15:02:54 #startmeeting Infrastructure (2019-05-15) 15:02:54 Meeting started Thu May 16 15:02:54 2019 UTC. 15:02:54 This meeting is logged and archived in a public location. 15:02:54 The chair is smooge. 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 'infrastructure_(2019-05-15)' 15:02:54 #meetingname infrastructure 15:02:54 The meeting name has been set to 'infrastructure' 15:02:54 #topic aloha 15:02:54 #chair nirik pingou puiterwijk relrod smooge tflink cverna mizdebsk mkonecny abompard bowlofeggs 15:02:54 Current chairs: abompard bowlofeggs cverna mizdebsk mkonecny nirik pingou puiterwijk relrod smooge tflink 15:03:00 hi 15:03:03 morning party people. 15:03:08 I have only 1 job... 15:03:09 Hello! 15:03:17 .hello zlopez 15:03:18 mkonecny: zlopez 'Michal Konečný' 15:03:24 morning 15:03:27 .hello2 15:03:28 bcotton: bcotton 'Ben Cotton' 15:03:41 morning 15:04:43 * pingou 15:04:49 #topic New folks introductions 15:04:49 #info This is a place where people who are interested in Fedora Infrastructure can introduce themselves 15:04:49 #info Getting Started Guide: https://fedoraproject.org/wiki/Infrastructure/GettingStarted 15:05:29 Hey everyone, I'm new. Just finally sent my introduction to the mailing list yesterday but have been looking over projects and guides for a while. 15:06:30 Welcome here :-) 15:07:01 hello braydon_kains 15:07:27 #topic announcements and information 15:07:28 #info Zombieland reboots of important servers have been done. 15:07:28 #info bodhi-4.0.0 beta is in staging. Please test because we ship soon 15:07:28 #info only N days until F31 freeze 15:07:28 #info bowlofeggs will be going on extended leave in June 15:07:29 #info nirik will be on leave from 2019-06-03 to 2019-06-07 15:07:30 #info many people will be at a community platform engineering f2f 2019-06-10 to 14 15:07:32 #info Another blogpost about release-monitoring.org is published (https://communityblog.fedoraproject.org/stories-from-the-amazing-world-of-release-monitoring-org-5/) - mkonecny 15:07:35 #info FAS2 has been moved to openshift (puiterwijk) 15:07:37 #info Ipsilon has been moved to openshift (puiterwijk) 15:07:38 welcome braydon_kains! 15:07:39 #info community openshift is progressing 15:07:43 #info next reboot cycle? 15:07:50 braydon_kains, I can add you to the apprentice gorup after meeting if you aren't already 15:08:09 smooge: Yeah I'm not in that yet, would appreciate it thanks. 15:09:02 OK we have a LOT of things starting to go on in the next 3 weeks 15:09:04 .hello2 15:09:05 rpbarlow: Sorry, but you don't exist 15:09:14 #info new getfedora.org has been deployed. Thank you to puiterwijk for deployment help 15:09:24 .hello bowlofeggs 15:09:25 rpbarlow: bowlofeggs 'Randy Barlow' 15:10:13 hello bowlofeggs/rpbarlow 15:10:18 heh 15:10:19 smooge: What is the N in "only N days until F31 freeze" 15:10:20 thanks relrod 15:10:33 i guess freenode bumped me from being my usual self but that's ok because i have more than one personality 15:11:01 relrod: I worked on translation and reported few things I encountered 15:11:26 https://fedoraproject.org/wiki/Releases/31/Schedule 15:11:36 2019-08-29... so we have a while. 15:12:03 mkonecny, looks like ~100 days til freeze and ~60 days til mass rebuild 15:12:25 oh, another info: 15:12:29 mkonecny: thank you 15:12:41 smooge: thanks 15:12:48 #info flocktofedora is aug 2nd-5th 15:12:49 relrod: You are welcome 15:13:02 no meeting that week either :) 15:13:08 sorry, 8th-11th 15:13:13 Hooray Flock \o/ 15:13:16 https://flocktofedora.org/ 15:13:32 everyone should start thinking about talks they might like to give... 15:13:56 I already have two in mind 15:14:09 nirik, ha! you had me there for a moment with 'Oh I can go!' 15:14:12 I think we should definitely do a infrastructure workshop, those are always good 15:14:27 mkonecny: oh yeah? 15:14:34 * pingou has at least one 15:15:01 * nirik isn't sure. There's a bunch I could do, but not sure what sounds good this year... 15:15:15 community openshift? 15:15:20 nirik: One is about release-monitoring.org with a little magic added :-) 15:15:35 mkonecny: are you going to wear a pointy hat? 15:15:55 The second one should be about privacy, but I'm not sure if this is a good speech for Flock 15:16:29 pingou: Unfortunately I don't have any, but I can borrow one from my brother :-D 15:16:36 pingou: yeah, could... with a lot of help from puiterwijk, we now have it up and running (initial install/not for public use yet) 15:17:49 mkonecny: sounds cool. Magic is great! 15:18:10 * nirik almost gave a privacy talk a few years back, but decided not to. 15:18:29 pingou, community openshift is the replacement for community openstack that is most of Fedora Infra Cloud 15:18:52 or was that a different question? 15:19:25 smooge: it was a talk suggestion for flock :) 15:19:30 sorry 15:19:33 smooge: I think he was suggesting I give a talk on it... 15:19:54 smooge: no worries :) 15:19:55 I came back from a ping and thought it was about the topic line 15:20:08 #topic Oncall 15:20:09 #info https://fedoraproject.org/wiki/Infrastructure/Oncall 15:20:09 #info nirik is on call from 2019-05-09 -> 2019-05-16 15:20:09 #info relrod is on call from 2019-05-16 -> 2019-05-23 15:20:09 #info smooge is on call from 2019-05-23 -> 2019-05-30 15:20:10 #info ?????? is on call from 2019-05-30 -> 2019-06-06 15:20:11 #info no one is on call from 2019-06-06 -> 2019-06-20 15:20:15 #info ?????? is on call from 2019-06-20 -> 2019-06-27 15:20:17 #info Summary of last week: (from nirik ) 15:20:38 a fair pile of pings... on various topics 15:20:48 fas had a few people as it was moved and needed tweaking 15:21:03 i bothered you a lot this round too! 15:21:04 there was some bodhi tweaking to work around issues with other packages 15:21:10 yeah. ;) 15:21:32 a fair number of incoming tickets to triage... 15:21:44 I think we are still over 100... 15:22:01 yeah, 105 15:22:36 you going to take the coffee cup smooge ? 15:22:48 relrod said he would 15:23:00 * nirik reads up more, ah... ok. ;) 15:23:02 I take it next week so he can do something during that week 15:23:23 I'm going to be out early june, so I can't take the week after that 15:23:26 * smooge needs to see if relrod has stopped partying from graduation 15:23:30 smooge: actually... so I mixed up my weeks. Is there any way we could flip? I won't be in tomorrow and this weekend 15:23:48 zodbot: take_coffee_cup_of_power 15:24:06 smooge: not much partying but a small trip to Dayton OH this weekend 15:24:17 did you go to the air force museum? 15:25:06 no, I need to 15:25:38 Two years ago when I went, I stayed in a hotel that basically sits on the afb though. 15:25:55 zodbot: oncall_take_us 15:25:55 smooge: Kneel before zod! 15:26:08 * smooge goes to make an alias for coffe_cup 15:26:19 ok schedules are changed 15:26:26 thank you and sorry 15:27:13 nirik, puiterwijk how does 2019-06-20 -> 2019-06-27 look for either of you 15:27:34 I should be able to do that week. 15:28:01 I don;'t think I have any plans to go anywhere after the face to face meetings and until flock 15:28:18 but... we may want to figure out phx2 site visit sometime 15:28:21 it's been a while since I took one, I should put my name in the hat 15:29:08 nirik, yeah.. I was figuring mid July for the phx2 site visit 15:29:20 pingou, I will put you on that one 15:29:24 we can change later 15:29:40 #topic Monitoring discussion 15:29:40 #info https://nagios.fedoraproject.org/nagios 15:29:40 #info Go over existing out items and fix 15:30:06 not sure whats up with those buildvmhost-aarch64's... 15:30:11 rhit firewall? 15:30:12 smooge: I expect I'll have some days off on that week, I'll just be slower at ticket triaging 15:31:21 3 boxes show raid issues, likely need rebooting to re-add drives? 15:31:22 the raids should be fixed on reboots 15:31:31 there's also one more machine with a dropped drive 15:31:54 autocloud-backend-libvirt2 15:32:27 I tried to do them via the drac but they will need someone on console to clear hw controller 15:32:38 oh I didn't know that one.. oof 15:32:53 nirik, is that a service which is going away after 28 is EOL? 15:33:04 no, after f29 15:34:39 so that brings us to reboots... perhaps next week? 15:34:57 when the fallout from the latest cpu attacks is more known? 15:35:02 yeah 15:35:12 in case we have to move everything to AMD and ARM 15:36:04 Do the usual cadence? Monday Staging, Tuesday External, Wednesday Internal, Weekend Build? 15:36:28 most of that cadence was because it was easier for just one person to do parts as they had time 15:36:44 yeah, that works for me, unless others have better ideas 15:36:49 and also to try and not make it a OMG fire to meet an outage window. 15:36:58 * pingou has to step out shortly 15:37:04 good seeing you pilotmatt 15:37:10 good seeing you pingou 15:37:12 :) 15:37:17 auto-complete.. you complete me 15:37:23 pilotmatt++ tho 15:37:39 ok I will put in tickets after meeting 15:37:52 #topic Tickets discussion 15:37:52 #info https://pagure.io/fedora-infrastructure/report/Meetings%20ticket 15:38:25 dammit. session crashed. 15:38:36 so I wanted to bring up something here which nirik has been bringing into work force 15:39:00 if it takes longer than a few minutes or needs more than 1 person, it needs to have a ticket 15:39:10 no ticket, no work 15:39:42 yeah, it might be a hassle, but it really helps us keep better track of things and how much people are doing. 15:40:25 +1 15:40:46 also related: 15:40:54 it also helps keep people working efficiently by not being interrupted 15:41:04 if you have a ticket, please keep that updated with questions/progress... 15:41:32 and please do NOT update the ticket and immediately ping people about it. Tickets send email, people get it and can act when they are able. 15:42:26 ideally we make tickets efficent enough people don't feel they need to do that. 15:43:08 and conversely.. opening a ticket and then pinging people about it lowers it in the queue 15:43:23 the cabal queue 15:44:34 haha yes that's a good policy 15:44:58 ok anything else on this? 15:45:10 #topic Open Floor 15:46:09 Well thank you all for attending. And thank you all for your work 15:46:20 small announcement - i have bodhi-backend02's tasks in openshift in stg (though they currently crash - investigating) 15:46:22 thanks smooge 15:46:28 so with bodhi 4, we should be able to drop backend02 15:46:28 cool 15:46:37 thanks smooge 15:46:58 more resources for my fedora block chain package signing tool 15:47:25 smooge-- 15:47:31 haha 15:47:45 Looking forward to being part of the team! Thanks everyone! 15:47:56 is it impossible or just not worth it to move backend01 into openshift? 15:48:00 so I figured that if we have all these people wasting electricity to 'mine' things why not let them do builds and package signing for us 15:48:16 i know there's something weird abotu the NFS share, but i confess that i don't really understand it 15:48:31 wait wait wait.. I am cracking jokes and rpbarlow is asking serious questions 15:48:35 hahaha 15:48:43 well, we could in theroy... but it does pungi runs... 15:48:47 yeah 15:48:51 so right now it's a vm with a ton of memory/cpu 15:49:00 yeah more than 90 GBs! 15:49:01 I am not sure how pungi would handle being in a container 15:49:10 yeah i don't know either 15:49:12 there was talk about moving that into koji... 15:49:30 that's true i guess we could consider that 15:49:31 we could make them runroot tasks possibly. 15:50:22 is a runroot task basically a "please run this shell command for me koji?" 15:50:24 I'm not sure how difficult that would be, but it sounds like it should work 15:50:30 yep. 15:50:34 ah 15:50:34 with these mounts, etc 15:50:44 well that is basically all i'm doing with bodhi's composer anyway 15:50:51 instead of asking koji it asks the OS 15:50:58 but you know, maybe it wouldn't be that hard? 15:50:59 random task: https://koji.fedoraproject.org/koji/taskinfo?taskID=34778643 15:51:07 see the 'command' there 15:51:17 bodhi does directly write files itself too though 15:51:21 to add the updateinfo.xml 15:51:40 is there a "koji please accept this updateinfo.xml that i made for you" command? haha 15:52:00 nope. ;( But if there was a seperate tool to do that it could run it as part of the runroot task? 15:52:03 i.e., it's not just pungi though that's the bulk of the work 15:52:06 or it could run a second one to do that 15:52:26 it would be hard for it to be a separate tool because it uses bodhi's DB 15:52:34 well, unless that tool also used bodhi's DB i guess 15:53:13 ok lets move this part to #fedora-admin. There will be another meeting here in 5 minutes 15:53:15 maybe a topic for further discussion another day i suppose - that's all i have on it ☺ 15:53:21 ok cool 15:53:38 thank you all again. see you next week. same bat-channel same bat time 15:53:42 #endmeeting