Tuesday, 2022-07-26

opendevreviewIan Wienand proposed zuul/zuul-jobs master: ensure-python: pull in venv package on Debuntu  https://review.opendev.org/c/zuul/zuul-jobs/+/85095700:19
opendevreviewIan Wienand proposed openstack/project-config master: translation updates: refactor Python 3.8 workarounds  https://review.opendev.org/c/openstack/project-config/+/85096201:08
opendevreviewIan Wienand proposed zuul/zuul-jobs master: ensure-python: pull in venv package on Debuntu  https://review.opendev.org/c/zuul/zuul-jobs/+/85095701:16
opendevreviewIan Wienand proposed openstack/project-config master: translation updates: refactor Python 3.8 workarounds  https://review.opendev.org/c/openstack/project-config/+/85096201:22
*** ysandeep|out is now known as ysandeep01:55
opendevreviewIan Wienand proposed zuul/zuul-jobs master: ensure-python: pull in venv package on Debuntu  https://review.opendev.org/c/zuul/zuul-jobs/+/85095703:00
opendevreviewIan Wienand proposed openstack/project-config master: translation updates: refactor Python 3.8 workarounds  https://review.opendev.org/c/openstack/project-config/+/85096203:03
opendevreviewIan Wienand proposed zuul/zuul-jobs master: ensure-python: pull in venv package on Debuntu  https://review.opendev.org/c/zuul/zuul-jobs/+/85095703:13
*** ysandeep is now known as ysandeep|afk03:21
opendevreviewIan Wienand proposed zuul/zuul-jobs master: ensure-python: pull in venv package on Debuntu  https://review.opendev.org/c/zuul/zuul-jobs/+/85095703:22
*** Guest6022 is now known as rcastillo04:22
opendevreviewIan Wienand proposed openstack/project-config master: translation updates: refactor Python 3.8 workarounds  https://review.opendev.org/c/openstack/project-config/+/85096204:55
*** ysandeep|afk is now known as ysandeep05:03
*** jpena|off is now known as jpena07:34
*** jpena is now known as jpena|off07:40
*** jpena|off is now known as jpena07:42
*** ysandeep is now known as ysandeep|afk08:02
*** pojadhav is now known as pojadhav|lunch08:06
*** pojadhav|lunch is now known as pojadhav08:19
*** tosky_ is now known as tosky09:10
*** rlandy|out is now known as rlandy10:34
opendevreviewIan Wienand proposed zuul/zuul-jobs master: linters: fix spaces between filters  https://review.opendev.org/c/zuul/zuul-jobs/+/85101410:51
opendevreviewIan Wienand proposed zuul/zuul-jobs master: linters: add names to blocks  https://review.opendev.org/c/zuul/zuul-jobs/+/85101510:51
opendevreviewIan Wienand proposed zuul/zuul-jobs master: linters: rename loop variable  https://review.opendev.org/c/zuul/zuul-jobs/+/85101610:51
opendevreviewIan Wienand proposed zuul/zuul-jobs master: linters: update to ansible-lint 6  https://review.opendev.org/c/zuul/zuul-jobs/+/85101710:51
*** marios_ is now known as marios11:00
*** dviroel__ is now known as dviroel11:05
fungipython 3.11.0b5 is out now. next stop: release candidates!12:04
*** ysandeep|afk is now known as ysandeep12:30
*** dasm|off is now known as dasm13:19
opendevreviewSlawek Kaplonski proposed openstack/project-config master: Add new project "tripleo-neutron-tempest-plugin" in the x/ namespace  https://review.opendev.org/c/openstack/project-config/+/85103113:39
opendevreviewSlawek Kaplonski proposed openstack/project-config master: Add new project "tripleo-neutron-tempest-plugin" in the x/ namespace  https://review.opendev.org/c/openstack/project-config/+/85103114:07
opendevreviewSlawek Kaplonski proposed openstack/project-config master: Add new project "tripleo-neutron-tempest-plugin" in the x/ namespace  https://review.opendev.org/c/openstack/project-config/+/85103114:37
opendevreviewSlawek Kaplonski proposed openstack/project-config master: Add new project "tripleo-neutron-tempest-plugin" in the x/ namespace  https://review.opendev.org/c/openstack/project-config/+/85103114:43
*** dviroel is now known as dviroel|lunch14:55
*** ysandeep is now known as ysandeep|out15:00
clarkbfungi: ianw: I guess 850957 is a low impact change to maybe address the issue 850917 is tryingto solve?15:27
clarkbthere are a number of changes related to that from overnight and I'm wondering what the path of least resistance is15:27
fungiclarkb: i'll admit i haven't been able to look at them closely enough yet15:42
*** dviroel|lunch is now known as dviroel16:01
*** marios is now known as marios|out16:16
*** jpena is now known as jpena|off16:33
ianwclarkb: right, 850957 is required by https://review.opendev.org/c/openstack/project-config/+/850962 which hopefully explains it19:02
clarkbaha thats the piece I was missing. Will review post meeting19:03
opendevreviewMerged zuul/zuul-jobs master: ensure-python: pull in venv package on Debuntu  https://review.opendev.org/c/zuul/zuul-jobs/+/85095719:27
opendevreviewIan Wienand proposed zuul/zuul-jobs master: linters: fix spaces between filters  https://review.opendev.org/c/zuul/zuul-jobs/+/85101419:53
opendevreviewIan Wienand proposed zuul/zuul-jobs master: linters: add names to blocks  https://review.opendev.org/c/zuul/zuul-jobs/+/85101519:53
opendevreviewIan Wienand proposed zuul/zuul-jobs master: linters: rename loop variable  https://review.opendev.org/c/zuul/zuul-jobs/+/85101619:53
opendevreviewIan Wienand proposed zuul/zuul-jobs master: linters: update to ansible-lint 6  https://review.opendev.org/c/zuul/zuul-jobs/+/85101719:53
opendevreviewIan Wienand proposed zuul/zuul-jobs master: upload-artifactory: fix version comparision  https://review.opendev.org/c/zuul/zuul-jobs/+/85108719:53
opendevreviewIan Wienand proposed zuul/zuul-jobs master: setup.py: override modules to stop auto-detection  https://review.opendev.org/c/zuul/zuul-jobs/+/85108819:53
opendevreviewIan Wienand proposed zuul/zuul-jobs master: test/upload-pypi : install venv  https://review.opendev.org/c/zuul/zuul-jobs/+/85108919:53
opendevreviewIan Wienand proposed zuul/zuul-jobs master: test/upload-pypi : install venv  https://review.opendev.org/c/zuul/zuul-jobs/+/85108920:08
opendevreviewIan Wienand proposed zuul/zuul-jobs master: linters: fix spaces between filters  https://review.opendev.org/c/zuul/zuul-jobs/+/85101420:08
opendevreviewIan Wienand proposed zuul/zuul-jobs master: linters: add names to blocks  https://review.opendev.org/c/zuul/zuul-jobs/+/85101520:08
opendevreviewIan Wienand proposed zuul/zuul-jobs master: linters: rename loop variable  https://review.opendev.org/c/zuul/zuul-jobs/+/85101620:08
opendevreviewIan Wienand proposed zuul/zuul-jobs master: linters: update to ansible-lint 6  https://review.opendev.org/c/zuul/zuul-jobs/+/85101720:08
opendevreviewIan Wienand proposed zuul/zuul-jobs master: ensure-kubernetes: install cri-dockerd  https://review.opendev.org/c/zuul/zuul-jobs/+/85109320:33
opendevreviewClark Boylan proposed opendev/system-config master: Run our base playbook on jammy  https://review.opendev.org/c/opendev/system-config/+/85109420:40
opendevreviewIan Wienand proposed zuul/zuul-jobs master: ensure-kubernetes: install cri-dockerd  https://review.opendev.org/c/zuul/zuul-jobs/+/85109321:03
clarkbianw: I'm looking at https://review.opendev.org/c/openstack/project-config/+/850962 and it modifies jobs on both focal and bionic but I think maybe we only want ot update the jobs on bionic? Specifically I'm worried we'll break the requirements update jobs on focal. But maybe the work in the pre.yaml playbook is happy for both?21:16
clarkboh except the new bits are only for bionic. I think maybe we need to keep the focal job the sam ethen?21:17
clarkbya I'll leave a note21:18
ianwclarkb: i think the focal path should be the same, effectively?21:18
clarkbianw: we won't install python3.8 on the focal job if I read it correctly21:18
clarkbI just posted a comment21:19
ianwright, but we don't need to do that on focal?21:19
clarkbwe do because the requirements jobs are trying to build requirements for both 3.8 and 3.921:19
clarkbthats what the comment on line 827 is about21:19
ianwwell i'm not sure, i don't see where python3.9 was being installed21:20
clarkbon line 836 of the old diff in the job def21:21
ianwoh hrm.  i think you're right, but actually it's from line 83421:22
clarkbbut I think the vars are merged into child jobs so we have to split the inheritance maybe?21:23
clarkbI'm not actually sur ehow to fix this immediately21:23
ianwsetting python_version will pull in the python3.9 packages21:25
*** rlandy is now known as rlandy|biab21:27
ianwsetting python_version on the job i think really confuses the matter over what's going on.  because it's still using python 3.8 too21:28
opendevreviewIan Wienand proposed openstack/project-config master: [dnm] additional discussion for requirements proposal  https://review.opendev.org/c/openstack/project-config/+/85109821:35
ianwclarkb / fungi: ^ i think we probably want to squash that change into it?  i think this makes it clearer overall what's going on21:36
opendevreviewClark Boylan proposed opendev/system-config master: Run our base playbook on jammy  https://review.opendev.org/c/opendev/system-config/+/85109421:36
fungilookin21:36
clarkbianw: I think that is still not installing python3.8 on focal though?21:38
ianwfor focal 3.8 is the base python which will be installed21:38
clarkbI guess it is the default on focal so maybe its already there?21:38
clarkbya that21:38
clarkband ya I think squashing is a good idea to avoid any further confusion with additional job failures if we land the two changes apart from each other21:39
ianw++21:39
fungia big continued concern of mine is that the requirements proposal jobs would be better run separately per interpreter version and combined after rather than trying to combine multiple versions in a single run21:42
fungia lot of the inherent complexity is trying to find ways to do it all in a single build21:43
ianw++++++++++++21:43
ianwmultiple people proposing changes to install multiple interpreters on multiple platforms via multiple roles has created quite the matrix21:45
ianw(not "throwing shade" ... everyone just trying to get things to work with the information we have at the time)21:48
fungiwell, it nearly came to a head this past openstack cycle when for a bit we were faced with a need to build requirements from python versions which we couldn't sanely install all of on a single platform21:48
fungiianw: anyway, in the near term, sure it seems like a way forward if we mash that in21:50
opendevreviewIan Wienand proposed openstack/project-config master: translation updates: refactor Python 3.8 workarounds  https://review.opendev.org/c/openstack/project-config/+/85096221:54
ianw^ that squashes everything, and adds another few paragraphs of changelog description 21:55
ianwas if it wasn't long enough21:55
clarkbit was definitely easier when all we had to worry about was python2.7 and a 3.x version21:55
ianwthanks for picking that up though!21:59
opendevreviewIan Wienand proposed zuul/zuul-jobs master: ensure-kubernetes: install cri-dockerd  https://review.opendev.org/c/zuul/zuul-jobs/+/85109322:03
opendevreviewIan Wienand proposed zuul/zuul-jobs master: ensure-kubernetes: install cri-dockerd  https://review.opendev.org/c/zuul/zuul-jobs/+/85109322:14
*** dasm is now known as dasm|off22:17
opendevreviewClark Boylan proposed zuul/zuul-jobs master: DNM check if this fixes artifactory testing  https://review.opendev.org/c/zuul/zuul-jobs/+/85110122:24
opendevreviewIan Wienand proposed zuul/zuul-jobs master: ensure-kubernetes: install cri-dockerd  https://review.opendev.org/c/zuul/zuul-jobs/+/85109322:26
clarkbnb01 and nb02 both appear t ohave full /opts so ansible is unhappy with them. nb03 is not responding to ssh as well ( Iwonder if if those hosts finally got cleaned up?) I'm going to stop the nodepool builders on nb01 and nb02 and clean them up22:26
clarkbI think we have a lot more fedora-35 images in /opt/nodepool_dib on both nb01 and nb02 than we really should have22:33
clarkbOnce I've done the typicaly cleanup I'll look at those images and see if maybe nodepool leaked them22:33
clarkbError loading json data from image build /nodepool/images/fedora-35/builds/0000007464 I'll manually delete that node22:36
clarkbI've removed the problematic znode and resetarted builders after clearing out /opt/dib_tmp22:44
clarkbI believe the builders are chewing through the fedora-35 images in zk and cleaning them up now, but it may take some time22:45
clarkbwas 500 images a couple of minutes ago and now down to 491. Only a small faction of which are taking up disk space22:45
clarkband now it is caught up and happy and disks have room22:51
clarkbnow to look at nb0322:51
clarkbopenstack apis are still there. nb03 and mirror are both in a shutoff state. Going to try startingthem now22:53
clarkband now nb03 is up and running a builder. I think that concludes the "why is our nodepool deploy job failing" debugging22:55
fungilast week i saw deploy jobs failing and nb03 in shutoff was the cause22:58
fungiopenstack server start brought it right up22:58
clarkbyup that is what I did, but also nb01 and nb02 were unhappy because fedora-35 builds weren't pruning and filled the disks22:59
opendevreviewMerged openstack/project-config master: translation updates: refactor Python 3.8 workarounds  https://review.opendev.org/c/openstack/project-config/+/85096223:07
*** rlandy|biab is now known as rlandy23:17
opendevreviewIan Wienand proposed zuul/zuul-jobs master: ensure-kubernetes: install cri-dockerd  https://review.opendev.org/c/zuul/zuul-jobs/+/85109323:23
clarkbservice coordinator election email sent23:26
*** dviroel is now known as dviroel|out23:29

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