14:31:40 #startmeeting Fedora CI SIG 14:31:40 Meeting started Wed Jun 3 14:31:40 2020 UTC. 14:31:40 This meeting is logged and archived in a public location. 14:31:40 The chair is jbair. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:31:40 Useful Commands: #action #agreed #halp #info #idea #link #topic. 14:31:40 The meeting name has been set to 'fedora_ci_sig' 14:31:47 .hello2 jimbair 14:31:48 jbair: Sorry, but you don't exist 14:31:58 .hello siddharthvipul1 14:31:59 siddharthvipul: siddharthvipul1 'Vipul Siddharth' 14:32:06 jbair: just .hello :) 14:32:12 .hello jimbair 14:32:13 jbair: jimbair 'Jim Bair' 14:32:15 there we go 14:32:17 ty :) 14:32:18 .hello2 14:32:19 bgoncalv: bgoncalv 'Bruno Goncalves' 14:32:25 .hello2 when your irc nick == fas id :D 14:32:34 hello everyone \o 14:32:36 I didn't read into it :) 14:32:37 hi! 14:32:39 morning 14:32:45 https://etherpad.gnome.org/p/fedora-ci 14:32:54 not sure if the agenda items are for this week or for the last meeting 14:33:00 but we can chat if we don't have any set agenda 14:33:15 I'll assume "next" means today :) 14:34:04 #topic datacenter migration 14:34:32 #link https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/thread/L3MAYSUM52C4G6PFVDZ6SEFKHCHGUUZS/ 14:34:43 The above email has a good overview of upcoming events 14:34:53 Looks like a few services will bounce to finalize the migration 14:35:30 .hello2 14:35:31 bookwar: bookwar 'Aleksandra Fedorova' 14:35:40 bookwar: hellow :) 14:35:54 there will be a lot of disruptions due to datacenter move 14:36:09 looks like they will occur this week and next week 14:36:21 so we should be expecting infra failures during that time 14:36:44 btw, maybe it is a good time to identify out infra dependencies 14:37:22 agreed 14:37:33 if there something unexpected which starts to fail because of some service outage, we need to record it somewhere that we have this dependency on the service 14:37:52 Is there a good place for it? The etherpad? 14:38:09 maybe just issue in Pagure? 14:38:59 So long as we have an action with the issues - something to do after finding the dependency 14:39:05 like updating code to account for outages or something 14:39:53 good point 14:40:09 the action would be to document them in our ci docs i think 14:40:46 outages start today, so keep your eyes open everyone :) 14:41:20 this may be a good time to switch to the centOS cluster topic since it's infra-related? 14:41:42 let's do this :) 14:42:01 +1 14:42:02 #topic ocp4 on centos cluster 14:42:09 siddharthvipul: how's the cluster going? 14:42:19 so the cluster is deployed with valid certs, automated storage and things 14:42:26 a little bit of petting needed 14:42:40 but it's good enough time to start discussing migration plan 14:43:02 we have added ACO authentication to the cluster (but it's still not quite there to be used by a general user) 14:43:20 ACO auth? 14:43:50 accounts.centos.org (apart from direct kubeadmin login) 14:44:30 ah 14:44:36 so like FAS but for centOS :) 14:44:51 yep, different ipsilon instance 14:44:54 siddharthvipul: can you create a test project inside that cluster and give access to us (at least jim) so that we can start playing with it? we don't need auth yet, but we would use persistent volumes and stuff 14:45:33 bookwar: yes we can 14:45:43 are we keeping those two forms of auth separate for awhile? Apologies if a silly question - somehow I haven't heard of it until now =) 14:45:53 and you can call the project in OCP fedora-ci - keeps it simple :) 14:46:08 bookwar: we are still working on nfs parts for persistent volumes 14:46:17 we have local storage pieces for now 14:46:38 siddharthvipul: do you have also ceph? or is it just nfs for now? 14:46:54 no, we are not using rook/ceph, just nfs 14:47:02 ok 14:47:52 jbair: bookwar we have created a ticket in "JIRA" and will update you all once we have a test namespace ready for y'all 14:48:11 thanks. I think getting the project early is helpful. We won't migrate anything yet, but we will check what other requirements we have for the migration and provide feedback on it 14:48:16 jbair: can you create a centos account for now and send me the id 14:48:31 bookwar: yes, definitely :) 14:48:32 siddharthvipul: I can :) 14:48:43 I'll do it after the meeting and ping you with the username 14:48:47 mostly debating jimbair vs jbair lol 14:48:56 jbair: take your time, no hurry at all :P 14:49:11 jbair: probably keep it same as FAS (will be easier in future haha) 14:49:42 well, with future it's noggin and probably won't have that problem, but different discussion 14:50:01 ack :) 14:50:09 anything else on the new cluster? 14:50:47 #topic Monitoring: reading fedmsg and counting it 14:50:51 nothing much, I just wanted to "raise awareness" that this is coming along and we should plan.. a good action item on us to give you namespace and play around is good for now 14:50:52 bookwar I think this one is yours 14:51:01 siddharthvipul: got it! 14:51:21 jbair: yes, that's from last time 14:51:31 i just wanted to share this project https://github.com/bookwar/fm-metrics 14:52:02 i have created a small app which counts messages on the message bus and classifies them by topic and exposes them in prometheus form 14:52:18 oh nice 14:52:20 i haven't done much with it yet 14:52:49 my goal is to add it to prometheus infra set up by mvadkert, so we can get metrics on ci messages in grafana 14:53:05 so just for your info, and contributions and feedback is welcome 14:53:21 nice! That should help for sure 14:53:29 I'll have to clone it and try it out 14:53:35 it is also an example of how one can use twisted app with fedora message bus listener 14:54:25 The last topic we have without an owner, assuming it's valid is about installability and rpmdeplint 14:54:38 not sure if that's msrb or someone else :) 14:54:44 i have added it 14:55:40 it is just a note that we are running "empty" installability and rpmdeplint pipelines to test the infrastructure, they currently not running any tests, just report "true" 14:56:02 #topic nstallability and rpmdeplint pipeline post green results, but don't do anything yet 14:56:09 we are still waiting for the backend from Testing Farm team 14:56:20 yeah, and the console output does say that which is helpful 14:56:30 I was looking into other failures and was glad to see those were expected when I first saw them a few weeks ago :) 14:56:53 do we have any update on when that backend is expected? 14:57:19 jbair: yes, we switched them to "pass" rather than fail, so that people get used to see them green, and when they start failing - this means there is smth real there 14:57:22 mvadkert: ping 14:57:22 bookwar: Ping with data, please: https://fedoraproject.org/wiki/No_naked_pings 14:57:35 mvadkert: is there an update on TFT backend? 14:58:30 i guess he is not here today 14:58:51 tflink: i'll reach out to him, and send an update to ci@ mailing list 14:59:08 bookwar: thanks 14:59:31 I had heard a few things but nothing had been said publicly or definitively 15:00:09 tflink: yep, we need a status update on this 15:00:42 #topic open discussion 15:00:57 I know we're at time, but in case there's anything else to discuss, now is your chance :) 15:03:45 #endmeeting