Thursday, 2020-12-17

*** tosky has quit IRC00:01
*** brinzhang has joined #opendev01:03
*** brinzhang0 has quit IRC01:06
openstackgerritIan Wienand proposed opendev/system-config master: bazelisk-build: build output in pwd
openstackgerritIan Wienand proposed opendev/system-config master: [wip] summary plugin
*** hamalq_ has quit IRC01:24
*** knikolla has quit IRC01:24
*** mnasiadka has quit IRC01:24
*** diablo_rojo_phon has quit IRC01:24
*** parallax has quit IRC01:24
*** Open10K8S has quit IRC01:24
*** jbryce has quit IRC01:24
*** cap has quit IRC01:24
*** jroll has quit IRC01:24
*** mlavalle has quit IRC01:26
*** hamalq_ has joined #opendev01:30
*** knikolla has joined #opendev01:30
*** diablo_rojo_phon has joined #opendev01:30
*** mnasiadka has joined #opendev01:30
*** parallax has joined #opendev01:30
*** Open10K8S has joined #opendev01:30
*** jbryce has joined #opendev01:30
*** cap has joined #opendev01:30
*** jroll has joined #opendev01:30
*** DSpider has quit IRC01:51
openstackgerritIan Wienand proposed opendev/system-config master: bazelisk-build: build output in pwd
openstackgerritIan Wienand proposed opendev/system-config master: [wip] summary plugin
*** hamalq_ has quit IRC03:00
*** ysandeep|away is now known as ysandeep03:24
openstackgerritIan Wienand proposed opendev/system-config master: bazelisk-build: build output in pwd
openstackgerritIan Wienand proposed opendev/system-config master: [wip] summary plugin
*** gnuoy has quit IRC04:19
*** gnuoy has joined #opendev04:19
openstackgerritIan Wienand proposed opendev/system-config master: [wip] summary plugin
*** ykarel has joined #opendev05:00
ianw#status log restarted nb01/02 with dib 3.5.0 in builder container to fix centos-8 builds05:30
openstackstatusianw: finished logging05:30
ianwi'm refreshing /opt/dib_tmp, where we store the git caches, as some trees seemed to be giving "dangling commit" messages05:31
*** marios has joined #opendev06:06
*** jaicaa has quit IRC06:28
*** jaicaa has joined #opendev06:30
ianwWOO HOO!
ianwplugin installing, clicked on, screenshot taking ... everything finally wired up!06:47
*** ralonsoh has joined #opendev07:14
*** eolivare has joined #opendev07:32
*** slaweq has joined #opendev07:34
*** slaweq has quit IRC07:38
*** sboyron has joined #opendev07:49
*** slaweq has joined #opendev07:50
*** slaweq has quit IRC08:01
openstackgerritIan Wienand proposed opendev/system-config master: gerrit: Install zuul-summary-results plugin
*** slaweq has joined #opendev08:08
*** hashar has joined #opendev08:10
*** andrewbonney has joined #opendev08:11
openstackgerritDaniel Pawlik proposed opendev/system-config master: Added Fedora 33 mirror-update role
openstackgerritSlawek Kaplonski proposed openstack/project-config master: Update Grafana dashboard for Neutron jobs
*** tosky has joined #opendev09:12
*** DSpider has joined #opendev10:37
danpawlikfungi, ianw: hey, could you take a look on: please?10:46
*** tosky has quit IRC10:47
*** tosky has joined #opendev10:47
danpawlikianw: I asked dmsimard to check if ara needs f31 nodeset. Also made change for zuul-operator. I don't see more projects related to f31.  I decide to split the mirror update to two patch sets, so first one is here:
*** lbragstad has quit IRC10:52
*** lbragstad has joined #opendev10:55
*** rpittau|afk is now known as rpittau10:59
*** hashar is now known as hasharLunch11:36
*** lbragstad has quit IRC11:38
*** lbragstad has joined #opendev11:42
*** brinzhang_ has joined #opendev11:57
*** brinzhang has quit IRC12:00
*** ykarel_ has joined #opendev12:15
*** ykarel has quit IRC12:17
*** ykarel_ is now known as ykarel12:32
dmsimarddanpawlik: I wasn't awake yet :p13:35
dmsimardI can take care of it sometime soon13:35
danpawlikdmsimard: cool, thanks13:35
*** hasharLunch is now known as hashar13:47
*** lbragstad has quit IRC14:21
*** lbragstad has joined #opendev14:23
*** brinzhang0 has joined #opendev14:46
*** brinzhang_ has quit IRC14:49
*** pyn38 has joined #opendev14:59
*** ysandeep is now known as ysandeep|away15:17
*** nautik has joined #opendev15:25
*** hashar has quit IRC15:30
*** mlavalle has joined #opendev15:45
*** ykarel has quit IRC16:07
*** hashar has joined #opendev16:38
*** marios is now known as marios|out17:00
*** marios|out has quit IRC17:05
hasharcorvus: good morning. I have noticed you have setup Zuul on   do you manage the admin changes to zuul/ops there or should I ask someone else? :]17:07
*** pabelanger has joined #opendev17:15
corvushashar: yes, i do, and mordred, and some other folks17:15
hasharcorvus: great :]  the context is to enable Zuul for  plugins/zuul.git if you get some minutes to waste  :]17:16
corvushashar: oh i see a change from you, sorry i need to do something about my notifications from that gerrit17:16
hasharthat is the super old plugin written by Khai Do  which we have adapted for Gerrit 3.2 / Polymer17:16
corvushashar: cool; i should go ahead and write the same change for the new summary plugin17:17
hasharprobably yeah17:18
hasharit is quite nice to see Zuul on upstream Gerrit Gerrit since those two tools are so closely related17:19
openstackgerritAlbin Vass proposed zuul/zuul-jobs master: Zuul Cache role with s3 implementation.
*** pyn38 has quit IRC17:21
hasharthank you corvus !17:21
corvushashar, mordred, ianw: remote: Add zuul-results-summary plugin [NEW]17:22
hasharand I see `recheck` is replaced by a `Re-Run` link quite nice17:22
*** ralonsoh has quit IRC17:23
corvushashar: yeah, there are some nice things about the check plugin; i hope the new stuff ends up working for us17:23
mordredcorvus: submit at will17:23
hasharI attended the virtual summit a month or so ago17:23
hasharif I get it right they largely enhance it so that CI system can provide what is currently in the Zuul build result17:24
pabelangermordred: corvus: re: python-builder / python-base images.  I think it is worth the discussion to have them support for distros. I've found the ability to use them, in minimal Dockerfile powerful. As a quickly way to 'package' a python project in a container.  The main things for ansible use-case is to support centos:8 / ubi:8.  our fork really just added support for dnf:17:24
mordredpabelanger: cool - I don't have much skin in the game today - but I could see incorporating that work into the opendev versions - and that it could improve life for some people17:26
corvuspabelanger: that sounds reasonable to me; i think opendev would be a fine home for all of those; maybe ansible/opendev/tripleo could collaborate there and i'm sure we would love to publish them to quay as well17:26
pabelangeralong with remove some unused bits too17:27
pabelangersure, this is all coming for the use case of execution environments for ansible. So I imagine opendev / tripleo would actually end up using them too17:28
mordredpabelanger: do you use both centos and ubi? or just ubi?17:28
pabelangercorvus: I think I asked this before, could we still use depends-on from opendev/python-builder changes, in zuul.ansible.com17:28
pabelangermordred: both right now. We are discussing if we need both or can just move everything to ubi17:29
pabelangerit is the upstream / downstream talk basically17:29
pabelangercorvus: for container jobs17:29
fungipabelanger: yes, if you add a connection for the opendev gerrit in it should work17:30
hasharcorvus: the related zuul/config change is feel free to hijack it and add the other plugin ;]17:31
pabelangercool, I haven't tested that step yet17:31
fungiit doesn't have to be configured to report to opendev's gerrit17:31
fungibut it could also do that, serving as a third-party ci system for those repos17:31
corvuspabelanger, fungi: but not for the container artifacts17:32
pabelangerah, okay. yah. it would be the artifacts we'd want17:32
corvuspabelanger, fungi: so ansible-zuul could use depends-on to build its own python-builder container image, but would not be able to use the provides/requires feature to obtain an opendev-zuul built container image17:33
fungioh, artifacts... yeah that would be tricky17:33
corvusthat only works within a single tenant17:33
mordredthey might want to build their own images anyway ...17:33
mordredbut in this case they could be very slim downstream images17:33
mordredoh - although it still wouldn't trigger the full rebuild of upstream layers - so there are places where it could still be a hole17:34
pabelangermordred: so, use shared Dockerfile, but publish to different namespace17:34
* mordred goes back to his zoom meeting17:34
corvusi mean, if that's a use case we want to try to support, we can try to engineer something for it.  it'd essentially be cross-zuul-dependencies.  (that would be our third ground-breaking cross-*-dependencies system)17:35
mordredcorvus: cross-zuul container layer depends-on would be super cool. my brain balks at thinking about _how_ it would work :P)17:36
corvusmordred: that probably means it's a good idea :)17:37
*** hamalq has joined #opendev17:50
*** hamalq_ has joined #opendev17:52
*** hamalq has quit IRC17:56
prometheanfireglean seems to append to /etc/hosts every reboot on deb-10, not sure if this is something I should look into fixing in case someone else fixed it at some point18:16
fungisounds worth fixing, we likely don't see it because we (almost) never reboot job nodes18:18
*** rpittau is now known as rpittau|afk18:20
*** hashar is now known as hasharDinner19:03
*** andrewbonney has quit IRC19:11
*** AJaeger has joined #opendev19:18
AJaegerinfra-root, config-core, Sorry, I was a bit longer offline after my vacation and used the time for reflection: My job responsibilities have changed some time ago and I cannot anymore be as engaged with OpenDev, OpenStack, and Zuul. It's time for me to move on now.19:20
AJaegerI plan to stop being on IRC - but plan also to continue reviewing once a week parts of the open queue. I do not expect to drive any changes. If I should stop reviewing, please tell me.19:20
AJaegerIt has been an absolute pleasure beeing part of this team and I made many friends. Thanks everybody!19:20
AJaegerIf you have any questions, feel free to email me and I'm happy to jump on IRC or answer per email...19:21
fungiAJaeger: no need to apologize! and your help has been invaluable and will always be appreciated in whatever capacity you find19:22
AJaegerthanks, fungi!19:23
* AJaeger hopes to meet again some time - Berlin this year was not meant to be ;(19:25
fungiyeah, i'm hoping we still get to have a conference there again soon after travel becomes possible again19:26
AJaegerlet's see when that will be...19:27
AJaegerFor now, I'd like to wish everybody all the best and merry christmas!19:33
fungithanks, and to you as well AJaeger, schöne feiertage!19:41
*** dviroel has quit IRC19:43
corvusAJaeger: thanks and merry christmas! stop in on irc and say 'hi' any time :)19:43
AJaegercorvus: thanks, corvus19:45
zbr|roverAJaeger: sorry to see your going. Only good wishes!19:45
AJaegerfungi, danke!19:45
AJaegerthanks, zbr|rover19:45
* AJaeger signs off again and waves good night!19:56
*** AJaeger has quit IRC19:56
*** hasharDinner is now known as hashar20:11
*** eolivare has quit IRC20:32
openstackgerritIan Wienand proposed opendev/system-config master: gerrit: get files from bazel build dir
openstackgerritIan Wienand proposed opendev/system-config master: gerrit: Install zuul-summary-results plugin
ianwinfra-root: the stack of changes to implement the summary plugin is ready for review @
ianwboth builders are spinning trying to build opensuse-15 at this point20:53
ianwgiven everyone's disappearance, might be best to pause that to try and keep the builders happy over break20:54
*** slaweq has quit IRC20:59
mordredianw: in the new gerrit - there is a thing on the right "relation chain"21:05
mordredianw: are the patches on the top of that list the patches on the top of the stack or the bottom?21:05
mordred(that is - is gerrit: Install zuul-summary-results plugin first or last?21:06
mordredI'm guessing last - just reading the summaries21:06
ianwmordred: yeah, is the last, and is the first21:07
mordredianw: that's sexy21:11
mordred(I mean that both honestly, because it's a super cool system, and ironically because it's just Loading...)21:11
ianwmordred: ahh, yes i'd welcome contributions on the selenium21:12
mordreddude - I'm impressed it exists at all21:12
ianwi can't quite figure out the best hook to wait for.  polygerrit/polymer uses shadowDom everywhere which isn't that well supported21:13
ianwfor example, you can't use xpath to select elements inside the shadow dom.  so basically all selenium is seeing is a "<gr-app>" tag21:14
*** hashar has quit IRC21:14
ianwin i figured out how to click by navigating the shadowDom via repeated javascript .querySelector().shadowRoot.querySelector.shadowRoot. chain21:15
ianwi guess we probably need to write a custom javascript "wait" function that polls a shadowRoot chain like that, then run it with selenium's "execute_script"21:17
mordredianw: yes - I think that's probably right. it's a much more real set of words that is sort of what my non-informed brain was trying to get at in my inline comment on that patch21:19
mordredianw: that's an awesome stack btw21:19
mordredianw: out of curiosity - your testing right now is just leaving a comment in zuul format - do you have any plans to make a version of the job that stands up a zuul (sort of like how quick-start works?)21:21
ianwit's friday; on wednesday i had never typed "querySelector" ever I'm sure, had no idea what a shadowDom was, and the last time i used xpath was with xml probably 20 years ago at university.  things change quickly in the web world :)21:21
mordredianw: you're a javascript expert now21:21
mordredianw: last year I made a travel website for sandy and I - super simple - there's literally three pages. EXCEPT - I now somehow have a react site that uses graphql to load data that drives a mapbox-gl map - and i'm running a tileserver to back it up. silly me for thinking I'd just make a simple map :)21:23
ianwmordred: it's not an immediate plan; i figured it's easy enough to fiddle the comment injection if/when we find things that trip up the plugin to keep it sane.  i'd like to add a older-style 3rd party comment in the testing too21:23
ianwmordred: lol21:23
*** sboyron has quit IRC21:24
ianwthe last time i made a travel site, i used, i can't even remember what it's called.  where you could split an image up and make parts of it clickable21:24
ianwlike 1999 era21:24
ianwtake that javascript21:27
fungioh, yep, i remember doing that21:30
fungiyou also find echoes of it in svg nowadays, since you can associate target hrefs with elements or regions within svg files21:31
openstackgerritIan Wienand proposed openstack/project-config master: nodepool: pause opensuse-15 builds
*** slaweq has joined #opendev22:20
*** DSpider has quit IRC22:23
*** slaweq has quit IRC22:33
*** dviroel has joined #opendev22:58
*** pyn38 has joined #opendev23:09
pyn38Hi Jeremy23:09
pyn38I have a small query23:09
pyn38regarding the new project I had created
pyn38I am not able to merge the patches to this project repo. Can you please let me know what might be the reason?23:12
*** lbragstad has quit IRC23:18
fungipyn38: did you see the comments i added in response to the questions on your initial change?23:19
fungiyou need to add zuul jobs in the first change and then if they pass in the check pipeline, approve the change so that zuul will rerun a final set of jobs and merge it for you when they succeed. i linked to the relevant sections of our infra manual23:20
pyn38Oh okay! Thanks Jeremy. Somehow missed those messages. Let me check.23:35
fungino worries, i'm happy to answer more questions if you have any, but we've gone to a lot of effort to document this23:36
*** pyn38 has quit IRC23:40
ianwfungi: would you mind a look at to pause the opensuse-15 builds that are failing?23:40
ianwwe have publicly available logs, but at this point i can't see it getting fixed for a prolonged period due to breaks23:41
*** pyn38 has joined #opendev23:41
ianwi don't want to leave it in a failing loop23:41
pyn38Thanks fungi. I had missed to add .zuul.yaml file in my initial commit. Just to be sure my should have been something like this commit right?23:43
*** tosky has quit IRC23:43
ianw2020-12-17 22:38:58.463716 | TASK [ensure-java : Install JDK]23:46
ianw2020-12-17 22:55:51.106866 | ubuntu-bionic | changed23:46
ianwa couple of times i've seen the gerrit jobs timeout in this step23:46
ianwwell, due to the long time that step takes23:46
ianwboth ran in airship citycloud, i wonder if the mirror is particularly slow23:49
ianwFetched 273 MB in 16min 22s (278 kB/s)23:49
ianwthat host does not seem to want to let me in23:50
fungipyn38: you might ought to leave out the equivalent of the stx-cert-manager-armada-app-upload-git-mirror job and the associated secret for now, someone in starlingx with access to the github credentials will need to propose that addition23:53
pyn38Yes, Thanks for that input. The mirroring part and the secret keys I'll leave for now and rest should be okay I believe. AM I right?23:56

Generated by 2.17.2 by Marius Gedminas - find it at!