17:00:21 <mboddu> #startmeeting RELENG (2018-10-11)
17:00:21 <mboddu> #meetingname releng
17:00:21 <mboddu> #chair nirik tyll sharkcz masta pbrobinson pingou puiterwijk maxamillion mboddu Kellin dustymabe
17:00:21 <zodbot> Meeting started Thu Oct 11 17:00:21 2018 UTC.
17:00:21 <zodbot> This meeting is logged and archived in a public location.
17:00:21 <zodbot> The chair is mboddu. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:21 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
17:00:21 <zodbot> The meeting name has been set to 'releng_(2018-10-11)'
17:00:21 <zodbot> The meeting name has been set to 'releng'
17:00:21 <zodbot> Current chairs: Kellin dustymabe masta maxamillion mboddu nirik pbrobinson pingou puiterwijk sharkcz tyll
17:00:22 <mboddu> #topic init process
17:00:41 <nirik> morning
17:00:46 <mboddu> nirik: Morning, how goes?
17:00:48 <puiterwijk> hello
17:01:09 <mboddu> puiterwijk: Hallo :)
17:01:40 <nirik> living the life!
17:02:23 * sharkcz is partially here :-)
17:03:33 <mboddu> Okay, since sharkcz is here, lets start with alt arch updates
17:03:42 <mboddu> #topic Alternate Architecture Updates
17:04:16 <mboddu> So, I am +1 on bringing back 28 base container images for f28 now that they are working fine on 29 and rawhide
17:04:33 <mboddu> But the armhfp thing is still annoying
17:04:41 <sharkcz> all is looking good AFAIK, and look forward to use KVM for the s390x builder LPAR
17:05:08 <mboddu> And since its a non failable artifact for 28, its causing issues there
17:05:31 <mboddu> sharkcz: Have you created the ticket?
17:05:34 <nirik> yeah, looking forward to better things.
17:05:36 <sharkcz> wasn't pwhalen looking at the armv7 issue? I recall hearing something about it ...
17:05:46 <sharkcz> mboddu: nirik has one already
17:05:46 <nirik> ticket for what?
17:05:59 <mboddu> sharkcz: Ohh, I didn't know that
17:06:04 <sharkcz> for the z/VM to KVM
17:06:09 <nirik> yeah, arm has been failing for a long time.
17:06:17 <mboddu> nirik: For moving z/VM to KVM
17:06:28 <nirik> yeah, I have it, and have been talking to mainframe folks. It's on my list for after f29
17:06:39 <sharkcz> nirik: please add me to CC there
17:06:40 <mboddu> nirik++
17:06:58 <nirik> ok.
17:07:26 <mboddu> nirik: Please add me if you can, I want to follow the progress, thanks
17:08:13 <nirik> sure. will do
17:09:05 <mboddu> nirik: Thanks
17:11:02 <mboddu> #info nirik is working on using KVM for s390x builders on LPAR and expecting it be done post 29 GA
17:11:51 <mboddu> I have one ticket to discuss and 1 update/info/review
17:13:15 <mboddu> #topic #7856 Silverblue is in the wrong place on dl.fp.o
17:13:21 <mboddu> #link https://pagure.io/releng/issue/7856
17:14:09 <mboddu> nirik: Do you know/remember what decides whether something goes to /pub/fedora or /pub/fedora-secondary ?
17:14:32 <puiterwijk> mboddu: infra/releng?
17:14:51 <nirik> well, it used to be if it was primary or secondary, but thats a bit muddier these days...
17:15:00 <puiterwijk> Yeah
17:16:03 <nirik> There's tons of other things there that aren't blocking...
17:16:04 <mboddu> puiterwijk: Not who, but what, because of what nirik said :)
17:16:39 <puiterwijk> mboddu: well, because of what nirik said it's infra/releng... :)
17:16:42 <nirik> personally I'd love it if we just put everything in one place... but that would mean a lot more for mirrors
17:16:56 <mboddu> nirik: Yup, there are tons of things like spins, containers, AH which get synced to /pub/fedora
17:16:56 <puiterwijk> Yeah. The mirroring stuff is the problem there
17:17:02 <puiterwijk> Wait. Did the stuff the ostree repos in primary?
17:17:16 <nirik> hum?
17:17:19 <puiterwijk> looks like not. Just images
17:17:35 <puiterwijk> nirik: sorry, I just wanted to verify where they put the ostree repos. Since putting those in primary is a definite problem
17:17:43 <mboddu> puiterwijk: Just images, not repos
17:17:50 <nirik> ah, the repos themselves, yeah...
17:17:51 <puiterwijk> Yeah, just checked
17:18:15 * mboddu remembers there is a ticket to sync atomic repos to mirrors as well
17:18:43 <nirik> yeah, there was some discussion about how they wanted to do things...
17:18:58 <puiterwijk> Let's not get into that right now.
17:18:59 <nirik> but I think they want ostree trees now...
17:19:04 <mboddu> Anyway, coming back to this ticket...
17:19:10 <nirik> but yeah, thats a sidetrack for this ticket
17:19:11 <puiterwijk> +1
17:20:09 <nirik> if we want to go with 'only release blocking'' there we could move a bunch of stuff out.
17:20:24 <nirik> but moving anything is likely to annoy websites and adamw
17:20:35 <puiterwijk> Primarily adamw is what I'm afraid of
17:21:00 <mboddu> Yes, and I dont want to put anymore work on adamw
17:21:24 <mboddu> And 'only release blocking' is not a criteria
17:23:31 <puiterwijk> I'm inclined to give a shrug and just say "figuring out what goes where is hard"
17:23:42 <nirik> proposal: leave it as is for f29 and discuss more some plan for f30
17:23:54 <puiterwijk> sounds reasonable to me
17:25:16 <mboddu> nirik: +1
17:26:18 <mboddu> #info We will leave it as it is for F29 and will discuss and come up with a plan for F30.
17:27:10 <mboddu> #topic Open Floor
17:27:13 <mboddu> I have one thing
17:27:46 <mboddu> I sent out a patch for review for bodhi automated pushes and made adjustments as per nirik
17:29:06 <mboddu> If everyone is happy with it, I will push the change post 29 GA
17:29:36 <nirik> sounds good to me.
17:29:43 <nirik> what time did you have it going?
17:29:43 <puiterwijk> +1
17:31:27 <mboddu> nirik: 00:00 UTC as thats the time when freeze starts
17:31:48 <nirik> ok, fair. I think most any time will have tradeoffs.
17:34:19 <mboddu> nirik: Well, with this at least we dont have to worry about doing a push at exactly 00:00 UTC, its okay for us since it is evening for us, but for Patrick it will be late night
17:34:40 * nirik nods.
17:35:12 <nirik> well, it will mean most folks in the us have finished and submitted thier updates. But it also means if it fails it will happen in the middle of the night and we may not fix it until the next morning.
17:35:19 <nirik> but I am fine with that
17:36:37 <mboddu> nirik: Yeah, now that bodhi is very stable (/me knocks the wood) we might not to worry about failing pushes
17:36:52 <nirik> they are more rare for sure.
17:37:13 <mboddu> #info mboddu will push the patch for automated bodhi pushes post 29 GA
17:37:26 <mboddu> And, yay, no more daily pushes
17:37:37 <nirik> hurray!
17:37:45 <pwhalen> sorry I missed earlier, the armv7 container issue in f29 seems to be oom, install hangs right after grabbing squashfs/treeinfo
17:37:49 <mboddu> nirik dont have to push updates anymore on a weekend :D
17:38:50 <pwhalen> the image creation is successful when using an nfs mounted squashfs
17:40:03 <pwhalen> I wanted to see if it would be possible to use an iso, but havent had time
17:40:34 <nirik> huh, ok
17:42:09 <mboddu> Ohhh, okay
17:44:59 <mboddu> Anyway that we can fix this?
17:47:13 <pwhalen> that's the question. I tested the nfs mounted squashfs locally which worked, so thought using an iso could work as well
17:48:37 <pwhalen> here the output of a failed task - https://paste.fedoraproject.org/paste/8TEjGm2C9nVGWPWVIFOMOQ
17:48:58 <pwhalen> just hangs there, nothing further
17:51:52 <pwhalen> if anyone has suggestions, let me know. I have the same hardware locally so happy to test
17:53:16 <mboddu> pwhalen: Okay, thanks for offering the help, we will what we can do about it
17:53:32 <mboddu> we will see*
17:54:16 <mboddu> Has anyone got anything else?
17:55:57 <nirik> not me off hand.
17:56:40 <mboddu> Okay, I will give back 5 min of your time
17:56:44 <mboddu> Thanks everyone for joining
17:56:49 <mboddu> #endmeeting