19:01:14 #startmeeting Fedora Infrastructure Ops Daily Standup Meeting 19:01:14 Meeting started Wed Feb 22 19:01:14 2023 UTC. 19:01:14 This meeting is logged and archived in a public location. 19:01:14 The chair is phsmoura. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions. 19:01:14 Useful Commands: #action #agreed #halp #info #idea #link #topic. 19:01:14 The meeting name has been set to 'fedora_infrastructure_ops_daily_standup_meeting' 19:01:14 #chair nirik smooge 19:01:14 Current chairs: nirik phsmoura smooge 19:01:14 #meetingname fedora_infrastructure_ops_daily_standup_meeting 19:01:14 #info meeting is 30 minutes MAX. At the end of 30, its stops 19:01:14 The meeting name has been set to 'fedora_infrastructure_ops_daily_standup_meeting' 19:01:14 #info agenda is at https://board.net/p/fedora-infra-daily 19:01:47 hello 19:02:45 #info reminder: speak up if you want to work on a ticket! 19:02:45 #topic Tickets needing review 19:02:53 #info https://pagure.io/fedora-infrastructure/issues?status=Open&priority=1 19:02:56 morning everyone 19:03:09 .ticket 11142 19:03:10 phsmoura: Issue #11142: Inactive provenpackagers - fedora-infrastructure - Pagure.io - https://pagure.io/fedora-infrastructure/issue/11142 19:03:24 low low ops... just needs doing sometime 19:03:39 +1 19:04:06 we dont have releng issues 19:04:14 #topic Planning, Upcoming work and Open floor 19:05:11 cool. ;) 19:05:46 So lets see... this week continues to be anoying... but hopefully it's settling down. :) 19:05:46 #info friday is a recharge day 19:05:52 #info we are in f38 beta freeze 19:06:37 I plan to work some more on my efforts to free up space on the koji netapp volume. Then, I have some misc stuff piled up like isolating a fasjson bug more, and tickets. 19:07:36 How about you phsmoura? 19:07:47 'recharge day' in infrastructure... does that mean anything 19:08:30 there's less other people around to bug you? ;) 19:09:07 I was taking a look at planet and was thinking how to migrate those users also the .ini file is a bit different when compared to venus 19:09:16 or there are more people because they aren't working 19:10:04 phsmoura: sounds good. 19:11:40 also couldnt figure it out how venus is building .ini from every user there must be something on cron maybe? I dont have access to it. do you know how it works nirik? 19:12:39 just to make sure if Ill have to do some scripting from scratch 19:12:50 venus is what is already in use? 19:13:00 yes thats it 19:14:22 yes, it's a cron. it's in ansible repo 19:14:45 roles/planet/files/ 19:14:51 ah cool. Ill take a look at it thanks :) 19:16:22 phsmoura: looking at the cron list I think it is whatever gets put in as /usr/local/bin/pull-run-planet-config.sh and a python script also 19:18:04 the same files I see in ansible repo 19:18:35 yep should be 19:18:55 but we still need to migrate those users right? 19:19:07 those that are in planet using vennus 19:19:42 well there is going to need to be some way to get the users configs into a place that the container can see 19:19:43 to add their blog they need to ssh to fedora blog and add the file 19:19:51 according with the docs 19:20:44 basically they have to log into fedorapeople.org and then add a file which the script consolidates into a new ini every hour 19:20:46 no no. 19:20:57 Thats the current process yes. 19:21:09 there will need to be a new process I guess 19:21:18 but we want the new one to be: update your blog url in accounts.fedoraproject.org 19:21:20 sorry I will let nirik explian 19:21:20 and we pull from that 19:21:47 so, yes, once we have the new process working, we will need to tell users about it and setup a cutover time. 19:22:19 so the new script needs to do some sort of ldap query filtered to that field and create a new ini file (or similar) that pluto? will use 19:22:25 ahh cool. did some queries at fasjson but that blog string was null for lot of users 19:23:41 curl -L --negotiate -u : https://fasjson.fedoraproject.org//v1/users/kevin/ | jq | grep website 19:23:41 shows mine for example 19:24:21 yeah, likely not many people know about it... because it currently doesn't do anything. 19:24:46 we should probibly open a fasjson ticket explaining that we want to get website from all 'contributors' and ask if a endpoint could be added to get that easily. 19:24:52 then it could just be one call... 19:25:37 yeah.. better than my ldapsearch idea 19:25:52 nirik: so this new process will need also another script to create .ini file. will it be at cron of fedora planet? inside the container? 19:27:08 yeah, I would think it would make sense to just have it be part of the cron... 19:27:24 ie, 'get list from fasjson' then update ini, then get feeds? 19:28:14 makes sense 19:28:52 There's 5470 users in the 'fedora-contributor' group. :) But I bet almost none of them have website set. 19:29:05 phsmoura: can you file that fasjson ticket? 19:29:32 https://github.com/fedora-infra/fasjson/ 19:30:08 Ill see what could be reused from those scripts in ansible repo, but because the ini file is a bit different maybe it will need some changes 19:30:25 yeah. 19:30:46 Diego H (he/him): did you get mbs building? I guess we need to schedule upgrading it after freeze now? 19:30:53 yes 19:31:04 nirik: what ticket? sorry didnt follow 19:31:22 https://koji.fedoraproject.org/koji/buildinfo?buildID=2154017 19:31:44 "we should probibly open a fasjson ticket explaining that we want to get website from all 'contributors' and ask if a endpoint could be added to get that easily." 19:32:12 Diego H (he/him): nice. Good detective work on finding out what was breaking it. Can you update the ticket on it? 19:32:19 sure 19:32:57 I will also add on getting it built on EL8 and EL9, it's viable, but needs more dependencies, if it's something we need in the future :) 19:34:08 yeah, would be nice... 19:34:25 ah ok, sure I can 19:34:41 Diego H (he/him): if you are up for more packaging work, we could use some work on infra-sig packages. ;) 19:34:51 phsmoura: excellent. thanks. 19:35:15 sure :D just point at something and i'll go for it ,) 19:36:12 I know that a cleanup was needed, but it's hard to do without knowing which packages specifically are being used and which are dependencies for those 19:36:29 nirik: just one more quick question. Planet will be using a service account to query from fasjson? 19:36:43 on a short term/easy scale, there's a number of packages that have updates pending. I usually just update them in rawhide unless there's a need to build elsewhere. 19:37:06 but yeah, on a longer term, infra-sig has a bunch of packages and there's very likely some we could drop/don't care about anymore. 19:37:34 so I guess we need to make some kind of dependency graph and look at it vs our ansible repo and see whats the intersection 19:38:03 or look at our big apps and follow deps down? like bodhi needs X,y,z and koji needs a,b,c, etc 19:38:18 Not sure. I guess we could start a thread on the mailing list or setup some meeting on it. 19:38:52 phsmoura: yeah, once it's in our infra. Not sure what to do on communishift. I guess for testing we could get your a staging keytab? 19:39:43 ok :) 19:40:04 the dependency graph idea might be cool, i'll look into that to see how hard it gets :) 19:40:41 There's some packages I think we have because of TurboGears, but we have no TG apps left I don't think. 19:40:43 (for example) 19:41:16 also we have a few we should add... like I think rpmautospec has main admin of nils... 19:41:27 anyhow, sorry for sidetracking. 19:41:47 end meeting? 19:42:07 +1 19:42:26 sorry to bomb you with so many questions today 19:42:48 thank you all for joining :) 19:42:51 not at all... questions are great! 19:43:11 #endmeeting