15:00:17 #startmeeting Infrastructure (2019-06-20) 15:00:17 Meeting started Thu Jun 20 15:00:17 2019 UTC. 15:00:17 This meeting is logged and archived in a public location. 15:00:17 The chair is smooge. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:17 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:00:17 The meeting name has been set to 'infrastructure_(2019-06-20)' 15:00:17 #meetingname infrastructure 15:00:17 The meeting name has been set to 'infrastructure' 15:00:17 #topic aloha 15:00:17 #chair nirik pingou puiterwijk relrod smooge tflink cverna mizdebsk mkonecny abompard bowlofeggs 15:00:18 Current chairs: abompard bowlofeggs cverna mizdebsk mkonecny nirik pingou puiterwijk relrod smooge tflink 15:00:31 Hello 15:00:34 o/ 15:00:51 morning. 15:01:14 I am expecting a small group as I forgot to email out the agenda 15:01:25 and no one added anything I could see 15:02:06 I was going to talk a short bit about workflow changes on the horizon... 15:02:18 cool 15:02:21 #topic announcements and information 15:02:21 #info bowlofeggs will be going on extended leave in 2019-06 15:02:21 #info cverna will be going on extended leave in 2019-06 15:02:21 #info Flock2Fedora 2019-08-08 -> 2019-08-11 15:02:21 #info Site trip to PHX2 will be 2019-07 15:02:23 but I didn't add it. :) 15:02:54 we really need to start planning that site visit if we want to get it in next month 15:03:03 #info Community Platforms Engineering team (CentOS+Fedora Infra) had a face to face meeting last week 15:03:18 yep. I am thinking it will be august 15:03:32 but... we might have to move datacenters, in which case a site visit in aug will be not much use 15:03:39 yep 15:04:10 so lets see what we hear in the next week or two 15:04:22 so after the meeting we should make a plan we will tear up in 2 weeks 15:04:38 #topic Oncall 15:04:38 #info https://fedoraproject.org/wiki/Infrastructure/Oncall 15:04:38 #info pingou is on call from 2019-06-20 -> 2019-06-27 15:04:38 #info ?????? is on call from 2019-06-27 -> 2019-07-04 15:04:38 #info ?????? is on call from 2019-07-04 -> 2019-07-11 15:04:39 #info Summary of last week: (from smooge ) 15:05:07 I am not sure giving pingou on-call when he isn't at the meeting is fair.. 15:05:32 .hello zlopez 15:05:33 mkonecny: zlopez 'Michal Konečný' 15:05:33 so I will wait til he pops back in 15:06:31 sure. I can tak 27th-5th 15:06:33 4th 15:06:39 or the one after 15:07:11 ok putting ou on 04 15:07:32 I will take after 15:07:56 .hello2 15:07:57 bowlofeggs: bowlofeggs 'Randy Barlow' 15:08:03 .hello3, even 15:08:11 Summary: The last couple of weeks has been meetings, more meeting and more meetings. We had a major outage this morning which was fixed within 10 minutes 15:08:46 this was due to a config push which broke some things 15:09:04 #topic Monitoring discussion 15:09:04 #info https://nagios.fedoraproject.org/nagios 15:09:04 #info Go over existing out items and fix 15:09:43 * nirik clicks 15:10:02 the fas messages is likely the openshift move. 15:10:14 been meaning to mention that one to puiterwijk. Guess I should file a ticket 15:10:22 yep. and I am expecting the greenwave is similar 15:10:24 no idea on greenwave, should bug them again 15:10:34 Yes, due to openshift move. 15:10:36 ppc8-02 and vh18 have disks that need replacing 15:10:56 they said the last 2 times that at the time no messages were being omitted for them to act on 15:11:14 I think they are wrong. :) 15:11:17 but perhaps 15:11:32 odcs-frontend01 is low on disk. We could ask them to clean up or give it more space 15:13:26 I think it needs more disk 15:13:56 I will make a ticket after this meeting 15:14:01 nirik, I think it is you now 15:14:02 or at least more cowbell 15:14:19 oh its gluster 15:14:49 ffs 15:15:07 me now? odcs? 15:15:10 * smooge remembers why he keeps forgetting about this system 15:15:21 nirik, sorry odcs was that 15:15:28 idea: work with odcs folks to move to nfs volume? 15:15:31 #topic Nirik Talks 15:15:48 Ha. ok. 15:16:34 so if you follow the community blog you know that we have been looking at how we work (when I say we here I mean the entire CPE team, which is all the Red Hat folks working on fedora and centos) 15:17:08 and we just had some face to face meetings last week on this, so just a heads up that changes are on the way... 15:17:38 I don't think these are going to be too radical or the like, just things to help us work better and avoid problems we currently have. 15:18:22 nirik: well the bit about how we will only accept requests written in Klingon is a bit radical i thought 15:18:48 bowlofeggs: It's Klingon now? 15:18:57 Some things that we are already trying to implement: ticket or it doesn't exist. Getting work items from irc or email just irq's us and causes silos where only some people get all the requests. 15:19:05 I thought we decided on Elvish? 15:19:26 I thought it was Dwarven 15:19:29 Q-plah! 15:19:36 it also means we can track things, prioritize them, etc. 15:19:55 yeah ticket-or-it-doesnt-exist++ 15:20:24 We are also trying to move to a model of 'one thing, finish it all the way, then take another' instead of trying to work on 12 tasks at once and never finish any of them 15:21:02 look for a lot more blog posts, documents and such in the coming weeks/months to explain things and hopefully get us on track. 15:21:05 That basically means that putting in a ticket and then pinging people in 3 channels to get it worked on is also not going to work 15:21:29 right. 15:21:38 The ticket has to get groomed (possibly by management), taken by a person with no current tasks, and then worked on to completion 15:21:49 So don't be offended if people ask you to update a ticket instead of asking on irc, or you only get a reply in ticket not irc. 15:22:36 I think bowlofeggs said it best... we can enforce the workflow we want by simply doing the work in the place we want... when people learn it's tickets, thats where they will go. 15:23:22 There's likey to be other changes down the road, but those are the near ones. 15:23:39 ☺ 15:23:42 so, to discuss here... 15:23:44 yeah i think it will be good 15:24:31 how can we track which exact thing people are working on? assignment on tickets? if so, we need to unsassign things because I think people have just taken tickets in the past that they thought they might work on someday 15:24:41 so it's a poor indicator of what they are actually working on 15:25:36 or we could do a tag... 15:25:49 'active' or 'being worked' or 'in progress' or something 15:26:13 in a remedy system I was in, we had 2 levels of assignment. We had a "will work on" queue and a "active" queue 15:26:47 yeah i recommend only assigning tickets you are actually working on 15:27:03 that way other people ahve more options when looking for something to do 15:27:17 yeah, if we had a kanban thing it might look a lot nicer. 15:27:21 the 2 level thing sounds cool though 15:27:39 yeah i guess a tag could help 15:28:02 we could do a custom field too perhaps. 15:28:07 imo any system where people can tell teh difference between "bowlofeggs is interested in working on this" and "bowlofeggs is currently working on this" will work 15:28:32 …on this 15:28:36 let me play around with a stg instance and see what I can come up with 15:29:49 anyhow, thats all I wanted to bring up today I think... 15:31:12 anything from anyone else on this?? 15:35:20 not me 15:35:24 #topic Open Floor 15:35:53 thanks nirik .. sorry my home internet went down for some reason 15:36:31 * smooge wonders if he is really on or just ghosting still 15:36:47 i see you smooge 15:36:52 ok 15:36:57 me too 15:37:29 thank you all for coming this week. I don't have anything else to discuss this week. I figure the changes in CPE are going to be going on for a while 15:38:35 yeah. 15:38:35 thanks smooge for running this :-) 15:38:49 and hopefully everyone will be patient with us while we figure out how to work better. 15:39:45 or be patent with us and we can patent how we work 15:40:15 bowlofeggs: https://pagure.io/pagure/pull-request/4507 15:40:22 you might be happy with that one :) 15:40:25 #endmeeting