| rdogerrit | rdo-trunk proposed rdoinfo master: Bump rdoinfo wallaby tags to latest upper-constraints https://review.rdoproject.org/r/c/rdoinfo/+/43781 | 00:04 |
|---|---|---|
| rdogerrit | rdo-trunk created rdoinfo master: Promote CBS tags update for xena-8s-release https://review.rdoproject.org/r/c/rdoinfo/+/43956 | 00:06 |
| rdogerrit | rdo-trunk created rdoinfo master: Promote CBS tags update for wallaby-8s-release https://review.rdoproject.org/r/c/rdoinfo/+/43957 | 00:06 |
| rdogerrit | rdo-trunk created rdoinfo master: Promote CBS tags update for victoria-8s-release https://review.rdoproject.org/r/c/rdoinfo/+/43958 | 00:07 |
| *** rlandy|bbl is now known as rlandy | 00:17 | |
| *** rlandy is now known as rlandy|out | 00:20 | |
| *** dviroel|out is now known as dviroel | 01:56 | |
| *** dviroel is now known as dviroel|out | 02:18 | |
| *** ysandeep|holiday is now known as ysandeep | 04:14 | |
| rdogerrit | Matthias Runge proposed openstack/gnocchi-distgit wallaby-rdo: Drop python3-pandas from BR https://review.rdoproject.org/r/c/openstack/gnocchi-distgit/+/43945 | 06:28 |
| *** ysandeep is now known as ysandeep|afk | 06:31 | |
| *** dmsimard0 is now known as dmsimard | 06:44 | |
| *** amoralej|off is now known as amoralej | 07:13 | |
| jcapitao[m] | amoralej: hi, wrt ^ | 07:27 |
| jcapitao[m] | DLRN is building sable/4.4 | 07:27 |
| jcapitao[m] | instead of stable/4.3 | 07:28 |
| jcapitao[m] | because yoga tag is set for pinned job | 07:28 |
| jcapitao[m] | https://github.com/rdo-infra/rdo-jobs/blob/79e7d52490e9439d7ffb0368f814a1a717060d2f/zuul.d/jobs.yaml#L161 | 07:28 |
| amoralej | jcapitao[m], ah, in ci | 07:29 |
| amoralej | right? | 07:29 |
| jcapitao[m] | yes | 07:29 |
| jcapitao[m] | pinned job can be apply only if all releases are pinned to same version IIUC, right ? | 07:32 |
| amoralej | i think we thought it mainly for master | 07:32 |
| amoralej | would it be possible to make it work with all releases | 07:32 |
| amoralej | instead to a fix one? | 07:32 |
| jcapitao[m] | you mean to have a pinned job for each release ? | 07:34 |
| amoralej | i mean that the job reads the pin for each release | 07:34 |
| amoralej | and adapt it | 07:34 |
| amoralej | instead of getting it as a parameter | 07:35 |
| rdogerrit | Adriano Vieira Petrich proposed openstack/validations-libs-distgit xena-rdo: Update to version 1.7.0 of the validation libs https://review.rdoproject.org/r/c/openstack/validations-libs-distgit/+/43876 | 07:38 |
| apevec | ready to merge? https://review.rdoproject.org/r/c/openstack/gnocchi-distgit/+/43945 | 07:43 |
| rdogerrit | Adriano Vieira Petrich proposed openstack/validations-libs-distgit victoria-rdo: Update to version 1.6.0 of the validation libs https://review.rdoproject.org/r/c/openstack/validations-libs-distgit/+/43874 | 07:44 |
| jcapitao[m] | amoralej: hmm, it's the default behavior https://github.com/softwarefactory-project/DLRN/blob/master/scripts/run_tests.sh#L105 | 07:46 |
| amoralej | jcapitao[m], so you are afraid that patch is wrong for stable/4.3? | 07:47 |
| jcapitao[m] | yes as amqpd is not available in 4.3 | 07:48 |
| jcapitao[m] | i think it will fail during real build | 07:48 |
| amoralej | jcapitao[m], send a patch to rdo-jobs to adjust to wallaby | 07:49 |
| amoralej | and run with a dnm | 07:49 |
| amoralej | sorry | 07:49 |
| amoralej | not dnm | 07:49 |
| amoralej | with depends-on | 07:49 |
| rdogerrit | Adriano Vieira Petrich proposed openstack/validations-libs-distgit wallaby-rdo: Update to version 1.7.0 of the validation libs https://review.rdoproject.org/r/c/openstack/validations-libs-distgit/+/43875 | 07:49 |
| jcapitao[m] | right, let's test that way | 07:49 |
| jcapitao[m] | apevec: I'll add a -W in the meantime | 07:50 |
| rdogerrit | Joel Capitao created rdo-jobs master: DNM set wallaby tag for pinned jobs https://review.rdoproject.org/r/c/rdo-jobs/+/43959 | 07:53 |
| rdogerrit | Joel Capitao proposed openstack/gnocchi-distgit wallaby-rdo: Drop python3-pandas from BR https://review.rdoproject.org/r/c/openstack/gnocchi-distgit/+/43945 | 07:54 |
| *** ysandeep|afk is now known as ysandeep | 07:59 | |
| jcapitao[m] | amoralej: it's failing, maybe we should add a conditionnal on version as a workaround ? | 08:03 |
| amoralej | on which version? | 08:04 |
| amoralej | ah, the amqp one | 08:05 |
| amoralej | jcapitao[m], or create another job | 08:05 |
| amoralej | pinned-wallaby | 08:05 |
| jcapitao[m] | if version > 4.0 we include the amqpd | 08:06 |
| amoralej | 4.4 | 08:06 |
| jcapitao[m] | yeah but we should also do the same for xena then | 08:06 |
| amoralej | well, given that wallaby is pinned at 4.3 | 08:06 |
| jcapitao[m] | sry 4.4 yes | 08:06 |
| amoralej | it's a bit weird | 08:06 |
| amoralej | but ok, would work | 08:06 |
| jcapitao[m] | yes it'd be better to add job | 08:09 |
| rdogerrit | Joel Capitao proposed rdo-jobs master: DNM set wallaby tag for pinned jobs https://review.rdoproject.org/r/c/rdo-jobs/+/43959 | 08:16 |
| rdogerrit | Adriano Vieira Petrich created openstack/validations-common-distgit xena-rdo: Update to version 1.7.0 https://review.rdoproject.org/r/c/openstack/validations-common-distgit/+/43960 | 08:21 |
| rdogerrit | Jiří Podivín proposed openstack/validations-libs-distgit rpm-master: Callback workaround removal https://review.rdoproject.org/r/c/openstack/validations-libs-distgit/+/42805 | 08:21 |
| rdogerrit | Joel Capitao proposed rdo-jobs master: DNM set wallaby tag for pinned jobs https://review.rdoproject.org/r/c/rdo-jobs/+/43959 | 08:24 |
| rdogerrit | Joel Capitao proposed rdo-jobs master: DNM set wallaby tag for pinned jobs https://review.rdoproject.org/r/c/rdo-jobs/+/43959 | 08:27 |
| rdogerrit | Adriano Vieira Petrich created openstack/validations-common-distgit wallaby-rdo: Update to version 1.7.0 https://review.rdoproject.org/r/c/openstack/validations-common-distgit/+/43961 | 08:32 |
| *** soniya29|ruck is now known as soniya29|ruck|lunch | 08:35 | |
| rdogerrit | Joel Capitao proposed rdo-jobs master: DNM set wallaby tag for pinned jobs https://review.rdoproject.org/r/c/rdo-jobs/+/43959 | 08:36 |
| rdogerrit | Joel Capitao created openstack/gnocchi-distgit xena-rdo: DNM test zuul template https://review.rdoproject.org/r/c/openstack/gnocchi-distgit/+/43962 | 08:40 |
| rdogerrit | Adriano Vieira Petrich proposed openstack/validations-common-distgit xena-rdo: Update to version 1.7.0 https://review.rdoproject.org/r/c/openstack/validations-common-distgit/+/43960 | 08:50 |
| rdogerrit | Adriano Vieira Petrich proposed openstack/validations-common-distgit wallaby-rdo: Update to version 1.7.0 https://review.rdoproject.org/r/c/openstack/validations-common-distgit/+/43961 | 08:51 |
| *** soniya29|ruck|lunch is now known as soniya29|ruck | 08:55 | |
| rdogerrit | Joel Capitao proposed rdo-jobs master: Edit gnocchi project templates https://review.rdoproject.org/r/c/rdo-jobs/+/43959 | 09:09 |
| rdogerrit | Joel Capitao proposed rdo-jobs master: Edit gnocchi project templates https://review.rdoproject.org/r/c/rdo-jobs/+/43959 | 09:12 |
| rdogerrit | Joel Capitao proposed openstack/gnocchi-distgit xena-rdo: Drop python3-pandas from BR https://review.rdoproject.org/r/c/openstack/gnocchi-distgit/+/43944 | 09:22 |
| rdogerrit | Joel Capitao proposed openstack/gnocchi-distgit wallaby-rdo: Drop python3-pandas from BR https://review.rdoproject.org/r/c/openstack/gnocchi-distgit/+/43945 | 09:23 |
| rdogerrit | Merged rdo-infra/ci-config master: Rename centos-9-quay job to ubi-9-quay job https://review.rdoproject.org/r/c/rdo-infra/ci-config/+/43758 | 09:31 |
| *** soniya29|ruck is now known as soniya29|ruck|afk | 10:18 | |
| rdogerrit | Joel Capitao proposed rdo-jobs master: Edit gnocchi project templates https://review.rdoproject.org/r/c/rdo-jobs/+/43959 | 10:19 |
| rdogerrit | Joel Capitao created openstack/gnocchi-distgit rpm-master: DNM test template with master branch https://review.rdoproject.org/r/c/openstack/gnocchi-distgit/+/43964 | 10:21 |
| *** rlandy|out is now known as rlandy | 10:27 | |
| rdogerrit | Joel Capitao proposed rdo-jobs master: Edit gnocchi project templates https://review.rdoproject.org/r/c/rdo-jobs/+/43959 | 10:30 |
| *** soniya29|ruck|afk is now known as soniya29|ruck | 10:40 | |
| rdogerrit | Joel Capitao proposed rdo-jobs master: Edit gnocchi project templates https://review.rdoproject.org/r/c/rdo-jobs/+/43959 | 11:19 |
| rdogerrit | Joel Capitao created openstack/gnocchi-distgit yoga-rdo: DNM test template with yoga branch https://review.rdoproject.org/r/c/openstack/gnocchi-distgit/+/43965 | 11:23 |
| *** dviroel|out is now known as dviroel|rover | 11:29 | |
| rdogerrit | User bshewale proposed rdo-jobs master: Reparent containers and standalone-ipa multinode jobs https://review.rdoproject.org/r/c/rdo-jobs/+/43897 | 11:44 |
| rdogerrit | Douglas Viroel created rdo-infra/ci-config master: Skip scenario010-kvm-internal-standalone-master job https://review.rdoproject.org/r/c/rdo-infra/ci-config/+/43966 | 11:51 |
| *** ysandeep is now known as ysandeep|break | 11:59 | |
| rdogerrit | Matthias Runge created rdoinfo master: Update gnocchi to the supported version https://review.rdoproject.org/r/c/rdoinfo/+/43967 | 12:15 |
| *** amoralej is now known as amoralej|lunch | 12:40 | |
| rdogerrit | Joel Capitao proposed gating_scripts master: exist_in_koji written in python https://review.rdoproject.org/r/c/gating_scripts/+/43902 | 12:44 |
| *** ysandeep|break is now known as ysandeep | 12:46 | |
| rdogerrit | Joel Capitao proposed gating_scripts master: exist_in_koji written in python https://review.rdoproject.org/r/c/gating_scripts/+/43902 | 12:46 |
| rdogerrit | Joel Capitao proposed gating_scripts master: Add option to not update all files https://review.rdoproject.org/r/c/gating_scripts/+/43952 | 12:57 |
| rdogerrit | Joel Capitao created deps/GitPython c9s-zed-rdo: Update to 3.1.27 https://review.rdoproject.org/r/c/deps/GitPython/+/43968 | 12:57 |
| *** dasm|off is now known as dasm | 13:02 | |
| rdogerrit | Merged rdo-infra/ci-config master: Skip scenario010-kvm-internal-standalone-master job https://review.rdoproject.org/r/c/rdo-infra/ci-config/+/43966 | 13:07 |
| *** amoralej|lunch is now known as amoralej | 13:41 | |
| rdogerrit | Joel Capitao created config master: Make the job to be able to run periodically https://review.rdoproject.org/r/c/config/+/43969 | 13:51 |
| rdogerrit | Joel Capitao proposed config master: Get change list from report-uc periodically https://review.rdoproject.org/r/c/config/+/43969 | 13:58 |
| rdogerrit | Joel Capitao proposed config master: Get change list from report-uc periodically https://review.rdoproject.org/r/c/config/+/43969 | 14:06 |
| rdogerrit | Douglas Viroel created rdo-infra/ci-config master: Revert "Skip scenario010-kvm-internal-standalone-master job" https://review.rdoproject.org/r/c/rdo-infra/ci-config/+/43687 | 14:13 |
| rdogerrit | Merged rdo-jobs master: Edit gnocchi project templates https://review.rdoproject.org/r/c/rdo-jobs/+/43959 | 14:16 |
| rdogerrit | Merged openstack/gnocchi-distgit wallaby-rdo: Drop python3-pandas from BR https://review.rdoproject.org/r/c/openstack/gnocchi-distgit/+/43945 | 14:32 |
| rdogerrit | Merged rdo-infra/ci-config master: Revert "Skip scenario010-kvm-internal-standalone-master job" https://review.rdoproject.org/r/c/rdo-infra/ci-config/+/43687 | 14:47 |
| jm1 | ykarel: hello :) you said we could simple use master versions of rpms in tripleo jobs by using different tags, e.g. zed instead of zed-uc. how do we use different tags? https://review.opendev.org/c/openstack/tripleo-quickstart-extras/+/847419/comments/54ebf109_dba689e5 | 14:55 |
| jm1 | rcastillo: fyi ^ | 14:56 |
| jm1 | oh, ykarel is off today. maybe amoralej has a hint how to do that? ^ | 14:57 |
| *** soniya29 is now known as soniya29|out | 14:58 | |
| amoralej | jm1, i'm lacking some context but | 14:59 |
| amoralej | i guess he refers to using unpinned projects? | 14:59 |
| amoralej | i mean build from master branch, instead of pinned tags? | 14:59 |
| jm1 | amoralej: yes, we want to run test without pinning openstacksdk and ansible-collections-openstack | 15:00 |
| amoralej | that's the difference between zed and zed-uc | 15:00 |
| amoralej | so, zed-uc is zed with the pins applied according to upper-constraints versions | 15:00 |
| amoralej | while zed is unpinned | 15:00 |
| jm1 | amoralej: how can we use zed instead of zed-uc? | 15:00 |
| amoralej | in projects.ini | 15:02 |
| amoralej | i'm not sure where it's set in those jobs | 15:02 |
| amoralej | it's in dlrn config | 15:03 |
| amoralej | done by the job | 15:03 |
| jm1 | amoralej: projects.ini? any example job or repo where i could look that up? | 15:03 |
| amoralej | can you point me to a log? | 15:03 |
| jm1 | amoralej: https://logserver.rdoproject.org/63/43963/2/check/periodic-tripleo-ci-centos-9-standalone-master/de5137b/ | 15:04 |
| amoralej | [build-test-packages : drop in the templated version of projects.ini] | 15:04 |
| amoralej | https://logserver.rdoproject.org/63/43963/2/check/periodic-tripleo-ci-centos-9-standalone-master/de5137b/job-output.txt | 15:04 |
| amoralej | probably there | 15:05 |
| *** ysandeep is now known as ysandeep|out | 15:05 | |
| amoralej | mmm where are dlrn logs? | 15:05 |
| amoralej | actually, that job is not building any package, i'd say | 15:06 |
| rdogerrit | Miguel Garcia created openstack/keystoneclient-distgit xena-rdo: Skip test_cms unit tests from %check https://review.rdoproject.org/r/c/openstack/keystoneclient-distgit/+/43970 | 15:07 |
| jm1 | amoralej: yeah, the patch that ykarel commented on above added a zuul variable which allowed us to say which packages to build from source. testproject which lists two packages to build from git: https://review.rdoproject.org/r/c/testproject/+/38639 | 15:08 |
| amoralej | jm1, so iiuc you want to add a tripleo job to gate changes in openstacksdk or ansible-collections-openstack? | 15:09 |
| rdogerrit | Miguel Garcia created openstack/keystoneclient-distgit wallaby-rdo: Skip test_cms unit tests from %check https://review.rdoproject.org/r/c/openstack/keystoneclient-distgit/+/43971 | 15:09 |
| rcastillo | amoralej: we want periodic jobs that run on both latest openstacksdk and latest ansible-collections-openstack | 15:09 |
| amoralej | jm1, sorry, i need more context to understand it | 15:09 |
| jm1 | amoralej: we want to add a tripleo job to ansible-collections-openstack repo which runs tripleo deployment with latest openstacksdk and given ansible-collections-openstack change, e.g. here: https://review.opendev.org/c/openstack/ansible-collections-openstack/+/844859 | 15:10 |
| amoralej | then, you probably want to unpin it? | 15:10 |
| rcastillo | (rdo pins openstacksdk to an older version in master) | 15:10 |
| amoralej | we pin openstacksdk to the version used in upper-constrains | 15:10 |
| amoralej | as we do for all the libraries | 15:10 |
| amoralej | so, the actual fix for that would be to unpin that in rdo | 15:11 |
| jm1 | amoralej: yes but for testing our code we want to use non-pinned rpms in our tripleo job in ansible-openstack-collection | 15:11 |
| amoralej | so | 15:11 |
| jm1 | amoralej: periodic jobs of tripleo ci should continue to use pinned rpms because else we would block everyone when we merge a breaking change | 15:11 |
| amoralej | so, wrt unping ansible-collections-openstack | 15:12 |
| amoralej | ok | 15:12 |
| amoralej | jm1 but then, if you merge a change in upstream that requires a newer version on openstacksdk | 15:13 |
| amoralej | that will break it | 15:13 |
| amoralej | so | 15:14 |
| amoralej | if you run a tripleo job in the ansible-collections-openstack repo | 15:14 |
| amoralej | that will build a rpm with the in-flight patch | 15:14 |
| amoralej | so, that should be fine in that front | 15:15 |
| amoralej | wrt testing it with openstacksdk from master | 15:15 |
| amoralej | i see it more as a problem that as a solution | 15:15 |
| amoralej | as it'd allow to merge changes that will break periodic | 15:15 |
| jm1 | amoralej: we need matching versions of openstacksdk and ansible-collections-openstack. tripleo ci jobs are pinned to older rpms of both. | 15:16 |
| amoralej | sorry, i don't follow you | 15:17 |
| rcastillo | basically master sdk and master collections aren't completely compatible yet. We want to have a pipeline that'll tell us when we're completely ready. | 15:17 |
| rcastillo | if we unpin sdk and collections right now, something _will_ break | 15:17 |
| amoralej | yes | 15:17 |
| amoralej | that's why u-c was invented | 15:17 |
| jm1 | amoralej, rcastillo: cannot type that fast, gimme some time^^ | 15:17 |
| rcastillo | problem is that breaking sdk is in upstream zed u-c, we pinned it in rdo | 15:18 |
| rcastillo | which leaves us in a weird spot | 15:18 |
| rcastillo | for the collections | 15:18 |
| amoralej | mmm | 15:18 |
| amoralej | what that means "breaking sdk is in upstream zed u-c" ? | 15:19 |
| jm1 | amoralej: in our ansible-collections-openstack repo we want to test our code with a tripleo job. this tripleo job currently installs the pinned openstacksdk which is too old. so we want a unpinned rpm of openstacksdk in our tripleo job in ansible-collections-openstack | 15:19 |
| amoralej | current version is bad? | 15:19 |
| rcastillo | yes, 0.99.0 is in upstream u-c, we pinned rdo to 0.64 or something like that | 15:19 |
| amoralej | jm1, if it's too old, request a new release upstream and push it to u-c | 15:19 |
| amoralej | mmm | 15:19 |
| amoralej | then, there must be some issue in rdo | 15:20 |
| amoralej | lemme check why we are not promoting it | 15:20 |
| jm1 | amoralej: no, we cannot push a new version to u-c because this would break tripleo ci jobs | 15:20 |
| amoralej | what would break? | 15:20 |
| amoralej | ah, we are in version-locked | 15:20 |
| amoralej | but that's wrong | 15:20 |
| amoralej | that may break other projects requiring openstacksdk | 15:21 |
| amoralej | what we need is to unbreack tripleo-ci with 0.99 | 15:21 |
| jm1 | amoralej: if we unpin openstacksdk (or raise its version) then tripleo ci jobs will use the new openstacksdk but the old ansible-collections-openstack rpm. this does not work. | 15:21 |
| amoralej | what's the problem? | 15:21 |
| amoralej | ok, gimme a while to think | 15:21 |
| rcastillo | problem is sdk has some breaking changes and we (collections maintainers) haven't had the time/resources to catch up | 15:22 |
| jm1 | amoralej: we have not yet finished porting ansible-collections-openstack to the new openstacksdk (>=0.99.0) which is why openstacksdk is pinned <0.99.0 in zed-uc | 15:22 |
| amoralej | jm1, current master branch of ansible-collections-openstack is ready for openstacksdk-0.99? | 15:22 |
| amoralej | so | 15:22 |
| jm1 | amoralej: nope | 15:22 |
| amoralej | so, you'd need to adapt it in a backwards-compatible way | 15:23 |
| jm1 | amoralej: ansible-collections-openstack is not ready yet for openstacksdk 0.99.0 | 15:23 |
| rcastillo | we wanted a way to test master for both packages as a way to gauge when we reach compatibility | 15:23 |
| jm1 | amoralej: nope, no manpower to adapt it backwards-compatible | 15:23 |
| amoralej | then, you may need to do it blindly | 15:23 |
| jm1 | amoralej: ansible-collections-openstack <2.0.0 works with openstacksdk <0.99.0 only. ansible-collections-openstack >= 2.0.0 works with openstacksdk >=0.99.0 only. but aoc 2.0.0 has not been released yet, we are not ready yet | 15:24 |
| amoralej | once you have something that is ready, we can unpin both together | 15:25 |
| jm1 | amoralej: yes, but it is harder: we have to patch tripleo as well for aoc 2.0.0 but for that we need this tripleo job in aoc repo | 15:25 |
| jm1 | amoralej: you see, its a mess ^^ | 15:26 |
| amoralej | yep | 15:26 |
| jm1 | amoralej: what we need is a way to build unpinned rpms of ansible-collections-openstack and openstacksdk in a periodic tripleo based job in our aoc .zuul.yaml | 15:27 |
| amoralej | but not periodic | 15:28 |
| amoralej | but gate, right? | 15:28 |
| amoralej | we don't want to unpin periodic? | 15:28 |
| amoralej | that's messing me | 15:28 |
| rcastillo | amoralej: not for now | 15:28 |
| rcastillo | eventually yes | 15:28 |
| amoralej | lol | 15:29 |
| jm1 | amoralej: check is important, once we have that periodic is easy | 15:29 |
| amoralej | i'm afraid you'll need to do some ad-hoc solution | 15:29 |
| amoralej | hardcoding somewhere to build ansible-collections-openstack | 15:30 |
| amoralej | from master | 15:30 |
| rcastillo | https://review.opendev.org/c/openstack/tripleo-quickstart-extras/+/847419 | 15:30 |
| amoralej | but again | 15:30 |
| rcastillo | this was the ad hoc solution | 15:30 |
| amoralej | if we are not unpining it, why do you want periodic? | 15:30 |
| amoralej | to test master a-o-c with master openstacksdk? | 15:30 |
| amoralej | so it's like non-voting periodic? | 15:30 |
| jm1 | amoralej: lets take periodic aside for a moment and focus on check | 15:31 |
| amoralej | non-voting -> not in component promotion | 15:31 |
| amoralej | ok | 15:31 |
| rcastillo | amoralej: yeah exactly | 15:31 |
| rcastillo | jm1: we do want a periodic line I think | 15:31 |
| rcastillo | what tripleo-ci calls a "dependency line" basically | 15:31 |
| rcastillo | non voting, just to let us know what would break if we unpin, so we can fix it before | 15:32 |
| amoralej | so, how long do you think it will take you to get this up to a compatible way with 0.99? | 15:32 |
| amoralej | note we are really in a very risky situation | 15:32 |
| jm1 | rcastillo: yeah but for now i am just trying to figure out how to run a tripleo job without pinning in aoc. the tripleo dependency line thingy is separate, kind of | 15:32 |
| amoralej | anything may need openstack 0.99 now | 15:32 |
| amoralej | and block promotions for other components | 15:32 |
| amoralej | i'd try to focus on getting some minimal viable version of compatibility with 0.99 if that's possible | 15:33 |
| jm1 | amoralej, rcastillo: we are talking about two separate things here... (gimme time to type ;) ) | 15:33 |
| jm1 | amoralej, rcastillo: (a) this tripleo based check job in aoc which we will use to develop aoc 2.0.0 | 15:34 |
| *** marios is now known as marios|out | 15:35 | |
| jm1 | amoralej, rcastillo: (b) this tripleo dependency line thingy which afaiu will be used to port tripleo-ansible etc. to aoc 2.0.0 | 15:35 |
| *** dviroel|rover is now known as dviroel|rover|biab | 15:36 | |
| jm1 | amoralej, rcastillo: wrt (a), this job is located in aoc's .zuul.yaml and runs on check etc. we need a way to build unpinned rpm's of ansible-collection-openstack (based on current patch that is being checked) and openstacksdk (latest code from sdk's master branch). how do we do that? | 15:37 |
| jm1 | amoralej, rcastillo: what do you think? | 15:38 |
| amoralej | i'm not sure about the best way to do it, tbh | 15:39 |
| rcastillo | my two proposals for that was using a separate repo with sdk 0.99.0 and the build_test_packages hack | 15:39 |
| amoralej | i see it dificult by modifying existint build-test-packages | 15:39 |
| amoralej | as it will need a different config for dlrn for that package | 15:39 |
| amoralej | inject a new project which is not in the list of changes, etc... | 15:40 |
| amoralej | but i'm not sure, tbh | 15:40 |
| amoralej | let me think if we may consume the package created from outside | 15:41 |
| amoralej | of openstacksdk | 15:41 |
| amoralej | in the past we used to have a rdo trunk repo following master, but we didn't create it for cs9 as it was never actually used | 15:41 |
| rcastillo | amoralej: we can, I created a repo manually and uploaded it to images.rdoproject.org | 15:42 |
| rcastillo | and included it with quickstart, it's possible | 15:42 |
| rcastillo | https://review.opendev.org/c/openstack/tripleo-quickstart/+/845375 | 15:42 |
| amoralej | well, but that may become obsolete | 15:43 |
| rcastillo | maybe we just set a periodic job that rebuilds just those two with dlrn? | 15:43 |
| amoralej | we may stand up a dlrn builder for only that package | 15:44 |
| amoralej | let me do a quick test | 15:44 |
| jm1 | amoralej, rcastillo: why not build both packages on the fly like rcastillo's patch did? (see depends-on patch for impl details) https://review.rdoproject.org/r/c/testproject/+/38639 | 15:46 |
| amoralej | mmm that may be the easiest | 15:47 |
| amoralej | actually, the one for a-c-o will should not be added | 15:47 |
| amoralej | as it's the tested repo | 15:47 |
| amoralej | only the openstacksdk | 15:47 |
| rcastillo | amoralej: yeah, that was more for the testproject | 15:47 |
| amoralej | yes, i understand | 15:47 |
| amoralej | lemme check the logs | 15:48 |
| amoralej | https://logserver.rdoproject.org/39/38639/22/check/periodic-tripleo-ci-centos-9-standalone-master-force-collection-master/fcf901d/logs/delorean_logs/component/clients/af/86/af86313c83f631d212ffc0a14e4bf00a6d058582_dev/rpmbuild.log.txt.gz | 15:49 |
| amoralej | there we have 0.99 | 15:49 |
| rcastillo | so you think we should go with this approach? | 15:50 |
| amoralej | so that's probably all we need :) | 15:50 |
| amoralej | yes | 15:50 |
| amoralej | will be standalone jobs? | 15:50 |
| amoralej | well, anyway, i assume if it's multinode or ovb, packages will be updated wherever needed | 15:50 |
| rcastillo | for now yes, we might test ovb later. But it should work as long as we build images, right? | 15:51 |
| rcastillo | right | 15:51 |
| amoralej | ah, that depends on https://review.opendev.org/c/openstack/tripleo-quickstart-extras/+/847419 | 15:51 |
| amoralej | so that's where we close the circle :) | 15:52 |
| amoralej | so, artg_extra_change_list is a new variable | 15:52 |
| amoralej | there is no way to "inject" changes in the original change list? | 15:52 |
| rcastillo | not that I can see, it gets overwritten at some point | 15:53 |
| amoralej | rcastillo, jm1 actually, isn't artg_change_list just an empty list initially? | 16:00 |
| amoralej | we may just use that variable | 16:00 |
| amoralej | althoug i'm not sure, tbh, i'm not very familiar with the details of that role | 16:01 |
| jm1 | amoralej: except for the impl details, you agree that rcastillo's approach with build-test-packages is the way to go? no way we can simply switch from zed-uc to zed tag for both rpms in our job? | 16:02 |
| amoralej | i think the approach of https://review.rdoproject.org/r/c/testproject/+/38639/22/.zuul.yaml is better | 16:02 |
| amoralej | and much easier | 16:03 |
| amoralej | to implement with current roles | 16:03 |
| rcastillo | amoralej: I tried doing that, didn't work, the logic in the role is hard to follow but it gets overwritten at some point | 16:03 |
| rcastillo | hence the new var | 16:03 |
| amoralej | ok | 16:03 |
| amoralej | i'd say in "artg_change_list: "{{ artg_change_list_prep }}"" it's overwriting the initial list | 16:05 |
| amoralej | actually, this is like adding a depends-on | 16:06 |
| amoralej | on a review in openstacksdk | 16:06 |
| jm1 | amoralej: ok, so we know its not as simple as switching some tag. then we stick to rcastillo's patch and work on that. cool, thanks! 😎 | 16:08 |
| amoralej | wdym "as switching some tag"? | 16:09 |
| amoralej | https://logserver.rdoproject.org/39/38639/22/check/periodic-tripleo-ci-centos-9-standalone-master-force-collection-master/fcf901d/logs/undercloud/home/zuul/dlrn.log.txt.gz | 16:10 |
| amoralej | that's the dlrn log | 16:10 |
| amoralej | of that patch, imo it's a very clean solution tbh | 16:11 |
| gthiemonge | amoralej: jcapitao[m]: Hi Folks, could we force-merge https://review.rdoproject.org/r/c/openstack/octavia-distgit/+/43788 ? centos8 job failed but it is no longer supported by Octavia | 16:41 |
| amoralej | actually we removed them | 16:41 |
| amoralej | rechecking it should clean it | 16:41 |
| *** dviroel|rover|biab is now known as dviroel|rover | 16:42 | |
| rdogerrit | Merged gating_scripts master: exist_in_koji written in python https://review.rdoproject.org/r/c/gating_scripts/+/43902 | 16:51 |
| *** amoralej is now known as amoralej|off | 16:54 | |
| gthiemonge | amoralej|off: thanks! | 17:02 |
| rdogerrit | Merged openstack/octavia-distgit rpm-master: Update Octavia spec for secure-RBAC https://review.rdoproject.org/r/c/openstack/octavia-distgit/+/43788 | 17:04 |
| rdogerrit | Luigi Toscano created openstack/os-brick-distgit wallaby-rdo: Remove sysfs dependency (OSP only) https://review.rdoproject.org/r/c/openstack/os-brick-distgit/+/43972 | 17:18 |
| rdogerrit | Luigi Toscano created openstack/os-brick-distgit wallaby-rdo: DNM Just a test, will centoss8 jobs work? https://review.rdoproject.org/r/c/openstack/os-brick-distgit/+/43973 | 18:31 |
| rdogerrit | Luigi Toscano created rdoinfo master: Fix ansible-role-tripleo-modify-image yoga tag https://review.rdoproject.org/r/c/rdoinfo/+/43974 | 20:57 |
| *** rlandy is now known as rlandy|bbl | 21:59 | |
| *** dasm is now known as dasm|off | 22:14 | |
| *** rlandy|bbl is now known as rlandy|out | 23:47 | |
| *** ysandeep|out is now known as ysandeep | 23:58 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!