17:00:46 #startmeeting RELENG (2017-10-12) 17:00:46 Meeting started Thu Oct 12 17:00:46 2017 UTC. The chair is mboddu. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:46 Useful Commands: #action #agreed #halp #info #idea #link #topic. 17:00:46 The meeting name has been set to 'releng_(2017-10-12)' 17:00:46 #meetingname releng 17:00:46 The meeting name has been set to 'releng' 17:00:46 #chair dgilmore nirik tyll sharkcz masta pbrobinson pingou puiterwijk maxamillion mboddu Kellin 17:00:47 #topic init process 17:00:47 Current chairs: Kellin dgilmore masta maxamillion mboddu nirik pbrobinson pingou puiterwijk sharkcz tyll 17:01:21 morning 17:01:24 howdy all 17:02:56 * mboddu waiting for more people to join 17:05:24 .hello maxamillion 17:05:25 maxamillion: maxamillion 'Adam Miller' 17:06:49 I guess most of the people are in F27 Server Beta Go/No-Go meeting 17:06:54 ah 17:06:55 yeah 17:07:12 maxamillion: since you are here, I can get an update on a ticket 17:07:16 lets start 17:07:22 mboddu: go for it 17:07:34 #topic #6939 Tracking ticket for bodhi->pungi mashing 17:07:43 #link https://pagure.io/releng/issue/6939 17:07:54 maxamillion: I know you just started working on it, so how is it going? 17:09:41 mboddu: it's going pretty well, I actually accidentally drug puiterwijk into it with me ... I was asking questions of him about stage infra because I just didn't know and then next thing I know he was doing some stuff while I was in a meeting and we ended up in a shared tmux session in stage bodhi doing some pair-programming type stuff, he's now working on a patch for something and we'll 17:09:47 probably have something done before too long 17:10:07 mboddu: from there it's mostly about mirrormanager and getting necessary changes into prod bodhi when the time comes 17:11:23 maxamillion: you drugged Patrick? Thats bad Adam 17:11:31 maxamillion: Anyway, thanks for the update 17:11:47 mboddu: it's going, stuff needs to be adjusted but it's all thought out 17:11:52 and just needs code 17:12:54 #info maxamillion and puiterwijk are working together but the process is all figured out and just needs implementation. 17:13:01 puiterwijk: how are you doing? 17:13:16 puiterwijk: I have a question for you as well, mostly a status update 17:13:19 mboddu: I'm awake 17:13:39 mboddu: and ask away then. 17:13:48 #topic #6984 create a unified ostree repo for fedora 27 and beyond 17:13:53 #link https://pagure.io/releng/issue/6984 17:14:10 puiterwijk: afair, you and Dusty were working on it during flock, any update? 17:15:05 mboddu: no, no update yet, I was primarily waiting on an SOP suggestion, but looks like Dusty asked me what info is needed. 17:15:48 mboddu: honestly, I think that we should work on this post-F27-GA, when things calm down 17:16:06 So, unless people bug me about it, it's on pause until then for me 17:17:04 puiterwijk: Okay, and it does make sense to wait until f27 GA 17:18:43 #info puiterwijk is waiting on a SOP suggestion and probably the work will be paused until F27 GA. 17:19:12 And thats it from me 17:19:31 nirik: I think you got something to share, so I am starting the Open Floor 17:19:37 #topic Open Floor 17:19:51 nirik and anyone else who got something to share? 17:20:19 I was going to bring up an idea someone else had... about batched updates. Right now we treat f27 just like any other release. we may want to do things differently since it's a prerelease 17:20:48 but perhaps we just want to wait and see how it goes... since we haven't had batched in place very long. 17:21:38 nirik: what are batched updates? 17:21:52 * mboddu googling 17:22:50 https://lists.fedoraproject.org/archives/list/devel-announce@lists.fedoraproject.org/thread/UDXVXLT7JXCY6N7NRACN4GBS3KA6D4M6/ 17:26:43 I'm fine just leaving it for now I guess... 17:27:30 nirik: So, whats the use of keeping an update in batched updates which can be pushed next week? 17:27:50 * mboddu trying to understand the use case here 17:28:07 it helps reduce thrash & churn on the mirrors 17:29:03 masta: Okay, it doesn't help the process, just the resources 17:29:41 it helps people not have to apply updates every. single. day. 17:29:49 I was thinking if it is planned to use something like freeze 17:29:57 especially when many of them are things they don't care about or are not important 17:30:32 it might help the daily process, less work for the mashers on the daily... 17:31:01 probibly in the noise... mash has to mash the entire thing either way 17:31:24 although there may well be branches with no updates to mash more often... 17:32:08 Let's turn it on for f27, I like the idea 17:32:15 I like the idea too 17:32:27 this is pretty common for other OS vendors to have "Patch Thursday" or something 17:32:30 it's already on 17:32:49 the question was if we should do it any differently for f27 than f25/f26/epel7/epel6 17:33:50 like 27 stable Wednesdays, 26 Tuesdays? Or something else? 17:35:10 everything in the batched state is moved to request on tuesday evening 17:35:24 request-stable* 17:35:43 So they'll be in the wednesday push 17:35:48 puiterwijk: so the question is whether we should enable it for f25/f26/epel7/epel6? 17:36:07 mboddu: it's enabled for everything 17:36:14 And there's no code to do it per release 17:37:01 puiterwijk: Okay 17:37:40 nirik: ^ based on that, what do you mean by "any differently for f27 than f25/f26/epel7/epel6"? 17:37:59 mboddu: the quest ion is whether thre *should* be code to handle it per-release 17:38:30 well, we don't even need code. 17:38:39 we could just approve all batches every night 17:38:55 That still needs code in the form of a script to do that :) 17:39:08 The push-batched-to-requested script doesn't accept args I don't think 17:40:00 which isn't hard, but I don't think it's there now 17:42:17 I think we should "enable it" 17:43:30 what do you mean? 17:43:43 masta: its already enabled with the new bodhi release 17:43:55 it is already enabled. I am not sure how more clear I can say that. ;) 17:43:56 ok I give up 17:44:23 we can just move on. I can bring up changing f27 / branched handling of it later... 17:44:29 after we have all seen it work for a while 17:45:05 nirik: yeah, thats a good idea. But I dont see why branched should be handled differently 17:45:32 But, we might find something once we all seen it working 17:47:32 #info bodhi now supports batched updates and all the batched updates are requested stable on Tue at 3:00 UTC and all the batched updates are pushed stable on Wednesday's bodhi push. 17:47:39 Anything else? 17:49:16 Okay, next topic 17:49:28 #topic Alternative Architectures updates 17:49:53 nirik, puiterwijk : same old question, s390x update? :) 17:50:01 sharkcz: any updates from you on alt arches? 17:50:02 mostly fix. 17:50:04 fixed. 17:50:09 except it dropped a min ago. 17:50:11 woot woot 17:50:17 :( 17:51:34 nirik: so, I saw sometime early this week, s390x composes started working, but I remember someone saying it was a fluke 17:51:45 So, when did it actually got fixed? 17:51:59 nope, it should be ok now. The main problem was inactivity on the mount causing it to drop 17:52:11 It' now doing a keepalive 17:53:29 nirik: Oh okay 17:54:09 nirik: How sure are you that s390x mounting issues are fixed? so that I can add it to the meeting minutes here 17:55:19 well, nothing in life is 100% but it seems like it's fixed to me. 17:55:22 * mboddu dont want to jinx it by saying that it got fixed :D 17:55:47 nirik: hehe 17:55:47 I'll add a nagios check for it too. 17:55:56 nirik: okay 17:56:47 #info s390x mounting is resolved and nirik is planning on adding a nagios check. 17:57:00 Okay, I think thats it 17:57:24 Thank you everyone for joining 17:57:30 #endmeeting