14:04:03 <pwhalen> #startmeeting Fedora IoT Working Group Meeting
14:04:03 <zodbot> Meeting started Wed Nov 20 14:04:03 2019 UTC.
14:04:03 <zodbot> This meeting is logged and archived in a public location.
14:04:03 <zodbot> The chair is pwhalen. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:04:03 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
14:04:03 <zodbot> The meeting name has been set to 'fedora_iot_working_group_meeting'
14:04:11 <pwhalen> Good morning folks, who's here?
14:04:24 <puiterwijk> Morning
14:04:44 * tdawson is here.
14:04:54 <pwhalen> #chair pwhalen pbrobinson bcotton tdawson puiterwijk
14:04:54 <zodbot> Current chairs: bcotton pbrobinson puiterwijk pwhalen tdawson
14:04:54 <pwhalen> #topic roll call
14:05:35 * satellit listening
14:06:03 * pwhalen gives a few minutes for others to join
14:08:01 <pwhalen> #topic 1) ==== Working Group process and admin  ====
14:08:01 <pwhalen> #link https://docs.fedoraproject.org/en-US/iot/
14:08:23 <pwhalen> Does anyone have anything for admin?
14:08:42 <tdawson> Not I
14:08:48 <pwhalen> Nor I
14:09:20 <pwhalen> #info Nothing new.
14:09:47 <puiterwijk> No admin stuff from me
14:10:04 <pwhalen> #topic 2) ==== Fedora 31 status ====
14:10:05 <pwhalen> #info Latest compose available for testing
14:10:05 <pwhalen> #link https://kojipkgs.fedoraproject.org/compose/iot/Fedora-IoT-31-20191110.0/compose/
14:10:16 <pwhalen> so its been a while since we had a compose
14:11:17 <pwhalen> looks like F31 is failing on aarch64 runroot tasks, which might be the same issue we've been seeing in vanilla Fedora
14:12:01 <pwhalen> It is, read only mnt/koji
14:12:22 <puiterwijk> So, one case of readonly mnt/koji I fixed the other day
14:12:43 <puiterwijk> (Basically, there were new runroot builders added that didn't have their mount permissions set correctly)
14:12:58 <pwhalen> last attempt was on the 18th
14:13:05 <pwhalen> ok, so it should be good now, thanks puiterwijk
14:13:07 <puiterwijk> Right. I fixed it after that
14:13:11 <pwhalen> perft
14:13:24 <puiterwijk> I think Peter just didn't get around to kicking off a compose since then due to council hackfest
14:13:48 <pwhalen> right, so likely should have a new compose once he gets back home
14:13:52 <puiterwijk> Yep
14:14:00 <pwhalen> Anyone hitting any issue in f31?
14:14:15 <pwhalen> or anything else for f31?
14:14:34 <puiterwijk> I forget when I last tested f31, but at that point auto-unlocking (clevis) worked.
14:14:54 <puiterwijk> However, since then someone pushed tpm-tools 4.0 to Fedora 31, so if that got into a working compose, that's broken for now
14:15:01 <puiterwijk> (A fixes clevis is submitted, but not yet stable)
14:16:07 <pwhalen> #info tpm-tools 4.0 in Fedora 31 breaks automatic unlocking with Clevis. To be fixed in an update.
14:16:33 <pwhalen> thanks, I tried to use it again on my fitlet and it still fails in the same way :(
14:17:32 <pwhalen> #info Fedora 31 composes have been failing due to runroot issues. Should be resolved on next attempt.
14:18:02 <pwhalen> #topic 3) ==== Fedora 32 status ====
14:18:03 <pwhalen> #info Latest compose available for testing
14:18:03 <pwhalen> #link https://kojipkgs.fedoraproject.org/compose/iot/Fedora-IoT-32-20191107.0/compose/
14:18:19 <pwhalen> Similar issue here, compose is a little stale.
14:18:23 <puiterwijk> Looks like the last f32 compose failed due to... other amusing errors
14:18:28 <pwhalen> oh
14:18:48 <puiterwijk> Might also be read-only /mnt/kolji actually, just shown differently
14:19:04 <puiterwijk> Yeah, looks like it actually. Sorry for that previous sentence.
14:19:23 <puiterwijk> So yeah, f32 seems to be in the same state as f31, waiting new compose
14:20:18 <tdawson> Do we know if the initial user setup is available yet?
14:20:41 <tdawson> That sounds weird.  Basically, can we log in on a fresh image?
14:20:44 <pwhalen> does look different, looks like armhfp is the reason but the log is empty..I'll take your word for it :)
14:20:49 <puiterwijk> tdawson: on f32, initial-setup is gone, and we are trying to make Ignition work.
14:21:11 <tdawson> Yep, just wondering if it's ready yet.
14:21:13 <puiterwijk> So, for that, I hope to send an email to the mailing-list soon, as soon as I can figure out some.. collissions
14:21:34 <pwhalen> puiterwijk has been working on it. I think for now your best bet is to use your ssh key
14:21:40 <puiterwijk> So, I can give you ad-hoc instructions if you ping me on IRC. We have another solution incoming that I'm trying to work out
14:22:08 <tdawson> OK
14:23:01 <pwhalen> tdawson I think there is also an issue with selinux and machine-id on the images. Hit it on vanilla f32, selinux context is wrong
14:23:32 <pwhalen> should be fixed soon by the looks of the BZ
14:24:54 <pwhalen> That's all I had for f32, anything else?
14:25:22 <tdawson> Nothing from me, ya'll answered my questions.
14:26:12 <pwhalen> #topic 4) ==== Open Floor ====
14:26:40 <pwhalen> Anything for open floor?
14:26:59 <puiterwijk> Nope, not me at this time
14:27:49 <tdawson> Nothing from me.
14:28:24 <pwhalen> Nor I. Thanks for coming folks
14:28:27 <pwhalen> #endmeeting