Wednesday, 2022-08-24

*** ysandeep|out is now known as ysandeep04:25
*** soniya29 is now known as soniya29|ruck05:50
chandankumardpawlik: hello06:29
chandankumardpawlik: can I update this testproject https://review.rdoproject.org/r/c/testproject/+/4438806:29
chandankumarwith correct config06:29
chandankumar?06:29
dpawliksure chandankumar06:40
chandankumardpawlik: updated, thanks!06:52
*** ysandeep is now known as ysandeep|afk07:00
*** jpena|off is now known as jpena07:34
*** soniya29|ruck is now known as soniya29|ruck|lunch08:09
*** soniya29|ruck|lunch is now known as soniya29|ruck09:37
ysandeep|afkrlandy|out, retro clashes with today's operator mtg.. Should we cancel operator mtg?10:15
*** ysandeep|afk is now known as ysandeep|break10:16
*** rlandy|out is now known as rlandy10:37
rlandyysandeep|break: neither pojadhav pls move the retro to thursday as we discussed last time10:38
rlandypojadhav: hi - let's talk re: sched10:38
pojadhavrlandy, sure10:39
rlandybhagyashri - pls join if you are around10:39
rlandyhttps://meet.google.com/bae-jvvs-sfb?pli=1&authuser=010:39
rlandypojadhav: ^^10:39
frenzyfriday0/ new vm for downstream cockpit - 10.0.111.46 . After setting up the cockpit there I'll update bitwarden10:53
*** ysandeep|break is now known as ysandeep11:03
*** vkmc_ is now known as vkmc11:06
ysandeeprlandy, pojadhav: thanks for sorting out the schedule.11:06
*** tweining___ is now known as tweining11:08
*** pojadhav is now known as pojadhav|afk11:14
rlandyreview time anyone?11:17
ysandeeprlandy, joining11:18
rlandyysandeep: dropped11:19
rlandythere was nothing on today's list11:19
* rlandy looking for akahat's review11:19
ysandeeprlandy, ack11:19
rlandyakahat: do you still have a review that needs to be merged?11:20
rlandyhttps://review.rdoproject.org/r/c/rdo-jobs/+/4459111:22
rlandyakahat: thanks - looks good now11:22
rlandychandankumar: ^^ can you second core review - then we can merge11:22
chandankumarsure11:23
rlandyfrenzyfriday: hi - what's the status of internal cockpit11:23
frenzyfridayrlandy,  I have the vm, deployment is in progress11:23
chandankumarI thought it got merged11:24
rlandyjm1: do we need to keep the internal cockpit on hardware because of the domain name?11:24
rlandycan we assign that to a vm on PSI?11:24
rlandyfrenzyfriday++11:24
frenzyfridaywe might have to send a mail regarding the domain again11:24
akahatrlandy, that's from marios need to merge now.11:25
chandankumarfrenzyfriday: please make sure we get a patch out also for moving to podman & podman compose11:25
akahatrlandy, chandankumar thanks!11:25
frenzyfridayyep11:25
chandankumarfrenzyfriday: please keep the old vm rafeal wants to upgrade that11:25
*** dviroel is now known as dviroel|rover11:26
dviroel|rovero/11:26
soniya29|ruckhello11:26
rlandydviroel|rover: did a sync with soniya29|ruck ... we went over program call doc11:31
rlandythere is a legit new bug on internal11:31
rlandydviroel|rover: soniya29|ruck: actually, let's sync quickly - will be easier11:31
dviroel|roverack11:32
soniya29|ruckack11:32
rlandyhttps://meet.google.com/giv-cuob-wjs?pli=1&authuser=011:32
rlandyhttps://opendev.org/openstack/tripleo-ci/src/branch/master/roles/prepare-node/tasks/main.yaml#L4311:37
rlandyysandeep: hey - let's touch base about the confluence page11:46
rlandyI added some stuff there and then read your wiki11:46
rlandywill ping after I drop from program call11:46
ysandeeprlandy, ack let me know when you want to touch base11:47
jm1rlandy, frenzyfriday: we cannot assign that internal name to a psi vm but we could create a new domain name. dont know how that is handled on psi though11:49
rlandyjm1: a domain name is really useful but running off really hardware seems dangerous11:50
rlandyjm1: frenzyfriday: I think let's go with the vm for now11:50
rlandyget it set up 11:50
rlandyeven if we have to use an ip for now11:51
rlandythen we can try upgrade the vm on the hardware11:51
rlandyso we have one working env to refer to if we mess that one up11:51
rlandysound ok?11:51
jm1frenzyfriday: do you use scripts in ci-config to provision and configure the internal vm?11:51
rlandyfrenzyfriday: then we can look into how to get a domain name on psi11:52
jm1rlandy, frenzyfriday: i was not in the loop for the psi vm for downstream cockpit. how do you do it? manually? ansible?11:52
rlandyjm1: ansible11:53
rlandywith some internal repos11:53
* akahat back in hour.11:54
jm1rlandy, frenzyfriday: afaiu sagi provisioned the downstream cockpit as a libvirt vm (in 2019?). the vm provisioning was probably done manually because we dont have code for that. the configuration inside the vm was (and is) done using ansible playbooks in ci-config repo. its actually the same code as for upstream cockpit with small changes.11:58
frenzyfridayjm1, I managed to find till here: https://github.com/rdo-infra/ci-config/blob/master/ci-scripts/infra-setup/grafana_playbook.yml11:58
frenzyfridayin out infra deployment workflow what actually runs this script?11:58
rlandyjm1: ack - the vm itself11:59
jm1rlandy, frenzyfriday: all our vms on vexxhost are using ansible in pull mode. so be careful when changing ci-config because our running vms will pull that code and update their config11:59
jm1^ applies to downstream cockpit as well12:00
rlandyjm1: that's a good thing12:00
rlandywe will get the updates12:00
rlandyin upstream and downstream12:00
rlandyit should be the same way12:00
rlandyjoin program call12:00
rlandyjoining12:00
rlandysorry - not an ask on anyone else (apart of rr to join)12:01
jm1rlandy, frenzyfriday: what i mean is that we have to be careful with manually doing stuff on vms because ansible is working in the background (in pull mode). we also have to be careful with updating ci-config for the downstream cockpit on psi because our downstream cockpit on our internal server will also pull these changes12:02
rlandyjm1: that should be the right way to operate12:02
frenzyfridayyep12:02
rlandyfor downstream to be able to pull any upstream changes12:02
frenzyfridayjm1, I can see https://github.com/rdo-infra/ci-config/blob/master/ci-scripts/infra-setup/tenant_vars/infra-tripleo/servers.yml#L55 - so looks like there should be a script which creates the Vm as well12:03
* rlandy will ask if we can get a new domain name on psi12:03
jm1frenzyfriday: right, this is for provisioning vms on vexxhost12:04
frenzyfridayjm1, do you have some time for a quick gmeet?12:05
jm1sure12:05
* frenzyfriday is looking for "the origin" of playbooks12:05
frenzyfridayjm1, https://meet.google.com/zgm-pnqt-zaw12:06
rlandysoniya29|ruck++ nicely done12:23
soniya29|ruckrlandy, thanks12:24
rlandydasm|off: pojadhav|afk: soniya29|ruck: I will miss the planning meeting today - have a clashing meeting12:30
soniya29|ruckrlandy, okay12:30
rlandysoniya29|ruck: pls will you create that BZ for standalone component so we have it for the CIX call. thanks12:31
soniya29|ruckrlandy, ohh sorry, i left that i middle12:31
rlandypojadhav|afk: pls ping when you have the planning section created and I'll add my notes12:32
rlandyysandeep: time now to talk about confluence?12:33
ysandeeprlandy, sure12:33
soniya29|ruckarxcruz, rlandy, kopecmartin_ , chandankumar, ysandeep, mind if i move tempest meeting to next wednesday since clashing with retro?12:33
rlandyysandeep: https://meet.google.com/?hs=197&pli=1&authuser=012:33
rlandysoniya29|ruck: pojadhav|afk is moving the retro12:33
rlandyysandeep: sorry - https://meet.google.com/wnn-qmda-yhp?pli=1&authuser=012:34
soniya29|ruckshe is moving? i can see it still on calendar..so i got confused12:34
arxcruzrlandy so, can we keep it? we have a guest today 12:34
soniya29|ruckrlandy, ack, no problem12:34
*** pojadhav|afk is now known as pojadhav12:35
pojadhavrlandy, ack12:36
pojadhavrlandy, added new section for new sprint @line 31 - https://hackmd.io/P5sTPMSBREe5kBS_4jDhIg 12:37
pojadhavsoniya29|ruck, dasm|off, ysandeep, chandankumar : please add new new topics if any @line 31 - https://hackmd.io/P5sTPMSBREe5kBS_4jDhIg12:44
soniya29|ruckrlandy, https://bugzilla.redhat.com/show_bug.cgi?id=212109412:47
rlandyysandeep: dviroel|rover: so we have anything on the agenda for today's podified meeting?12:50
rlandyif now - maybe we can run through the plans for the POC?12:51
rlandyI'll move the meeting minutes in my afternoon  - so you can take notes on the hackmd for today12:51
rcastilloo/12:51
rlandysoniya29|ruck: thanks12:51
ysandeeprlandy, nothing on agenda, sounds good to discuss POC12:51
rlandyrcastillo: hey - just confirming - retro is tomorrow12:51
rlandyysandeep: we should also mention move of info to confluence12:52
rlandyrcastillo: you are up as the fearless retro leader, right?12:52
rcastillorlandy: yup12:53
dviroel|roverrlandy: ysandeep: agree12:54
rlandysoniya29|ruck: pls raise a CIX for that BZ13:00
soniya29|ruckrlandy, done13:00
arxcruzkopecmartin_ soniya29|ruck will you join?13:01
rlandysoniya29|ruck: email to rhos-dev? 13:01
rlandysoniya29|ruck: to get a CIX we need that from a BZ13:02
soniya29|ruckrlandy, yeah, its done13:03
soniya29|ruckrlandy, i have forwarded you the email, please check13:04
rlandysoniya29|ruck: thank you - idk  I didn't see that before13:05
soniya29|ruckrlandy, no problem :)13:06
*** kopecmartin_ is now known as kopecmartin13:10
*** dasm|off is now known as dasm13:27
dasmo/13:27
* soniya29|ruck needs a break..will be back in 10-15 mins13:27
*** soniya29|ruck is now known as soniya29|ruck|afk13:27
dviroel|roversoniya29|ruck|afk: rlandy: downstream issue is related to https://opendev.org/openstack/tripleo-ci/commit/f80cf14b624f421d407854b1adddd423aa53fa0f13:29
dviroel|roveri will propose a fix on downstream13:30
rlandydviroel|rover++ thank you13:32
rlandydasm: hello ...13:33
dasmrlandy: o/13:33
rlandyleft notes on planning meeting13:33
dasmack13:33
rlandyhave a clash13:33
*** soniya29|ruck|afk is now known as soniya29|ruck13:40
dviroel|roverrlandy: there are more jobs affected, fixing13:44
rlandydviroel|rover: need help there?13:45
dviroel|roverrlandy: no, on review only13:47
frenzyfridayjm1, hey, do you know what server is provisioner here https://github.com/rdo-infra/ci-config/blob/master/ci-scripts/infra-setup/provision-all.yml#L8? localhost?13:48
dviroel|roverrlandy: chandankumar: https://code.engineering.redhat.com/gerrit/c/openstack/tripleo-ci-internal-config/+/42593614:23
rlandydviroel|rover: chandankumar: ^^ looks ok to me  - waiting on chandankumar to vote  - then will merge14:40
chandankumardone14:41
chandankumardviroel|rover++ for fixing it14:42
jm1frenzyfriday: provisioner is defined here, its ansible_connection=local aka executed on the ansible controller. https://github.com/rdo-infra/ci-config/blob/master/ci-scripts/infra-setup/inventories/inventory.ini14:45
jm1frenzyfriday: you might have to call ansible-playbook with "-i inventories/inventory.ini" to get it working14:45
*** ysandeep is now known as ysandeep|dinner14:46
dasmhmmm, why ovb one time works, but the other doesn't? (rhetorical question)_14:58
pojadhavchandankumar, ysandeep|dinner soniya29|ruck : UA/TC/PM sync15:02
dasmjm1: some time ago you mentioned about errors spilling out to journalctl for telegraf. I tried to find it in the history but failed. Can you point me where is this instance?15:10
dasmjm1: wrt TRIPLEOCI-115515:10
chandankumarsee ya!15:17
dasmchandankumar: o/15:17
dviroel|rovero/15:18
* pojadhav leaving for the day15:18
*** pojadhav is now known as pojadhav|out15:18
jm1dasm: frenzyfriday is working on downstream cockpit, not sure what the status is. i dont want to interrupt her work so i cannot check atm. 15:22
jm1dasm: i am not sure if journalctl was spilling out the errors or docker logs.15:23
frenzyfridayjm1, you can work on the centos7 vm if you want. I am working on the new vm15:23
dasmjm1: nah, i wanted to check that, but i don't remenber the instance :)15:23
frenzyfridayin the centos7 vm telegraf should throw errors if you try to rebuild15:23
jm1frenzyfriday: ack but the old instance is still running?15:24
*** dviroel|rover is now known as dviroel|rover|lunch15:24
frenzyfridayyep15:24
frenzyfridayjm1, we are just not getting new patches15:25
frenzyfridayjm1, ansible-playbook infra-setup/provision-all.yml  -i inventories/inventory.ini did not work as well15:25
frenzyfridaysomewhere I am missing host entries15:25
jm1frenzyfriday: we can debug that tomorrow if you want15:25
frenzyfridayjm1, cool, I'll ping you tomorrow morning15:26
jm1dasm: login to tripleo-cockpit.lab4.eng.bos.redhat.com and get root. then 'docker logs --since "2022-08-01" telegraf_py3 2>&1 | less' shows you some errors and journalctl shows others15:26
jm1frenzyfriday: sure!15:26
dasmjm1: cool! i was looking for that instance. thx15:27
* soniya29|ruck leaving for the day15:31
frenzyfridayjm1, it works now, I was in the wrong dir XD Also, we were wondering why the directory was called infra-tripleo - it has to be the name of the tenant on openstack15:32
*** ysandeep|dinner is now known as ysandeep15:36
dasmfrenzyfriday: you mentioned about issues to build docker? we're still running cs8 with py3.6 on telegraf. is that why?15:36
frenzyfridaydasm, no, docker pull quay.io/centos/centos:stream9 doesnt work on the centos 7 vm15:37
frenzyfridaydue to manifest version mismatch15:37
dasm:(15:37
dasmack15:38
frenzyfridaySo when you try to rebuild telegraf it fails at docker pull15:38
dasmso, there are no updates for last few weeks :(15:38
dasmgotcha15:38
frenzyfridayBut the already build image should not cause any problems15:38
frenzyfridayyep, exactly :(15:38
dasmhmmm... can we revert back to docker repo for cs9 until we have a new vm up and running?15:39
dasmwould it be useful?15:39
dasmhmm, probably not: > Last pushed a year ago15:39
* frenzyfriday running the server setup role for incockpit - hope nothing breaks in psi15:47
ysandeeprlandy, dviroel|rover|lunch fyi.. After you left there were suggestion that we do things in baby steps, and as a first step just build check/gate jobs without promotion pipeline(and use the current job that promote the latest operator image in post merge). I will update the confluence page in my morning.15:48
rlandyysandeep: dviroel|rover|lunch: let's meet on friday and talk about what makes sense for us15:51
rlandylisten to all suggestions and see what is the right approach15:51
rlandyysandeep: dviroel|rover|lunch: I'd like to see your take on this  - w/o a whole group meeting15:52
ysandeepThere are already jobs that push the latest operator image in post merge step(there are some github action jobs that does that) and we are already attempting to first setup check jobs.. So I think its not a bad idea and actually more practical for Nov meetup timeline. but let's chat more on Friday - I wan to listen dviroel|rover|lunch take on this as well.15:55
ysandeepThey are not saying no for promotion pipeline - just as the next step15:56
* ysandeep out, see you tomorrow 15:58
*** ysandeep is now known as ysandeep|out15:58
rlandyysandeep|out: dviroel|rover|lunch: ok - book us a slot for friday16:02
rlandylunch - biab16:05
*** dviroel|rover|lunch is now known as dviroel|rover16:21
*** jpena is now known as jpena|off16:42
jm1frenzyfriday: @infra-tripleo tenant directory name: very good observation! please please document that in readme :)17:33
* dasm is going for a walk. bbiab17:49
* dasm is back18:19
dviroel|roverrdo zuul seems to be slow again18:37
rlandyslow how?18:43
dviroel|roverrlandy: lots of jobs on queue, but there isn't so many jobs running18:43
rlandyk - see messages on rhos-ops18:43
dasmdviroel|rover: i see some nod failures on my jobs. some of them stiil haven't started.18:45
dasmit's been in that state for last x minutes.18:45
dasmit's kinda bothering18:46
dviroel|rovervexx is working on another fix18:47
dviroel|roverlet's see18:47
dasmack18:47
rlandydasm: looking at planning meeting notes ... https://hackmd.io/P5sTPMSBREe5kBS_4jDhIg - line 4118:54
rlandyhow is that relevant? to 17.1?18:55
dasmrlandy: i thin it was mis copied. pojadhav|out ^18:56
dasmrlandy: oh, i think i know why it is there. it was used as a reference why we're in need to enable 17.118:56
rlandyoh ok18:57
dviroel|roverrlandy: already validated that BZ 2121094 is fixed, i will add logs and close19:13
rlandydviroel|rover++ 19:13
rlandydviroel|rover: kicked a few testprojects downstream for some missing jobs19:14
dviroel|roverrlandy: yeah, i saw, thanks19:14
rlandydviroel|rover: I guess we'll try rdo when zuul gets back from vacation19:14
dviroel|roveryep, a have a list of jobs to rerun alredy19:15
dviroel|roverlets see if remains stable19:15
dviroel|roverit is back already19:15
rlandyok - starting to read operator testing docs19:16
rlandydasm: you need anything?19:16
dasmrlandy: working zuul?19:17
dasmthat's really the issue atm. jobs either not starting, or taking forever to start.19:17
rlandyright19:17
rlandysorry - am pretty helpless there19:17
dasmnone of my jobs finished today. just trying to do other things -- not requiring zuul atm19:18
dasmrlandy: i know, it's not your fault. just venting out my disappointment :)19:18
dasmin the same time i know vexx people work to fix anything and everything. i've been there too :)19:18
rlandyk19:20
* jm1 have a nice evening #oooq 🍻19:22
* jm1 signing out19:23
dasmjm1[m]: o/19:25
rlandydviroel|rover: rdo still hosed?20:53
* rlandy tries testproject20:55
dviroel|roverrlandy: no, seems to be working now20:55
rlandydviroel|rover: ok - I want to test https://review.opendev.org/c/openstack/tripleo-ci/+/854426 on wallaby20:56
rlandyc8 and c920:56
dviroel|roverrlandy: ok, i will kick master again in a few minutes - last run was a mess20:58
dviroel|roverand I give up my almost promoted hash20:58
rlandynp - my testproject will just be two container builds20:58
rlandyand pushes20:58
*** dasm is now known as dasm|off21:32
dasm|offtake care team! i'm going offline!21:32
dviroel|roveroh tripleo - master is all red21:33
dviroel|rovercontainer-puppet-nova_metadata21:34
* dviroel|rover dinner22:30
*** dviroel|rover is now known as dviroel|rover|brb22:30
*** dviroel|rover|brb is now known as dviroel|rover23:26
* rcastillo leaves for the day23:52
rcastilloo/23:52

Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!