15:01:39 #startmeeting Infrastructure (2019-11-07) 15:01:39 Meeting started Thu Nov 7 15:01:39 2019 UTC. 15:01:39 This meeting is logged and archived in a public location. 15:01:39 The chair is mkonecny. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:01:39 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:01:39 The meeting name has been set to 'infrastructure_(2019-11-07)' 15:01:39 #meetingname infrastructure 15:01:39 The meeting name has been set to 'infrastructure' 15:01:39 #topic aloha 15:01:39 #chair nirik pingou relrod smooge tflink cverna mizdebsk mkonecny abompard bowlofeggs 15:01:39 Current chairs: abompard bowlofeggs cverna mizdebsk mkonecny nirik pingou relrod smooge tflink 15:01:48 .hello zlopez 15:01:50 mkonecny: zlopez 'Michal Konečný' 15:01:55 hello 15:02:11 morning 15:02:14 morning 15:02:22 morning 15:02:44 #topic Next chair 15:02:44 #info magic eight ball says: 15:02:44 2019-11-07: mkonecny 15:02:44 2019-11-14: smooge 15:02:44 2019-11-21: pingou 15:02:55 morning 15:03:29 Do we want to nominate someone for 28th November? 15:04:23 If not, we could decide this on next meeting 15:04:41 Yes next meeting sounds good to me 15:04:52 #topic New folks introductions 15:04:52 #info This is a place where people who are interested in Fedora Infrastructure can introduce themselves 15:04:52 #info Getting Started Guide: https://fedoraproject.org/wiki/Infrastructure/GettingStarted 15:05:00 Anybody new here? 15:06:07 28th nov is a holiday in the us... so we might skip that one. 15:06:24 nirik: Oh, I forgot 15:07:16 It looks like nobody new today, so moving to next topic 15:07:18 #topic announcements and information 15:07:18 #info f31 final freeze now over 15:07:18 #info aarch64 hardware mostly sorted, will be installing soon 15:07:18 #info koji backups are done on secondary host. Slowdown in koji still occurring. 15:07:18 #info mass update/reboot cycle completed 15:07:47 Any other announcement? 15:08:30 * nirik has nothing off hand. 15:09:37 Moving to Oncall then 15:09:39 #topic Oncall 15:09:39 #info https://fedoraproject.org/wiki/Infrastructure/Oncall 15:09:39 #info nirik is on call 2019-10-31->2019-11-07 15:09:39 #info cverna is on call 2019-11-07->2019-11-14 15:09:39 #info pingou is on call 2019-11-14->2019-11-21 15:09:40 #info Summary of last week: (from current oncall ) 15:10:21 wasn't too bad... really busy dealing with host updates/reboots, but not too many pings 15:10:35 most were related to the reboot no? 15:11:02 yeah, not too many otherwise. 15:11:13 .takeoncalleu 15:11:33 What is the magic command ? 15:11:40 although I think we are backsliding sometimes (me anyhow) and answering people on irc too quickly/etc 15:11:58 .oncalltakeus 15:11:58 mkonecny: Error: You don't have the alias.add capability. If you think that you should have this capability, be sure that you are identified before trying again. The 'whoami' command can tell you if you're identified. 15:12:07 nirik: well, this week you were allowed ;-) 15:12:19 I know the words, but I don't have right magic wand :-D 15:12:31 try while wearing your hat! ;-) 15:12:38 pingou: sure... 15:13:14 I also don't want us time but Europe. Even if I don't sleep much 😛 15:13:40 .oncalltakeeu 15:13:40 cverna: Kneel before zod! 15:13:41 there is 'oncalltakeeu' 15:13:45 :) 15:13:55 I don't think we have separate oncall for us and eu :-D 15:13:55 Cool 🙂 15:14:05 .oncall 15:14:07 cverna is oncall. My normal hours are 0700 UTC to 1600 UTC Monday through Friday. If I do not answer or it is outside those hours, please file a ticket (https://pagure.io/fedora-infrastructure/issues) 15:14:28 Ok I am all setup 15:14:30 Oh, I was wrong then 15:14:43 #topic Monitoring discussion 15:14:43 #info https://nagios.fedoraproject.org/nagios 15:14:43 #info Go over existing out items and fix 15:15:17 so, lets see. 15:15:27 we have the osbs-aarch64 boxes down (known) 15:15:38 we have virthost-os02 down (it lost 2 drives) 15:16:10 bugzilla messages are still not coming in. ;( I will poke that internal ticket again I guess... 15:16:54 * pingou forgot something for the announcements, will wait for the open floor 15:17:03 the rabbitmq mts thing I am not sure on... 15:17:15 perhaps the app needs a redeploy? 15:17:21 mts? 15:17:23 we could ping on the mts ticket 15:17:35 module tagging service iirc 15:18:12 thanks pingou 15:18:44 * nirik does a deploy 15:19:15 I think thats most / all of them? 15:19:49 What is the issue with bugzilla, those are fedmsg messages? 15:20:13 yeah, we listen to a STOMP endpoint and inject fedmsgs... 15:20:18 but it stopped sending to us. 15:21:15 it might be a expired cert... 15:22:32 The issue looks like it's on the bugzilla side, thanks for explanation 15:22:51 #topic Tickets discussion 15:22:52 #info https://pagure.io/fedora-infrastructure/report/Meetings%20ticket 15:22:52 #info no tickets marked for the meeting 15:22:52 #info Fedora CoreOS Team requests 15:22:52 #link https://hackmd.io/5uB7hOJKSjGUt65iLgPnbA#Existing-requests-for-Fedora-Infra 15:23:26 Do we have some tickets to be discussed? 15:23:56 Is somebody from CoreOS Team here? 15:24:38 * nirik checks the link 15:24:40 dustymabe, dustymabe, dustymabe 15:25:12 yeah, all those are things we would like to get done. ;) 15:25:13 hmmm doesn't look like summoning worked 15:25:17 * mkonecny beetlejuice, beetlejuice, beetlejuice 15:26:19 Let's move to backlog 15:26:20 #topic backlog discussion 15:26:20 #info go over our backlog and discuss and determine priority 15:26:20 #link https://pagure.io/fedora-infrastructure/issue/8178 15:26:20 #info topic: provision new aarch64 builders 15:26:20 #link https://pagure.io/fedora-infrastructure/issue/8157 15:26:21 #info topic: ansible: enable ansible-report as a hook 15:27:25 the aarch64 builders are finally all coming along... 15:27:47 No progress on the ansible report 15:27:52 on ansible-report... I think we were going to run it and send a initial to the list? 15:28:03 so we could ideally clean up things? 15:28:24 But since I am on call I'll take some time to do that 15:29:16 What about doing this as part of the end year initiative? 15:29:25 how long does one run take ? 15:29:54 I think it's pretty quick? 15:30:06 I did not manage to do a full run it add some errors 15:30:33 It is quite picky on what you give as an input 15:30:34 we had all errors fixed at the face2face meeting... 15:30:40 but it had a ton of warnings 15:31:23 Yes I think this what I saw lots of warnings 15:31:32 you can supress those with -q 15:31:32 Do we want to ignore them? 15:32:24 I think we want to fix them, but perhaps only block on error. 15:32:41 * nirik can't remember how to run this thing... it expects that you cat everything to it? 15:33:28 You need to use find and xargs 15:33:33 perhaps there are better tools now. :) 15:33:48 Find all the yml files and xargs them to the tool 15:34:13 ah right 15:34:23 Hm, not that straightforward like flake-8 in Python 15:34:35 could someone put that info into the ticket so that anyone can reproduce the warnings/errors without having to ask here first ? 15:35:11 karsten++ 15:35:12 mkonecny: Karma for karsten changed to 1 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 15:35:12 yeah, we should. 15:35:27 Yes, although the info on how to run is on the GitHub repo 15:35:38 But I am happy to update the ticket 15:36:29 * dustymabe waves 15:36:34 cverna++ 15:36:35 mkonecny: Karma for cverna changed to 5 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 15:36:38 cverna++ 15:36:39 nirik: Karma for cverna changed to 6 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 15:37:05 Moving to last topic 15:37:06 #topic Open Floor 15:37:08 sorry I was late 15:37:13 i've got nothing for today 15:37:14 Cookies :) 15:37:22 #info I've updated ci-resultsdb-listener today 15:37:22 dustymabe: welcome 15:37:30 dustymabe: no worries than 15:37:39 it now uses fedora-messaging and supports the new message format used by ci.centos 15:38:06 this means, we will be able to decommission resultsdb.ci.centos.org soon 15:38:12 as well as archive the rdbsync project 15:38:23 Nice 15:38:32 (rdbsync is already turned off on resultsdb01.qa) 15:38:52 excellent. 15:38:58 * pingou is actually just removing it 15:39:10 eol 15:39:56 CentOS CI should start using fedora-messaging soon as well 15:40:06 it was announced for next week 15:40:24 they are planning on updating the messaging plugin today and change its configuration next week 15:40:36 I was not sure about the dates 15:41:11 so, I have a question on that timing, if we are on to that... 15:41:33 I'm ok 15:42:34 so next week, I heard... tuesday is prod rawhide gating work? that includes redeploying bodhi-backend01 15:42:49 redeploying? 15:43:04 reinstalling? 15:43:22 or no? 15:43:28 not that I am aware of 15:43:30 abompard: ^ 15:43:33 I thought it is only an update 15:43:51 for me as well 15:43:58 Yes redeploying it 15:44:16 well, either way is fine, although it's f29 and we need to move it off that before eol 15:44:23 Because we need a more recent version of celery or something related 15:44:43 cverna: I forgot about the celery in the new Bodhi 15:44:47 ok, so it's going to f30? because f31 has not all the python2 needed for sigul 15:44:53 bodhi-backend01.stg is F30 15:45:02 * nirik nods 15:45:07 Yes f30 15:45:08 prod is still f29 indeed 15:45:14 30 is fine, just want to know. ;) 15:45:26 and finally to my question... 15:45:58 yes on Tuesday for bodhi 5.0 15:46:02 there's a new koji (1.19.0) I wanted to get updated/out, but... there's been issues with it, so they are doing a 1.19.1 next week hopefully. when would be a good time to do that that doesn't interfere 15:46:16 and/or do you need 1.19.x koji as part of this? 15:46:32 the koji that is stg seems to be working fine 15:46:41 1.18 I guess? 15:46:53 I upgraded it to 1.19.0 monday. :) 15:46:55 so we roll out the new bodhi on Tues 15:46:58 ah :) 15:47:09 well, still seems to perform fine for bodhi's needs 15:47:41 well, prod right now is 1.18.1 (well, it's a mix, but thats what is on the hub) 15:47:43 I'd recommend that we don't do koji 1.19.1 before Thu 15:48:05 We could try it in staging first 15:48:17 to give us a couple of days (Tue/Wed) to figure out if anything in prod is misbehaving 15:48:36 but I'd also recomment we do koji 1.19.1 before announcing rawhide gating 15:48:41 (which I have no ETA for atm) 15:48:54 but I doubt it'll be before Nov 18th 15:49:02 recommend* 15:49:16 ok... I'm fine waiting and doing koji upgrade week after next 15:49:24 Sounds like a reasonable date 15:49:34 nirik: end of next week should be fine I think 15:49:39 do we need a outage scheduled for tuesday? for bodhi at least/ 15:49:50 depending on how you like the idea to upgrade koji on a Thu or a Fri 15:50:10 yeah, not great, but doable 15:50:14 I think we should announce a outage, when we will do the redeploy of backend 15:50:32 to be on the safe side we can announce it 15:50:37 worst case, it's minimal 15:51:01 cool... can someone take that as a action. ;) I'm happy to help 15:51:02 or rather, worst case there is one, best case barely 15:51:08 on it 15:51:22 pingou++ 15:51:22 mkonecny: Karma for pingou changed to 2 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 15:51:34 hm I thought we had a template for outage on our issue tracker 15:51:55 pingou: we do, but currently the repo attached to fedora-infrastructure is the repospanner one thats not working. ;) 15:52:04 ahah, good point :) 15:52:07 ie, we would need to move the old one back to get the template. 15:52:21 yup 15:52:31 * pingou takes an old ticket as example 15:52:37 anyhow, I had one more quick item: 15:53:13 We are going to try and do some short projects before the end of the year... hopefully forming a few small teams and just getting them done. 15:53:31 we should have a list of them next weekish and will share for everyone to see and provide input on 15:53:46 EOF 15:53:58 nirik++ 15:54:40 We should look at our backlog when planning these short projects 15:55:22 There should be something reasonably small 15:55:55 mkonecny: oh yeah, there's tons of things. ;) 15:56:01 we have no lack of work 15:56:16 I know, the time is our master 15:56:34 Do we have something else for the last few minutes? 15:58:03 In this case, I hereby declare this meeting as finished 15:58:04 #endmeeting