterdrinkwater authored Oct 18, 2023
1 parent c942286 commit 50e5488
Showing 1 changed file with 138 additions and 4 deletions.
142 changes: 138 additions & 4 deletions sig-support/meetings/019-2023-10-11.md
Original file line number Diff line number Diff line change
Expand Up @@ -12,12 +12,16 @@

- Date: Wednesday, October 11th, 2023
- Time: 07:00 AM PT (Pacific Time)
- [Recording]
- [Recording](https://4bnijv3wt46kxmijhgq67y4tjt55h36c3yz7ubtakslbdyjord2q.arweave.net/4FqE13afPKuxCTmh7-OTTPvT78LeM_oGYFSWEeEuiPU)
- [Transcript](#transcript)

# Participants


- Andrey Arapov
- Naman Kabra
- Scott Carruthers
- Scott Hewitson
- Tyler Wright


# Notes
Expand All @@ -26,8 +30,7 @@
- Scott led the team through triaging the isues.
- Assigned issue 130 to Andy. He will investigate further and close.
- Andrey talked about issue 129. This was affecting Europlots provider.
- Documentation and code is needed. This will be signed to

- Documentation and code is needed.


## Support
Expand All @@ -47,3 +50,134 @@
- Artur will look into AuthZ issue 88.

# **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 to six supports biweekly number 19. During these support biweekly, special interest group meetings participants go through active and open tickets and the support repo. The group triage is those and assigned to individuals, whether it be from the overclock team, or from the community, to help solve those issues. And then again, if anybody has any questions about any issues that are open, that could feel free to ask questions here. After that, we talked to folks that are usually on the call from the insiders of vanguards that are interacting with people from the community and discord telegram and elsewhere. And we just see if there's any support claims worth having more discussion about directly from the community.

Tyler Wright: I know our tour is not here today, so I'm gonna hand it over to Scott to kick us off with a riassing open issues. A new issues that have appeared in the last couple of weeks in the support repo. So I'll hand over to you Scott.

Scott Carruthers: Alright, thanks Tyler. Let me share my screen.

Scott Carruthers: Okay, is everyone able to see my Screen.

Andrey Arapov: Yeah.

Scott Carruthers: Okay, so I think this is going to be pretty quick. If we look at the issues that been open over the last couple of weeks, there's only a couple and I think these are pretty quick, discussion points.

Andrey Arapov: Yeah, I think I can just talk about them quickly. So the first one is the provide responsibility upgrades. do you need to explain details? What's the reason or just where we are? The state?

Scott Carruthers: Yeah, we could just since we don't really have anything,…

Andrey Arapov: The state. Yeah.

Scott Carruthers: we can just go very quickly. So that's this the docs for this is definitely open or have been completed. The only reason I kept open was for the second bullet point.

Andrey Arapov: that's true. So basically I just need to double check the sandbox I already checked the production is good and the sandbox basically when and upgrades are enabled and they're always some enabled by default on Ubuntu servers. and, the silently can upgrade some packages and break some stuff. We haven't seen this previously because we haven't been using convidia drivers. But now since that, we have providers that support GPU, including ours, whenever, in media driver gets upgraded

Andrey Arapov: And get stuck this nvdp container and the Kubernetes cluster kind of gets in the state where I just like, waiting for something, but doesn't get through and the only way to solve it, you need to reboot the note, But doesn't impact they already running application so that's why we created this dog. Scott posted it in move to Dogs a cushioned network thing, thank you. and they providers so that they disabled the content upgrades

Scott Carruthers: Yeah, and kind of being proactive. And every time during this support session, we also talk about the support issues that have occurred over the last couple weeks. I would say that this is probably the most significant thing this come up on our radar over the last couple weeks so we can talk about it a little bit more during that portion as well. But yeah, I would say that some providers being impacted by this may been our most significant issues. We're pretty late support activity over the last couple of weeks.

Andrey Arapov: Okay.

Scott Carruthers: So yeah, I think we're good with that. And then we also have the provider says, insufficient capacities, sporadically.

Andrey Arapov: Yeah, I think it's possibly was related to this, and they ended upgrades was enabled on the provider because I've seen it the providers meeting constantly, but I just need to look up the locks once again just to double check this, and then I can close it.

Scott Carruthers: What it makes sense…

Andrey Arapov: So, Yeah.

Scott Carruthers: since you want to go through that additional validation, before we do look any more code investigation or investigating. This deeper DME assign us to you for now and…

Andrey Arapov: Yes, yes. Thank you.

Scott Carruthers: you can okay. Sure.

Scott Carruthers: Okay, and then, I think our area,…

Andrey Arapov: The airport.

Scott Carruthers: our last issue is Euro plots, starts bid in

Andrey Arapov: I just checked on what's on the bottom of this. I think I started it.

Scott Carruthers: Okay. Yeah. And that you said that I think I did read through this.

00:05:00

Andrey Arapov: What we see next again.

Andrey Arapov: Yeah,…

Scott Carruthers: I won't

Andrey Arapov: yeah, I remember basically there is a metal B container that is for ipleasing. And if it breaks, basically, it breaks the provider because he told me someone that's me on the call. Because we basically have a cash provider that is expecting, some information, it's talking to the operators and one of the ators, the inventory, operator and enter service and it basically cannot get iPS and because of that just provider

Andrey Arapov: Basically, it doesn't solve self recover after you fix this metal will be stuff. So it's kind of interconnected and I think I already asked after about it in Internal Death channel some other day ago. So probably need to create the follow-up for this just to make sure the provider can self recover basically from this issues and update the Presentation Troubleshooting section because here all the comments are necessary for interval shooting this issue. Yep.

Scott Carruthers: Okay, sounds good. Do you want me to? So it sounds like there's some documentation and also some code investigation needed. So as far as assignment, should I assign both UN Archer to this issue.

Andrey Arapov: You can assign trust me and I will create a follow-up for Arthur that I was excited to him. So, Okay,…

Scott Carruthers: Okay. Sounds good. Sure.

Andrey Arapov: thank you.

Scott Carruthers: And yeah, Tyler with that, I believe that reviews, all the issues since we met two weeks ago and just looking at the ones that appreciate that as well, just make sure there weren't any other ones that they weren't dealt with, during our last thinker that may have come in in a couple days after our last meeting unless we have assignment for everything and the span rate before our last two weeks as well. So yeah, I think as far as issues were in good shape

Tyler Wright: Perfect. Thank you Scott. Thank you, Andre.

Tyler Wright: Before I move on to an update from those that are actively supporting folks in discord and telegram. I do want to remind folks issues that go in the support, repo are related to the core products of again, the cash network. There is a separate repo that if anybody has any really issues related to console, they could put those issues there is also for folks that are using if you have any most related issues, I know Cloud most has been in transition and is now fully open source and within the Akash Network Organization on GitHub, feel free to drop an issues specifically into the cloud most repo. So that again, we can continue to maintain those three places where issues go and we can properly segment them, we will in the future again. Start to have Most issues as

Tyler Wright: Of this, if the number of issues grows. But for right now, again, support repo until the cash network is for issues related to the court product. And then there's separate issues for folks that are respectively using console and cloudmouse inside those repos. So, I just wanted to remind folks of that. Cool moving along at this point during the agenda. Again we usually hear from folks in the community in the insiders and vanguards. I know we have a lighter crew today. Scott's I'll hand it back over to you. And I know Andre also very active working with folks in the community and discord. So, I'll hand it back to you. I think to talk a little bit about the issues that you just talked about issues 131 as well. If anything else that you've seen over the last couple of weeks in the community,

Scott Carruthers: So as I was thinking of this session and wanted to highlight anything that's occurred over the last couple weeks of significance, it This is kind of been our state since we stabilize. I mean that sex and got through, GPU provider, builds the last I would say month has been A majority of support activity, is just assisting users with SQL issues or maybe some provider, build assistance, and i problems in the code base or code issues, but just assistance within the community. So that's certainly been a bulk of the activity. So, as I was thinking over the last couple weeks, the issues that we just tackled and discuss with the Nvidia GPU drivers, automatically updating, and that affecting

00:10:00

Scott Carruthers: A new bit activity. And the documentation that we covered is really the only thing of significance as far as an issue that I can think of that's come over up over the last couple weeks and that's back by our review of the issues and there really wasn't anything else creative from a code level. So again, with documentation is now in place for the disabled men of the unintended upgrades for Nvidia GPU drivers. So that should just be a part of the normalized provider bill process. I think we should be in good shape on that issue. And again I really don't have anything else to report as far as core issues as mostly just some handholding of new users and some provider builds. Assistance. Andy, can you think of anything else that you've supported over the last couple of weeks that would be worth mentioning?

Andrey Arapov: I guess maybe just a small thing but it's not an issues. Just rather the sinners provider of had made observation before. and again basically he noticed that he has high iops on the disc and we've checked together and this are just some customers decay based notes like kava or osmosis or aka they do have the right a lot on the disk and read from the risk. So this is normal behavior for customers decay based notes. So, if anyone any provider notices, this kind of thing, they can just use a IO top command. It's particular flags there and quickly identify which process is writing but it's like nothing unusual, really.

Andrey Arapov: And other than that, I just want to say that it's been quite some time since we haven't had any drastic, big issues other than the very old issues that I reported before.

Scott Carruthers: Yes I think to kind of wrap up. He has been a very stable period and again we're just assisting new users. And some users through image build processes, or provider bill processes, and things like that. But yeah, rather

Tyler Wright: Perfect, thank you very much. Again, just want to shout out, Scott and Andre who have been working very closely with the community and discord for a while. as the cash network continues to grow. I know that Overclock has a small team Andre and Scott both heavy number of responsibility. the vanguards and the insiders have been doing a great job of really being hands-on and helping out the community with again, getting started any specific issues. The goal is over time to continue to leverage, the community to really be hands-on and discord and other channels. So that again folks Andre and Scott can continue to build new features. Continue to again work closely with providers and advance the Akash network.

Tyler Wright: So I do want to continue to have folks whether it be listening in later know that they can participate. Again if you haven't joined the insiders program if you eventually want to become a vanguard, please feel free to join the inside program. Come to a sick community meeting get involved in discord. Certainly looking for more people that can help. with documentation whether it be just again, taking ideas that may become from the core team or vanguards and inside and insiders and helping to update documentation or again, as you see issues in the community, making sure that documentation is reflecting the most up-to-date stuff because again, we've been relying on Scott affair deal for that and Andre supporting. And again, we want to continue to have this be A team effort so that they can continue to work on other things. So I just want to shout that out.

Tyler Wright: It's also a natural segue naman. Who's on this call, I invited him here and if he has the ability, I would like him to continue to join this. moving forward on demand is part of the node, Ops team, note Ops is been building some getting started cooling, that leverages AI technology and it's been leveraging the documentation on a caution there. I know that they're also working on some other projects on demand. I know that we talked about potentially doing A short demo here. But I think this audience is already seen the product and it's already been testing it. I know that you did a demo for folks in SIG community yesterday which I know what very well and I know there's gonna be continued testing. I just maybe have you briefly just introduce yourself but naman if you are open to it this is a meeting that I think if you can come to in the future basically and during this biweekly meeting

00:15:00

Tyler Wright: The group, which is mostly folks from the core team and then inside is vanguards. But again, this is meeting, it's open to anybody, the group discusses open issues in the support repo related to core product, as well as talks about what's going on the community. I know that the tooling and Node OPs is trying to make it easier for developers to get started using a cache and obviously using other projects. But again specifically to Akash you're always thinking about and nodeops It's always thinking about ways to improve get started process. So this is a great meeting for you to be on the ground floor of what's going on. The community, what people are talking about? And this could lead to maybe some of those templates that are a part of your product, or again, streamlining, some of these things, whether it's

Tyler Wright: Many people asking the same kind of question as an example about sdls, then the Note ops tool can be improved to service But again, I think this is a great meeting for you to be at. And again, you can also hear what's going on in as real-time as fashion with documentation updates that affects support and just beyond. So naman, I don't want to again, have you do a demo because you've done when the past but just, maybe briefly introduce yourself to this group if you haven't already, because I know some folks, know you and I've been working closely with you, like Scott,

Naman Kabra: Yeah, absolutely. Thanks. Thank you so much. Tyler again, for inviting me as well. one of the very first six support girl, I'm attending but, really looks interesting to get some insights in terms of what's happening from the development side within the foundation and beyond as well in terms. So for all the contributors coming in, we're not actively trying to, simplify the developer onboarding at nodeops here and I've been actively working in this space since 2017. built out a couple of projects. One of them was acquired by circles and, was part of persistence team initially and then build out and NFT and currently, with no love sorrow objective as Tyler. Wrightly mentioned is to just, simplify the developer on boarding with two devs and

Naman Kabra: And software guys, looking to get started with, particular protocol and eventually just leveraging AI as an catalyst here, but the goal focus always remains on. We taking care of everything when it comes to infrastructure and developers can just focus on their use cases and applications. So yeah. And, really looking forward to launch this as early as possible.

Naman Kabra: You Tyler?

Tyler Wright: Thank you very much again. Usually there's a number of folks from the Insider vanguards community but I know a lot of folks are in Europe or just traveling around. But again naman, this is a great opportunity, a number of folks from the core team, like Scott's Archer, Andrey, a Neil sometimes attend this meeting and it's a little bit more technical. But again, you'll be on the ground floor of what are some of them open issues? What are some things that are the community is facing and hopefully you can help streamline. And again this is also a good group that is already been testing. I know that soon everyone here is already taking a look at nodes since been testing and again we'll continue to share that in discord and continue to get feedback.

Tyler Wright: All right. We've gone through the agenda. We've looked at open issues that haven't been triage in those issues have been triage. Thank you, Scott and Andre, we talked about support in discord and Telegramand, gave her brief intro of himself. And again, hopefully, we'll be able to continue to attend these meetings on demand. If anybody from that, has a go background and wants to help solve some issues for our costs. There are about toots available. So just if you want to pass that message along again you can work, I have more folks that want to work closely with the core team and help us.

00:20:00

Tyler Wright: I'll get more cloud parity. But again if there's nothing else then we can certainly adjo This meeting for today and then again I'll follow up with notes recording and transcript make those available and then we'll meet in two weeks. Does anyone else have anything else that they want to cover?

Tyler Wright: again, Scott much, appreciate you filling in and leading us to victory. I hope everybody has a great rest of the day and we'll certainly talk online. If anybody has any questions comments, please feel free to drop them in six support or again open up a support issue or again reach out in the community a number of channels but much appreciate everybody. Hope everybody has a great rest of the day.

Andrey Arapov: Thank everyone.

Tyler Wright: Thank you.

Naman Kabra: Thank you.

Meeting ended after 00:21:21 👋

0 comments on commit 50e5488

Please sign in to comment.