16:59:50 <mboddu> #startmeeting RELENG (2017-10-19)
16:59:50 <zodbot> Meeting started Thu Oct 26 16:59:50 2017 UTC.  The chair is mboddu. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:59:50 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
16:59:50 <zodbot> The meeting name has been set to 'releng_(2017-10-19)'
16:59:50 <mboddu> #meetingname releng
16:59:50 <zodbot> The meeting name has been set to 'releng'
16:59:50 <mboddu> #chair dgilmore nirik tyll sharkcz masta pbrobinson pingou puiterwijk maxamillion mboddu Kellin
16:59:51 <zodbot> Current chairs: Kellin dgilmore masta maxamillion mboddu nirik pbrobinson pingou puiterwijk sharkcz tyll
16:59:51 <mboddu> #topic init process
16:59:54 <puiterwijk> hi
16:59:56 <dustymabe> .hello dustymabe
16:59:57 <zodbot> dustymabe: dustymabe 'Dusty Mabe' <dustymabe@redhat.com>
17:00:03 <masta> ahoy
17:00:08 <mboddu> dustymabe: you are quick :)
17:00:26 <dustymabe> mboddu: :)
17:00:29 <dustymabe> masta: o/
17:01:13 * dustymabe hoping ausil will join us today
17:01:49 <mboddu> dustymabe: I think he cant join today
17:02:00 <dustymabe> ??????
17:02:35 <maxamillion> .hello maxamillion
17:02:36 <zodbot> maxamillion: maxamillion 'Adam Miller' <maxamillion@gmail.com>
17:02:42 <dustymabe> dgilmore - can you join us today?
17:03:10 <puiterwijk> maxamillion: just a quick hint that I learned myself recently:
17:03:13 <puiterwijk> .hello2
17:03:13 <zodbot> puiterwijk: puiterwijk 'Patrick "マルタインアンドレアス" Uiterwijk' <puiterwijk@redhat.com>
17:03:14 <bowlofeggs> i'm gonna /nick maxabillion
17:03:31 <maxamillion> puiterwijk: oh, also .hello by itself apparently works now
17:03:34 <puiterwijk> bowlofeggs: "head, meet desk"
17:03:35 <maxamillion> puiterwijk: I just always forget
17:03:35 <mboddu> bowlofeggs: haha
17:03:36 <maxamillion> .hello
17:03:38 <zodbot> maxamillion: (hello <an alias, 1 argument>) -- Alias for "hellomynameis $1".
17:03:40 <maxamillion> oh,
17:03:42 <maxamillion> nvm
17:03:44 <puiterwijk> maxamillion: :)
17:03:45 <maxamillion> maybe it was hello2
17:03:48 <maxamillion> ¯\_(ツ)_/¯
17:03:49 <puiterwijk> Yeah
17:03:53 <bowlofeggs> hahaha
17:03:54 <puiterwijk> hello2 is .hello $mynick
17:03:57 <maxamillion> +1
17:04:05 <maxamillion> bowlofeggs: do it
17:04:06 <dustymabe> what's the nice
17:04:21 <maxamillion> bowlofeggs: MAX-A-ALL-THE-THINGS
17:04:24 <puiterwijk> zodbot: misc help hello2
17:04:24 <mboddu> maxamillion: Yeah, I came to know about it last week, diff between .hello2 and .hello
17:04:26 <zodbot> puiterwijk: (hello2 <an alias, 0 arguments>) -- Alias for "hello $nick".
17:04:28 <puiterwijk> maxamillion: ^
17:04:48 <bowlofeggs> .hello maxabillion
17:04:49 <zodbot> bowlofeggs: Sorry, but you don't exist
17:04:56 <puiterwijk> :-)
17:05:32 <mboddu> dustymabe: so, I think you need Dennis for https://pagure.io/releng/issue/7113? right?
17:05:39 <bowlofeggs> maxathousand**3
17:05:55 <dustymabe> mboddu: we can still have a discussion here
17:06:11 <mboddu> dustymabe: Okay
17:06:16 <mboddu> lets start then
17:06:38 <mboddu> #topic #7113 proposal to run bodhi updates pungi composes with `--no-label`
17:06:44 <mboddu> #link https://pagure.io/releng/issue/7113
17:06:54 <puiterwijk> I don't think that that's going to be useful, since Dennis is the person who told us to specify the label. So without the info as to why he wanted that, there's not much discussion to be had I don't think
17:06:56 * dustymabe will give people time to read the description
17:07:25 <dustymabe> if no one else can explain what a label is used for then I think that is useful information
17:07:27 <dustymabe> puiterwijk: ^^
17:07:40 <mboddu> puiterwijk: yeah, thats what I thought
17:07:46 <dustymabe> so i think it's worth chatting about
17:07:58 <puiterwijk> dustymabe: I can tell you that it's used in the PDC import for information purposes on its end
17:08:21 <dustymabe> 'information purposes' is pretty generic
17:08:26 <maxamillion> ah
17:08:32 <puiterwijk> What I cannot tell you is how important that is, or what the info is used for exactly other than things like fedfind
17:08:43 <maxamillion> PDC data is slated to drive a lot of Factory 2.0, so if it's needed there then that's going to remain important
17:09:39 <mbonnet> I don't think we have any dependency on update labels though
17:10:05 <lsedlar> PDC could import composes without label fine, it already has nighlies without a label
17:10:19 <puiterwijk> Yes, but it is extra info that PDC makes available
17:10:33 <dustymabe> puiterwijk: so a string that exists in PDC?
17:10:35 <puiterwijk> And as I said, I do not know how useful that info is, because I do not know everything that uses PDC
17:10:45 <maxamillion> ultimately I think the questions about it being necessary would be for dgilmore and threebean
17:11:31 <dustymabe> If it's useful them I'm ok with leaving it. it will require some changes to pungi for us.
17:11:37 <dustymabe> i'm just trying to make sure it is worth the effort
17:11:55 <bowlofeggs> yeah i think we should defer to dgilmore on this
17:11:59 <puiterwijk> And that is a valid question, but I'm not sure we're going to get that answered here.
17:12:12 <puiterwijk> So I'd suggest to ask Dennis to comment on the ticket.
17:12:20 <dustymabe> puiterwijk: will do
17:12:25 <mboddu> dustymabe: I understand, but you should ask Dennis to comment on it
17:12:28 <dustymabe> thanks for the discussion
17:13:38 <mboddu> #info dustymabe will check with dgilmore on this topic
17:14:04 <mboddu> dustymabe: I guess the other one is also waiting on Dennis comments
17:14:23 <mboddu> dustymabe: I will move to other topics then, is that okay?
17:15:05 <dustymabe> mboddu: well
17:15:09 <dustymabe> it's been waiting for a week
17:15:11 <dustymabe> no input there
17:15:19 <dustymabe> I'd like to not block on dennis forever
17:15:42 <dustymabe> can we agree to timebox it?
17:16:05 <dustymabe> i.e. if it's not commented on by dennis by next tuesday, we approve it
17:16:14 <dustymabe> or something like that
17:17:03 <mboddu> dustymabe: I dont want to do it, but I can check with him about it next week and I can get something by next Thu, is that okay?
17:17:57 <dustymabe> sure. i just don't want to wait for a month and then have someone say no to something that everyone else is ok with
17:18:10 * dustymabe just wants input :)
17:18:16 <dustymabe> ok move on to other topics
17:18:56 <mboddu> dustymabe: sorry about it, but I will get you something by next Thu
17:18:57 <mboddu> Okay, moving on
17:19:28 <mboddu> #topic #6939 Tracking ticket for bodhi->pungi mashing
17:19:39 <mboddu> #link https://pagure.io/releng/issue/6939
17:20:30 <bowlofeggs> i def don't want to approve the --no-label flag without dgilmore's input, even if he doesn't respond by next tuesday
17:20:30 <maxamillion> dustymabe: define "everyone else"
17:20:34 <maxamillion> sorry, reading backlog
17:20:36 <bowlofeggs> since he did explicitly tell us to do that
17:21:20 <dustymabe> bowlofeggs: we are refeerring to https://pagure.io/releng/issue/7100
17:21:22 <dustymabe> maxamillion: ^^
17:21:31 <dustymabe> which as discussed last week and has been waiting on input
17:22:36 <puiterwijk> So, on the current topic (bodhi->pungi mashing), what is there to discuss there?
17:22:40 <puiterwijk> mboddu: ^
17:22:59 <maxamillion> dustymabe: yeah, I'm looking at that one and I don't see any vote count or anything ... I'll check the meeting logs
17:22:59 <mboddu> puiterwijk: Oh my irc got disconnected, so I wanted to check if we can close it now?
17:23:09 <puiterwijk> mboddu: I'd say yes
17:23:16 <mboddu> puiterwijk: okay
17:23:35 <puiterwijk> Bodhi 3.0 is live, and we've sent out updated repos for everything
17:23:39 <dustymabe> maxamillion: i linked to the meeting logs in my last comment
17:23:40 <mboddu> #info puiterwijk and bowlofeggs completed the work on bodhi calling pungi to mash.
17:23:54 <dustymabe> puiterwijk++ bowlofeggs++
17:23:56 <zodbot> dustymabe: Karma for bowlofeggs changed to 22 (for the f26 release cycle):  https://badges.fedoraproject.org/tags/cookie/any
17:23:57 <maxamillion> dustymabe: +1
17:24:36 <mboddu> Okay, moving on
17:24:55 <mboddu> #topic Open Floor
17:25:01 <mboddu> Anybody has anything to share?
17:25:11 <puiterwijk> Bodhi 3.0 is live?
17:25:43 <maxamillion> yeah, I'd say that is note worthy :)
17:25:55 <mboddu> #info Bodhi 3.0 is live and for this week puiterwijk is going to do the pushes to debug any issues that might occur
17:26:15 <puiterwijk> #info so far, no significant issues that aren't a result of me being silly have occured
17:26:22 <puiterwijk> Sorry.
17:26:25 <puiterwijk> #undo
17:26:25 <zodbot> Removing item from minutes: INFO by puiterwijk at 17:26:15 : so far, no significant issues that aren't a result of me being silly have occured
17:26:30 <mboddu> puiterwijk: haha
17:26:40 <puiterwijk> #info So far, no significant issues have occured, and no issues that aren't a consequence of me being silly either
17:26:47 <bowlofeggs> #info puiterwijk is a hero
17:27:16 <mboddu> puiterwijk: But, starting next week its going to be normal schedule for pushes, right?
17:27:28 <puiterwijk> mboddu: that is the plan, yes
17:27:41 <mboddu> bowlofeggs: ^^ I want to leave that in out meeting minutes
17:27:57 <bowlofeggs> haha ☺
17:28:29 <mboddu> bowlofeggs: But I guess, I shouldn't do that, sorry Patrick ;)
17:28:31 <mboddu> #undo
17:28:32 <zodbot> Removing item from minutes: INFO by bowlofeggs at 17:26:47 : puiterwijk is a hero
17:29:02 <mboddu> #info Starting next week, we will be back on normal schedule for pushing updates
17:29:25 <puiterwijk> #info Patrick thinks Mohan should be a little less strict about meeting logs
17:29:49 <puiterwijk> #undo
17:29:49 <zodbot> Removing item from minutes: INFO by puiterwijk at 17:29:25 : Patrick thinks Mohan should be a little less strict about meeting logs
17:29:57 <puiterwijk> mboddu: see, I can help with cleaning up
17:30:25 <puiterwijk> :)
17:30:54 <mboddu> puiterwijk: hehe, the only problem is we put on rel-eng mailing list, which is public, hence the problem
17:31:21 <mboddu> puiterwijk: ^ we put our meeting minutes
17:31:21 <bowlofeggs> freenode is public too ☺
17:31:44 <puiterwijk> mboddu: meh. You should see the infra meeting logs. We keep some silly stuff in at the end at times, sometimes it makes people read them just to wonder what's the funny thing at the end this time :)
17:32:03 <mboddu> bowlofeggs: yeah, I guess I should be little bit liberal then
17:32:06 <mboddu> puiterwijk: Oh okay
17:32:17 <puiterwijk> mboddu: but it's fine, whatever you prefer, since you're chairing :)
17:32:21 <mboddu> #info puiterwijk is a hero
17:32:25 <mboddu> puiterwijk: there you go
17:32:42 * puiterwijk didn't mean that that one specifically had to stay. But I did win something :)
17:33:01 <bowlofeggs> hahah yesss
17:33:05 <mboddu> hahahah
17:33:17 <maxamillion> puiterwijk++
17:33:34 <mboddu> Anybody got anything else?
17:35:03 <dustymabe> mboddu: one thing real quick
17:35:23 <dustymabe> we don't currently build qcows/ISOs in the bodhi/pungi updates composes
17:35:35 <dustymabe> i don't know if we ever do want to do that (because of time), but you never know
17:35:57 <dustymabe> I'd like to create a compose that gets kicked off immediately after the updates compose completes
17:36:05 <dustymabe> i.e. listens to fedmsg and then kicks off a new compose
17:36:06 <mboddu> dustymabe: I never thought thats a plan
17:36:42 <dustymabe> does anyone have any tips for achieving that?
17:37:05 <puiterwijk> Well, what would be in that kicked-off compose?
17:37:06 <mboddu> puiterwijk: ^ ???
17:37:24 <dustymabe> puiterwijk: basically exactly what is in the twoweek stuff that runs at night today
17:37:41 <dustymabe> instead of kicked off via cron, we kick it off via fedmsg
17:37:44 <puiterwijk> Okay, then I think I would suggest to get that in the updates compose but failable.
17:37:59 <puiterwijk> I would say we create everything that's needed in the same compose
17:38:06 <dustymabe> yeah, the problem with that is time puiterwijk
17:38:14 <dustymabe> i'd love to do that too, but I don't think we'll be able to
17:38:28 <puiterwijk> dustymabe: well, for that I would first like to see how long the current composes take and how long those composes take.
17:38:52 <dustymabe> ok so basically you'd like to see the difference?
17:38:59 <puiterwijk> No, the combined times.
17:39:06 <mboddu> puiterwijk: I am not sure how useful it is, since we dont enable bodhi of branched for some time after branching, but we do nightlies right away
17:39:09 <dustymabe> i know we build atomic images, cloud images, and docker images "nightly"
17:39:19 <mboddu> puiterwijk: and no bodhi for rawhide
17:39:19 <dustymabe> in separate composes
17:39:24 <puiterwijk> So, I want to figure out 1. how long do the current updates mashes take, and 2. how long would it add to also combine nightly
17:39:47 <puiterwijk> mboddu: this is post-release.
17:39:51 <mboddu> puiterwijk: So, you can do it for couple of months in a release cycle
17:40:09 <mboddu> puiterwijk: Oh, just for atomic and cloud composes?
17:40:13 <puiterwijk> mboddu: yes
17:40:15 <dustymabe> puiterwijk: ok yeah that would take more "testing" in prod to fully determine that info puiterwijk
17:40:26 <dustymabe> because really stg does not tell us how long it will take in prod
17:40:28 <mboddu> puiterwijk: sorry, my bad, I was thinking of normal nightly composes
17:40:31 <dustymabe> as it is different hardware
17:40:44 <puiterwijk> dustymabe: so, I would first like to stabilize the current pushes and get that fully as they should (Right now there are still steps that I kick off manually to test)
17:40:53 <dustymabe> puiterwijk: indeed
17:40:59 <puiterwijk> dustymabe: so, my current proposal is to wait about a week, and then check the time the composes take.
17:41:02 <dustymabe> i'm not expecting this today
17:41:06 <puiterwijk> And then use that as input for this discussion
17:41:07 <dustymabe> i just wanted to start a discussion
17:41:26 <dustymabe> puiterwijk: of course
17:41:28 <dustymabe> thanks
17:41:33 <puiterwijk> Yeah. I just think that right now we do not have the data for that discussion. So let's wait for things to stabilize, and discuss this next week?
17:41:45 <puiterwijk> If possible, I'd love to just combine the two composes.
17:41:59 <dustymabe> puiterwijk: yeah, my feeling was that my experience with the composes tells me its going to take too long
17:42:09 <dustymabe> but real data would be better
17:42:15 <dustymabe> thanks for the discussion
17:42:18 <puiterwijk> dustymabe: that's a fair concern, and we need to take it into account indeed. But yeah, let's capture data :)
17:42:27 <puiterwijk> Right now is certainly too early for any real-use data
17:44:19 <lsedlar> this is real data so far: https://paste.fedoraproject.org/paste/tfluQ~bpHs2uKzYvVjw4pA with sample size of 2…
17:45:09 <puiterwijk> lsedlar: hah, thanks for that. But I think that right now the composes have been too fluctuating to be of any use
17:45:25 <lsedlar> puiterwijk, yeah, totally
17:46:30 <lsedlar> was there no F27 updates compose? I only see updates-testing
17:46:40 <puiterwijk> F27 is the branched nightly compose
17:46:51 <puiterwijk> So f27 updates "stable" in bodhi is just tagging stuff over right now
17:46:57 <lsedlar> doh, of course
17:47:03 <lsedlar> thanks
17:47:13 <dustymabe> ok we can probably close meeting out
17:47:14 <puiterwijk> No worries, I was wondering that myself too the other day :)
17:48:45 <mboddu> #info puiterwijk will look at merging nightly atomic and cloud composes along with updates push sometime after next week based on statistics gathered from new bodhi 3.0 push
17:48:56 <mboddu> Sorry guys, I have bad connection right now
17:49:00 <puiterwijk> mboddu: no, I disagree with that summary
17:49:05 <puiterwijk> #undo
17:49:05 <zodbot> Removing item from minutes: INFO by mboddu at 17:32:21 : puiterwijk is a hero
17:49:42 <puiterwijk> #info we should have a discussion next week about merging nightly atomic and cloud composes with updates push when we have statistics about how long updates composes take
17:49:57 <puiterwijk> mboddu: I am not saying I will look at it or fix it. I am just saying we should discuss it only then.
17:50:14 <mboddu> puiterwijk: oh okay
17:50:15 <puiterwijk> (I'm also not saying I won't, I'm just saying that I do not want to commit to doing it just yet)
17:51:04 <mboddu> #topic Alternate Architectures Update
17:51:11 <mboddu> sharkcz: Any update?
17:51:19 <dustymabe> mboddu: i have one
17:51:45 <dustymabe> now that we have bodhi+pungi officially talking to one another, we'll be able to ship ostree updates for multi-arch
17:51:48 <mboddu> dustymabe: sure
17:52:01 <sharkcz> mboddu: all looks good to me :-)
17:52:06 <dustymabe> #info now that we have bodhi+pungi officially talking to one another, we'll be able to ship ostree updates for multi-arch
17:52:19 <dustymabe> don't know if I'm chair so my #info might not count
17:52:30 <mboddu> dustymabe: sure?
17:52:35 <puiterwijk> #chair dustymabe
17:52:35 <zodbot> Current chairs: Kellin dgilmore dustymabe masta maxamillion mboddu nirik pbrobinson pingou puiterwijk sharkcz tyll
17:52:39 <dustymabe> #info now that we have bodhi+pungi officially talking to one another, we'll be able to ship ostree updates for multi-arch
17:52:58 <dustymabe> that might make pbrobinson happy
17:53:15 <mboddu> #info now that we have bodhi+pungi officially talking to one another, we'll be able to ship ostree updates for multi-arch
17:53:29 * puiterwijk wonders if that's twice or thrice in the logs now
17:53:33 <mboddu> #undo
17:53:33 <zodbot> Removing item from minutes: INFO by mboddu at 17:53:15 : now that we have bodhi+pungi officially talking to one another, we'll be able to ship ostree updates for multi-arch
17:53:45 * mboddu think it should be good now
17:54:02 <puiterwijk> We'll find out :)
17:54:05 <mboddu> #info Everything looking fine on Alternate arches
17:54:15 <mboddu> Okay, I am gonna close this meeting now
17:54:22 <mboddu> Anything else?
17:55:27 <mboddu> puiterwijk: haha
17:55:48 <mboddu> Okay, thats it then
17:55:58 <mboddu> Thanks for joining guys
17:56:01 <mboddu> #endmeeting