fedora_cloud_meeting
LOGS
15:11:29 <davdunc[m> #startmeeting fedora_cloud_meeting
15:11:29 <zodbot> Meeting started Thu Apr 27 15:11:29 2023 UTC.
15:11:29 <zodbot> This meeting is logged and archived in a public location.
15:11:29 <zodbot> The chair is davdunc[m. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions.
15:11:29 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:11:29 <zodbot> The meeting name has been set to 'fedora_cloud_meeting'
15:11:52 <Eighth_Doctor> .hello ngompa
15:11:53 <zodbot> Eighth_Doctor: ngompa 'Neal Gompa' <ngompa13@gmail.com>
15:12:03 <davdunc[m> #topic roll call
15:12:21 <davdunc[m> .chair ngompa
15:12:21 <zodbot> ngompa is seated in a chair with a nice view of a placid lake, unsuspecting that another chair is about to be slammed into them.
15:12:31 <Eighth_Doctor> huh
15:12:36 <Eighth_Doctor> that's new
15:12:48 <davdunc[m> #chair Conan Kudo
15:12:48 <zodbot> Current chairs: Conan Kudo davdunc[m
15:12:58 <Eighth_Doctor> #chair Eighth_Doctor
15:13:15 <Eighth_Doctor> oh right, not actually chair
15:13:16 <davidduffey[m]> here
15:13:27 <davdunc[m> #chair Eighth_Doctor
15:13:27 <zodbot> Current chairs: Conan Eighth_Doctor Kudo davdunc[m
15:13:29 <nzwulfin> .hello
15:13:29 <zodbot> nzwulfin: (hello <an alias, 1 argument>) -- Alias for "hellomynameis $1".
15:13:34 <davdunc[m> #chair davidduffey
15:13:34 <zodbot> Current chairs: Conan Eighth_Doctor Kudo davdunc[m davidduffey
15:13:40 <Eighth_Doctor> #unchair Conan Kudo
15:13:40 <zodbot> Current chairs: Eighth_Doctor davdunc[m davidduffey
15:13:47 <davdunc[m> chair nzwulfin
15:13:52 <nzwulfin> .hello nzwulfin
15:13:52 <zodbot> nzwulfin: Sorry, but user 'nzwulfin' does not exist
15:14:03 <Eighth_Doctor> #chair nzwulfin
15:14:03 <zodbot> Current chairs: Eighth_Doctor davdunc[m davidduffey nzwulfin
15:14:24 <Eighth_Doctor> nzwulfin: .hello uses FAS ID
15:14:25 <davdunc[m> my octal foo is missing today.
15:15:08 <davdunc[m> Eighth_Doctor: hence I am missing a closing bracket.
15:15:51 <mhayden[m]> .hello2
15:15:51 <zodbot> mhayden[m]: Sorry, but user 'mhayden [m]' does not exist
15:16:01 <mhayden[m]> .hello mhayden
15:16:02 <zodbot> mhayden[m]: mhayden 'Major Hayden' <mhayden@redhat.com>
15:16:07 <mhayden[m]> πŸ€¦β€β™‚οΈ
15:16:11 <mhayden[m]> i am a matrix noob
15:16:16 <nzwulfin> .hello mmicene
15:16:17 <zodbot> nzwulfin: mmicene 'Matt Micene' <nzwulfin@gmail.com>
15:16:36 <nzwulfin> i just need to remember all my user name croxxovers....
15:16:49 <davdunc[m> haha. I tell you!
15:16:57 <davdunc[m> #chair mhayden
15:16:57 <zodbot> Current chairs: Eighth_Doctor davdunc[m davidduffey mhayden nzwulfin
15:17:01 <mhayden[m]> πŸͺ‘
15:17:26 <davdunc[m> obudai: are you joining us?
15:18:10 <obudai[m]> .hello obudai
15:18:11 <zodbot> obudai[m]: obudai 'OndΕ™ej Budai' <obudai@redhat.com>
15:18:21 <davdunc[m> #chair obudai
15:18:21 <zodbot> Current chairs: Eighth_Doctor davdunc[m davidduffey mhayden nzwulfin obudai
15:18:43 <davdunc[m> thanks for your patcience! sorry I was late to the party.
15:19:08 <davdunc[m> #topic Action items from last meeting
15:19:36 <davdunc[m> mhayden and team to look at getting walagent into the Fedora cloud kickstart
15:19:53 <mhayden> so it's actually already in the azure kickstart file
15:20:00 <mhayden> which is awesome
15:20:14 <davdunc[m> excellent.
15:20:29 <mhayden> the changes to fedimg to get it built are done and releng said to poke them when we are ready for new artifacts
15:20:37 <davdunc[m> now we just need to get the project naming and the uploads completed.
15:20:40 <mhayden> now we need the change request approved :)
15:20:45 <Eighth_Doctor> we've got both x86_64 and aarch64 images?
15:21:02 <mhayden> #link https://fedoraproject.org/wiki/Changes/Fedora_Images_On_Azure
15:21:24 <mhayden> Conan Kudo: i think i added arm to that one -- let me go back and check
15:21:33 <davdunc[m> Ah yes. Thanks Conan Kudo for updating that!
15:21:59 <Eighth_Doctor> we should make sure we've got aarch64 images for everything
15:22:12 <Eighth_Doctor> with the new website, it should be much easier to make sure people can find them too
15:22:12 <mhayden> Conan Kudo: i missed that change for aarch64. I'll make a PR
15:22:18 <Eighth_Doctor> πŸ‘οΈ
15:22:24 <davdunc[m> Conan Kudo: I have a meeting topic for that.
15:23:27 * davdunc[m mhayden to draft a change proposal for removing standard
15:23:27 <davdunc[m> storage images https://pagure.io/cloud-sig/issue/295
15:23:36 <Eighth_Doctor> excellent
15:23:41 <davdunc[m> That's done too!
15:23:57 <mhayden> PR for aarch64 on azure is in the queue: https://pagure.io/pungi-fedora/pull-request/1169
15:24:05 <davdunc[m> excellent.
15:24:14 <davdunc[m> #link https://pagure.io/pungi-fedora/pull-request/1169
15:25:23 * davdunc[m nzwulfin and mhayden to sync up on changing gp2 > gp3 for
15:25:23 <davdunc[m> Fedora cloud image
15:25:52 <nzwulfin> we got a PR submitted to fedimg for the change https://github.com/fedora-infra/fedimg/pull/166
15:26:11 <nzwulfin> i think the docs still need work but mhayden said 'ship it!' :D
15:26:47 <mhayden> yeah, might as well get the change moving along. i'm happy to work on docs. likely don't need much here, perhaps just release notes?
15:26:51 <Eighth_Doctor> should gp2 just be outright not supported?
15:26:55 <Eighth_Doctor> in fedimg
15:27:08 <davdunc[m> it's fine.
15:27:12 <Eighth_Doctor> okay
15:27:38 <nzwulfin> the PR is a s/gp2/gp3/g basically
15:28:00 <davdunc[m> I don't see a reason really to remove it. It would be more difficult to guard against it as it's just a text config.
15:28:18 <mhayden> anyone could use gp2 at instance creation time if they really want to
15:28:28 <davdunc[m> it's in the storage device definition and it's really just a default.
15:28:37 <davdunc[m> exactly!
15:28:48 <nzwulfin> right, the only real code change is the default if unset is gp3
15:29:18 <davdunc[m> yes.
15:29:29 <davdunc[m> but gp3 is the better choice.
15:29:31 <Eighth_Doctor> I don't care that much, let's 🚒it
15:29:34 <nzwulfin> take a look at the PR, happy to make chnages
15:30:03 <davdunc[m> nzwulfin++
15:30:17 <nzwulfin> i'm really not even sure it's on the right branch rbh
15:30:34 <davdunc[m> :) that's something we can clear up with infra.
15:31:01 * nzwulfin apparently cain't type either
15:31:15 <davdunc[m> it's going aroudn
15:31:32 <mhayden> so much shipping. love this.
15:32:12 <Eighth_Doctor> mh
15:32:42 <davdunc[m> I got to one I don't remember and didn't get phrased well - by me. Action: davdunc chase down the modifications in the ansible and then
15:32:42 <davdunc[m> modify the flag if mhayden and team don't beat him to it.
15:33:13 <mhayden> i'll leave that one for you πŸ˜‰
15:33:24 <davdunc[m> I'll do it!
15:33:31 <nzwulfin> real life CTF
15:34:16 <davdunc[m> #action davdunc will review the raw notes from 2023-04-13-15 and determine what he was going to do to the ansible for infra before mhayden beat him to it.
15:34:36 <davdunc[m> okay.
15:34:58 <davdunc[m> #topic Story for F39 Release Eng cycle.
15:35:30 <davdunc[m> mattdm: wrote up a discussion article on how we should have a story for each release cycle.
15:35:48 <davdunc[m> I wanted our story for F39 to be aarch64 everywhere.
15:36:51 <davdunc[m> So we have been woefully neglectful of putting up the aarch64 configurations over the past couple of relases and I want to make sure they are available everywhere we can use them, from Vagrant all the way out.
15:37:21 <davdunc[m> Anyone have a story they would like to see as a focus for our release work towards F39?
15:38:01 <davdunc[m> I'll wait for the network to catch up.
15:38:18 <Eighth_Doctor> well, this cycle is about bringing everything to all the clouds, I guess?
15:38:23 <Eighth_Doctor> well, the big three ones anyway
15:38:39 <Eighth_Doctor> and private clouds
15:38:47 <davdunc[m> #link https://discussion.fedoraproject.org/t/objective-review-each-edition-has-a-story-for-each-release/79324
15:38:55 <mhayden> i keep thinking about more cloud integrations built into fedora -- things that make it easier to interact with the cloud under the instance and the services offered in that cloud
15:39:05 <nzwulfin> for the new folks in the room, can you say more about where to find said configs?
15:39:12 <mhayden> i've been working on that with vultr + hcloud CLI tools
15:39:42 <Eighth_Doctor> I filed a ticket a while back about also looking at Oracle cloud too
15:39:49 <Eighth_Doctor> but I've done no investigations on the difficulty of doing so
15:40:06 <Eighth_Doctor> I did it because people asked because of its rather generous free tier for ARM
15:40:11 <davdunc[m> yep and Oracle has aarch64 hardware.
15:40:35 <davdunc[m> we should get it there, but I don't know that we are ready for that tooling wise.
15:41:50 <davdunc[m> nzwulfin: you are talking about the kickstarts?
15:41:54 <davdunc[m> #link https://pagure.io/fedora-kickstarts
15:42:27 <davdunc[m> mhayden: Can we add the vultr + hcloud CLI tools to the fedimg work?
15:42:27 <nzwulfin> whatever you meant by ' woefully neglectful of putting up the aarch64 configurations '
15:42:39 <Eighth_Doctor> that's us not building and releasing images
15:42:42 <obudai[m]> I did some experiments with Oracle Cloud before. The standard KVM images work fine on VMs. Bare metal needs more work because of iSCSI.
15:43:01 <obudai[m]> I don't expect many folks to use bare metal with Fedora, though.
15:43:21 <Eighth_Doctor> we have the definitions, but we have to tell pungi to create the images for those targets
15:43:22 <davdunc[m> That's making the pungi configs for the specific clouds and building the images as part of the release.
15:43:24 <mhayden> davdunc[m: possibly
15:43:30 <Eighth_Doctor> that's what mhayden did for the Azure image
15:43:46 <mhayden> Hetzner has aarch64 now but they build their own Fedora image
15:44:09 <davdunc[m> yea. so we can do that for GCP, for Azure, and Vagrant, etc.
15:44:10 <Eighth_Doctor> eh?
15:44:12 <mhayden> davdunc: by "add those tools to the fedimg work", what are we looking to do?
15:44:39 <davdunc[m> mhayden: looking to automate the uploads alongside the others.
15:44:57 <obudai[m]> have we tried to reach out to Hetzner guys? Using the official Fedora images would be super-cool.
15:45:05 <mhayden> davdunc: uploading images for vultr is supported but hetzner doesn't allow external image uploads right now
15:45:08 <davdunc[m> let's talk about that after May 1 (after I have turned in my presentatinos for Summit) :)
15:45:16 <davdunc[m> I see.
15:45:30 <mhayden> i still use my rescue/snapshot hetzner hacks for getting fedora in there πŸ˜‰ πŸ‡©πŸ‡ͺ
15:46:29 <davdunc[m> okay. I resonate with Conan Kudo 's sentiment that we need to be working on getting Fedora everywhere.
15:46:47 <davdunc[m> He focuses on the big picture as usual. :)
15:47:05 <Eighth_Doctor> lol I'm the big picture guy now? :)
15:47:40 <nzwulfin> cue buzz lightyear meme with conak kudo's face
15:47:45 <mhayden> at least an 8x10" picture
15:47:51 <davdunc[m> heh. you're always thinking strategically. . . when you don't overwhelm yourself with deliverables. :)
15:48:19 <davdunc[m> he's a 20-200mm lens kinda guy.
15:48:27 <Eighth_Doctor> heh
15:48:37 <Eighth_Doctor> one day it'll be useful :)
15:49:15 <davdunc[m> well, I kinda think that your idea is a great North Star. I
15:49:32 <davdunc[m> I think it's basically the introduction to the PRF.
15:49:46 <davdunc[m> or PDR.
15:50:32 <mhayden> a PDF of the PDR?
15:50:47 <davdunc[m> so I was trying to tone it down a little and make it achievable and associated with  your goal.
15:50:53 <Eighth_Doctor> :)
15:51:17 <davdunc[m> yes. in the PDF of the PDR.
15:52:03 <davdunc[m> PRD is aroudn heer somewheer
15:52:09 <davdunc[m> #link https://fedoraproject.org/wiki/Cloud/Cloud_PRD
15:52:52 <nzwulfin> do we have a 'needit / got it' list for 'everything' and 'everywhere'?
15:53:42 <davdunc[m> nzwulfin: we don't have one that we maintain, but it would be nice to have a plan of record and that PRD is definitely the place where our directives and goals go.
15:53:48 <Eighth_Doctor> I think we have tickets in the cloud-sig tracker of some of our gaps
15:54:21 <davdunc[m> yea, but it would be nice to have a list we refer to. Probably something like the Infra team uses.
15:54:27 <Eighth_Doctor> yeah
15:54:51 <Eighth_Doctor> and let's not forget that we're working on our first layered cloud variant with the cloud-kde version
15:55:10 <davdunc[m> well now, that's right.
15:55:40 <davdunc[m> so we have a lot of irons in the fire. Which one do wee think is sufficiently big enough to be our story?
15:56:00 <davdunc[m> Let's consider it and come back to this next week.
15:56:06 <Eighth_Doctor> sounds good
15:56:11 <nzwulfin> sounds like we need a table regardless
15:56:42 <davdunc[m> should I just assign an action to nobody for that?
15:56:46 <Eighth_Doctor> lol
15:56:52 <davdunc[m> we can just work on it over email. :)
15:56:56 <nzwulfin> i'll take it :)
15:57:21 <nzwulfin> that way i can feel less bad about not starting the documentation discussion
15:57:35 <davdunc[m> #action nzwulfin to build out a needit / gotit table for the cloud team.
15:57:53 <mhayden> guess you can say that table is a 'needit' 🀭
15:57:55 <mhayden> 3 mins!
15:58:02 <nzwulfin> lol
15:58:15 <davdunc[m> #action cloud team to discuss story in next meeting and over email
15:58:21 <davdunc[m> #topic open floor
15:58:38 <davdunc[m> quick! get in your topic!
15:59:39 <davdunc[m> Spoke to Zach at GCP, we need support for the gve nic module.
15:59:39 <davdunc[m> That's going in a change proposal
15:59:57 <davdunc[m> #action davdunc to complete change proposal for GVNIC feature on GCP
16:00:28 <davdunc[m> #info for those of us who have the opportunity to go to summit, let's meet up.
16:00:35 <nzwulfin> :(
16:00:44 <davdunc[m> okay. we are over time.
16:00:52 <davdunc[m> #endmeeting