15:00:48 #startmeeting Infrastructure (2020-01-16) 15:00:48 #meetingname infrastructure 15:00:48 Meeting started Thu Jan 16 15:00:48 2020 UTC. 15:00:48 This meeting is logged and archived in a public location. 15:00:48 The chair is smooge. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:48 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:00:48 The meeting name has been set to 'infrastructure_(2020-01-16)' 15:00:48 The meeting name has been set to 'infrastructure' 15:00:48 #chair nirik pingou relrod smooge tflink cverna mizdebsk mkonecny abompard bowlofeggs 15:00:48 #info Agenda is at: https://board.net/p/fedora-infra 15:00:48 Current chairs: abompard bowlofeggs cverna mizdebsk mkonecny nirik pingou relrod smooge tflink 15:00:49 #topic aloha 15:01:00 .hello nphilipp 15:01:01 nils: nphilipp 'Nils Philippsen' 15:01:04 morning all 15:01:32 \o 15:02:02 .hello2 15:02:03 ks3: ks3 'Kevin Sandy' 15:02:28 mornin 15:02:35 morning* 15:02:50 morning 15:03:38 .hello2 15:03:38 karsten: karsten 'Karsten Hopp' 15:04:22 #topic Next chair 15:04:22 #info magic eight ball says: 15:04:22 #info 2020-01-16 - smooge 15:04:22 #info 2020-01-23 - ??? 15:04:23 #info 2020-01-30 - no meeting? 15:04:52 .hello2 15:04:53 bowlofeggs: bowlofeggs 'Randy Barlow' 15:04:56 who can take next weeks meeting? Or should we cancel due to everyone on travel and such 15:05:09 i can do it if it is going to happen 15:05:18 * nirik will be out traveling 15:05:28 I ll be around 15:05:41 i think nirik, relrod, smooge will be out next week. 15:05:45 but yeah I think we will have many people travelling 15:05:54 I will put bowlofeggs down for the meeting and it can be a record short one 15:06:05 +1 15:06:22 thanks bowlofeggs 15:06:27 #topic New folks introductions 15:06:27 #info This is a place where people who are interested in Fedora Infrastructure can introduce themselves 15:06:27 #info Getting Started Guide: https://fedoraproject.org/wiki/Infrastructure/GettingStarted 15:06:48 .hello ekulik 15:06:49 ernestask: ekulik 'None' 15:06:55 welp, no name 15:07:11 in any case, hi, I’m one of the ABRT devs 15:07:11 .hello2 15:07:13 abompard: abompard 'Aurelien Bompard' 15:07:21 I’m looking towards becoming an apprentice of yours 15:07:26 and working together more closely 15:07:32 as I need to take care of our part of infra sometimes 15:07:51 welcome ernestask! 15:07:59 welcome ernestask 15:08:10 * nirik notes the "None" is there if you set the 'privacy' setting in fas so it doesn't share your real name. 15:08:52 ernestask, I will add you after the meeting 15:09:07 #topic announcements and information 15:09:07 #info builder reinstalls/uprgades to f31 continuing. Should be done before holiday break - kevin 15:09:07 #info ops folks are trying a 30min ticket triage every day at 19UTC in #fedora-admin 15:09:07 #info smooge is on PTO 2020-01-13 -> 2020-01-23 15:09:08 #info the following people will be AFK for Devconf: cverna, pingou, mkonecny, mboddu, nirik, relrod, nils 15:09:09 #info the following people will be AFK for FOSDEM: pingou, mboddu 15:10:00 builders are all done now. ;) 15:10:20 * nirik forgot to update the board 15:11:04 smooge: wait.. you are already on pto? 15:11:19 nope 15:11:40 2020-01-13 was monday? 15:11:54 #info smooge is on PTO 2020-01-18 -> 2020-01-26 15:12:01 sorry about that 15:12:21 8, 6 and 3 look too much alike :P 15:12:40 :) 15:12:55 ok, just confused me. ;) 15:13:31 #info all of infrastructure was updated/rebooted in the last 72 hours 15:13:40 #info pkgs01.stg was rebuilt 15:15:06 oh wow, we rebooted all the world's roads and bridges too? 15:15:24 yeah, the bridges were tricky. 15:15:30 #info Fedora Infrastructure will be moving datacenters this year. This is your only notice beyond the one we have filed in Alpha Centauri filing cabinet 15:16:08 since we are being silly 15:16:30 .hello zlopez 15:16:30 mkonecny: zlopez 'Michal Konečný' 15:16:52 I don't have any other announcements 15:16:55 #topic Oncall 15:16:55 #info https://fedoraproject.org/wiki/Infrastructure/Oncall 15:16:55 #info smooge is oncall 2020-01-09 -> 2020-01-16 15:16:55 #info nirik is oncall 2020-01-16 -> 2020-01-23 15:16:56 #info ??? is oncall 2020-01-23 -> 2020-01-30 15:17:11 cverna, or bowlofeggs can you take that last week? 15:17:26 I can 15:18:01 .oncalltakeus 15:18:01 nirik: Kneel before zod! 15:18:25 oh I will be at devconf but only the weekend so that should be ok 15:18:42 also my travel time is much shorter 15:19:15 ok I have updated you in the form. 15:19:20 thanks 15:19:23 #info ???? is oncall 2020-01-30 -> 2020-02-06 15:19:33 if anyone wants to plan out ahead 15:19:50 #info Summary of last week: (from current oncall ) 15:20:06 #info dedicated-solutions (proxy11) has been having network problems a lot 15:20:16 #info smooge broke composes for EPEL 15:20:45 what is dedicated-solutions ? what it is doing ? 15:20:57 it is a colocation we get a server from 15:21:13 they are in germany? I think? 15:21:15 i think the data center is in europe 15:21:19 ha ok :) 15:21:32 and it has been up and down since Sunday though today it has been down the longest 15:22:20 #info Most .oncall were from 06:00 UTC -> 07:00 UTC when smooge is not available 15:22:25 and that is all I have 15:22:48 #topic Monitoring discussion [nirik] 15:22:48 #info https://nagios.fedoraproject.org/nagios 15:22:48 #info Go over existing out items and fix 15:23:14 lets see. 15:23:27 the fas messages missing... we still need to fix that 15:23:44 no fedora planet messages - we need to fix that 15:24:02 dbgserver we need to remove from monitoring... not sure why it's still showing up 15:24:20 qa-stg01 disk space... we need to ping tflink to look I think? 15:24:21 ok will do so 15:24:39 * tflink will look 15:24:49 for fas should we just wait for the new fas ? 15:24:50 and the check fedmsg gateway ones that relrod was going to fix. 15:24:55 IIRC, that host has needed to be deleted for a while now 15:25:18 dbgserver still shows up in monitoring because inventory/inventory:[dbgserver_stg] defines it 15:25:23 cverna: well, we could... but it's anoying as people change ssh keys and then our trigger to update them never fires. 15:25:39 smooge: can you fix that/remove it? 15:25:43 yep 15:25:44 nirik: ok I was not sure how it was used 15:26:08 I have added the koji wellness monitoring plugin from https://pagure.io/fedora-infrastructure/issue/6505 15:26:24 oh cool. I had that on my backlog a long time... thanks much! 15:26:46 it is checking if koji.fp.o works well, but nagios will report on koji01.fp.o 15:27:00 ok 15:27:38 anyhow I think thats it for nagios... would be nice to get it all clean someday again 15:28:09 #topic Tickets discussion [nirik] 15:28:09 #info https://pagure.io/fedora-infrastructure/report/Meetings%20ticket 15:29:27 ok we are doing daily meetings on tickets so I think this section might be covered? 15:29:46 well, we could still talk about any folks want. ;) 15:29:46 #topic backlog discussion 15:29:46 #info go over our backlog and discuss and determine priority 15:29:46 #link https://pagure.io/fedora-infrastructure/issue/8178 15:29:46 #info topic: provision new aarch64 builders 15:29:52 oops sorry 15:29:55 thats fine. 15:30:26 aarch64 boxes will get cables today. I will put in a ticket for networking after that 15:30:44 aarch64 boxes will then be able to get get built/run after that 15:31:06 oh wait its closed 15:31:07 cool. 15:31:25 I would like to try and tackle https://pagure.io/fedora-infrastructure/issue/6397 15:31:27 so those are... how many, I lost track? 15:31:48 oops sorry I thought you were done 15:31:58 nirik, no idea myself.. 15:32:05 cverna: thats going to be a big one. ;) needing lot of planning, but sure! 15:32:08 sorry my brain has swapped that out 15:32:20 I think it's 7 of them, but I can check my notes. 15:33:02 nirik: I can start looking at the planning bits, like what needs to be done 15:33:18 nirik: then we can probably do that in an incremental fashion :) 15:33:22 cverna: it would be great to get it done sometime for sure. 15:33:55 ideally if we redirect things we set a sunset on that and remove the redirects in a year or two or something (I hate having tons of redirects) 15:34:15 yes that sounds good to me 15:34:42 I ll try to reference what needs to move first 15:34:49 cool 15:35:41 cool 15:35:43 any others? 15:36:23 https://pagure.io/fedora-infrastructure/issue/8544 15:36:35 I just want to announce that this one is fixed 15:36:54 cool 15:36:57 I can close it 15:37:05 Quick fix was applied and longterm fix will be deployed next week 15:37:25 mkonecny: cool. Thanks for looking at it. 15:37:31 There is a topic on resultsdb ownership in the doc, but it also says to revisit after devconf 15:37:48 so otherwise.. I think we are at open floor 15:38:02 oh and thanks nirik for taking oncall 15:39:27 #topic open floor 15:39:33 * tflink has one thing 15:39:36 for open floor 15:39:42 its your floor 15:40:15 long story short, the keypairs for rabbitmq generated by infra can't work with jenkins. mostly because Java 15:40:32 hahahahaha 15:40:40 * tflink was wondering if there were any preferences on where to store the things we need to convert those keypairs into 15:41:00 is it for fedora messaging ? 15:41:02 to be clear, it's not difficult to convert the keypairs into something that Jenkins can use 15:41:20 cverna: yeah, to auth with the rabbitmq instances 15:41:33 there is a jenkins plugin for fedora-messaging 15:41:36 ah, bummer. 15:41:43 mkonecny: do you have the link for it ? 15:41:44 cverna: jms-messaging? 15:41:46 tflink: I updated jms-messaging-plugin for rabbitmq 15:41:52 ha yes 15:42:08 yeah. it can't use the keypairs generated by infra without modification 15:42:11 https://github.com/jenkinsci/jms-messaging-plugin RabbitMQ support is added in latest verson 15:42:26 cool, does that include the ability to read from rabbitmq queues? 15:42:38 tflink: We used this plugin for CentOS CI 15:42:48 yeah, I'm using it as well 15:42:51 tflink: It can read and publish 15:43:04 great, the current fork I'm using can't do that 15:43:34 If there will be any issue with the plugin, just let me know 15:43:38 but we digress. should the modified keypairs be stored in the ansible private repo? 15:43:59 i think so 15:44:37 ok, I can add them and their passwords to the repo so they're not just sitting on my laptop 15:44:39 we could sure. 15:44:50 if the unmodified ones are in private then the modified ones should be 15:45:13 * tflink also needs to write up the process of getting that all to work in a public place 15:45:24 I do need to add one more topic 15:46:14 #topic Hardware budget season 15:46:54 We are at the time of the year where we put in budget requests to Red Hat for replacing servers or additional capacity. 15:47:38 bowlofeggs needs a talos II workstation to build a CI system for the infra 15:47:44 If you know of project capacity you need additional capacity between 2020-03->2020-12 please let me know so I can put in requests 15:48:31 ok will put that in 15:48:47 speaking of capacity, we might want to add some nodes to our prod/stg openshift. 15:48:59 I have a hard deadline of 2020-02-01 15:50:06 so please give me some ideas of what and how many by then. [i just got given the deadline so my apologies for its shortness.] 15:50:10 EOF 15:50:33 nirik, yeah I will do so 15:50:40 #topic Back to open floor 15:50:47 ok anything else? 15:51:03 well, we will need to work out how to do that. ;) I'm not sure off hand, I guess it takes a redeploy? 15:51:20 * nirik has nothing else off hand 15:51:35 nirik, well I figure we will be redeploying to the new datacenter so setting up a OS4 there makes sense 15:52:13 ok eol 15:52:16 #endmeeting