15:02:54 #startmeeting RELENG (2020-06-30) 15:02:54 Meeting started Tue Jun 30 15:02:54 2020 UTC. 15:02:54 This meeting is logged and archived in a public location. 15:02:54 The chair is mboddu. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:02:54 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:02:54 The meeting name has been set to 'releng_(2020-06-30)' 15:02:55 #meetingname releng 15:02:55 The meeting name has been set to 'releng' 15:02:55 #chair nirik sharkcz pbrobinson pingou mboddu dustymabe ksinny jednorozec 15:02:55 #topic init process 15:02:55 Current chairs: dustymabe jednorozec ksinny mboddu nirik pbrobinson pingou sharkcz 15:03:19 hello, I'm here today 15:04:16 Hey sharkcz 15:04:20 How are you doing? 15:04:33 I'm fine 15:06:48 I think nirik is not here today 15:07:07 Good for him that he is taking some rest 15:07:19 #topic #9562 cloud/container image creation fails on s390x 15:07:25 #link https://pagure.io/releng/issue/9562 15:07:36 So, I haven't dug anything into it yet 15:07:45 any idea about this one? 15:07:45 sharkcz: Anything you observed? 15:08:11 I belive it first appeared after updating koji to 1.21 on the builders 15:11:19 sharkcz: What do you mean by https://pagure.io/releng/issue/9562#comment-662551 ? 15:12:09 I have meant like updating only koji, but not oz (in case koji 1.21 needs newer oz than koji 1.20) 15:12:41 sharkcz: Okay, let me check that quickly 15:13:51 because missing "image.os_plugin.abort()" should fail on all arches, there is (hopefully) nothing s390 specific there 15:16:18 I compared both s390x and x86_64 builders oz and imagefactory and both of them are running the same versions and the latest one's as well 15:16:44 ok 15:17:09 Maybe a koji issue? 15:18:21 it looks as code issue, I will probably need to look closer 15:19:22 maybe something changed there between 1.20.1 and 1.21 ... 15:21:09 https://koji.fedoraproject.org/koji/taskinfo?taskID=45808274 is when they started failing it seems 15:21:56 And it seems when we moved to iad2 dc and also when we upgraded to 1.21 15:22:11 yup 15:22:45 so in theory there could be some firewall rule missing too 15:24:48 Yup 15:25:54 * nirik waves, reads up 15:27:24 nirik: Go back to bed 15:30:53 naw, I'm fine 15:32:30 So, we think it might be some firewall issue or koji 1.21 or imagefactory or oz updates or iad2 .... 15:34:41 * nirik has no idea without looking. 15:37:22 ok. I know what it is. 15:37:35 imagefactory packaging issue. 15:38:53 f31 used python 3.7 and f32 uses python 3.8... and imagefactory tries to make these links in /etc/imagefactory/plugins.d/ that point to the files, but it makes them in post and sometimes does the wrong thing on upgrades. 15:40:09 But why s390x only? 15:40:22 Doesn't it should happen on all arches? 15:41:18 good question. :) All the other ones were reinstalled fresh with f32. 15:41:27 only s390x were upgraded 15:42:39 Ahhh okay, so, uninstalling imagefactory and reinstalling it again on s390 should fix this? 15:42:49 *might* fix this? 15:43:02 or editing the symlinks by hand 15:43:22 yeah, I am testing a reinstall... not sure if it does the right thing or not. 15:43:24 sharkcz: Its easy to write an adhoc command to do this 15:43:34 ok 15:43:39 this = uninstalling and installing of imagefactory 15:44:50 it's just the plugins... but yeah, removing it and installing again fixes it. 15:45:40 Okay, I will remove them and install them on s390x builders 15:46:15 mboddu: great, thanks. 15:46:24 #info Kevin found that it might be due to a packaging bug, so mboddu will uninstall and install imagefactory on all s390 builders 15:46:47 #topic Open Floor 15:47:12 nirik: Whats the status on installing the hardware that got shipped from phx2 last week? 15:47:39 we are working thru it... I hope to add more builders today/tomorrow. 15:48:58 nirik++ 15:49:00 Thanks nirik 15:49:43 oh I have one 15:50:00 mboddu: container base images are failing 15:50:23 siddharthvipul: On s390x or all arches? 15:50:30 If s390x we were just talking about it 15:50:31 s390x 15:50:39 I see 15:50:55 I will skip that in script.. thanks 15:51:25 siddharthvipul: nirik debugged the issue and I am fixing it as we speak 15:51:36 oh you all are awesome 15:51:43 I will bug you later then :) 15:53:36 Anything else? 15:53:45 nothing here :) 15:54:02 mboddu++ nirik++ for s390x issue.. and well, so manyother work 15:54:14 of course I am out of karma :p 15:54:39 Well, its all nirik, I am just running a simple adhoc command to fix it :) 15:55:24 Anyway, if nothing else, I will give back 5 of your min back 15:55:34 Thanks guys for joining 15:55:36 #endmeeting