| *** rlandy|ruck|bbl is now known as rlandy|ruck | 01:31 | |
| *** ykarel|away is now known as ykarel | 06:12 | |
| *** rpittau|afk is now known as rpittau | 07:49 | |
| *** bhagyashris_ is now known as bhagyashris | 08:05 | |
| zbr | please check https://review.opendev.org/c/opendev/system-config/+/737960 - i will abandon if you think is not needed. | 08:28 |
|---|---|---|
| *** bhagyashris_ is now known as bhagyashris | 09:35 | |
| *** ykarel is now known as ykarel|lunch | 10:25 | |
| *** jcapitao is now known as jcapitao_lunch | 10:40 | |
| *** ykarel|lunch is now known as ykarel | 11:52 | |
| gryf | hello. | 12:03 |
| gryf | I have a job definition in zuul, which inherits from several parents chain, and I'm wondering, why resulting local.conf doesn't contain services, which are defined in base job in vars. | 12:05 |
| gryf | according to the docs (https://zuul-ci.org/docs/zuul/reference/job_def.html#attr-job.vars), it should. and my understanding is that if I have in base job.vars.devstack_localrc.neutron = true, neutron should be enabled in job inherited from the base one. or am I mistaken? | 12:08 |
| opendevreview | Stephen Finucane proposed openstack/openstack-zuul-jobs master: Run lower-constraints to ubuntu-bionic https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/801773 | 12:10 |
| ykarel | gryf, i think for services enable disable, param is devstack_services | 12:12 |
| ykarel | and yes config is merged, so things should be available from parent unless and until overridden in child | 12:13 |
| gryf | ykarel, yeah, I have it under devstack_services (wrote out of my head, sorry), but it doesn't propagate. | 12:15 |
| opendevreview | Stephen Finucane proposed openstack/openstack-zuul-jobs master: Run lower-constraints to ubuntu-bionic https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/801773 | 12:15 |
| *** jcapitao_lunch is now known as jcapitao | 12:17 | |
| ykarel | gryf, it should propogate, likely it's overridden somewhere in chain | 12:18 |
| ykarel | can u share the job link | 12:19 |
| gryf | yup, https://review.opendev.org/c/openstack/kuryr-kubernetes/+/801618/5/.zuul.d/multinode.yaml#16 | 12:19 |
| gryf | it is quite lengthy inheritance: devstack-tempest -> kuryr-kubernetes-tempest-base -> kuryr-kubernetes-tempest-octavia-base -> kuryr-kubernetes-tempest -> kuryr-kubernetes-tempest-multinode | 12:19 |
| ykarel | ohkk multinode job | 12:21 |
| ykarel | looking | 12:21 |
| gryf | and here https://review.opendev.org/c/openstack/kuryr-kubernetes/+/801618/5/.zuul.d/base.yaml#b91 is quite services enabled | 12:21 |
| gryf | yet I've ended up with this: https://e46e23a6b1d1e3078e4b-2e621766bf34635f11548cb9d0caae28.ssl.cf2.rackcdn.com/801618/5/check/kuryr-kubernetes-tempest-multinode/0b87147/controller/logs/local_conf.txt | 12:21 |
| ykarel | you looking for etcd3 right? | 12:22 |
| gryf | not really, there is no nova/neutron/keyston enabled | 12:24 |
| gryf | there are only those services, which was enabled in host/group vars. | 12:24 |
| ykarel_ | gryf, okk so in that case it's not getting merged, setting those vars are overriding | 12:35 |
| gryf | overriding entire vars? that's… unexpected ;P | 12:36 |
| ykarel_ | so if vars: x: {1:2, 3:4}, host-vars: host1: x: 2, for that host x will override | 12:37 |
| ykarel_ | if vars was used in child, then it should have merged from parent | 12:38 |
| *** ykarel_ is now known as ykarel | 12:38 | |
| gryf | I thought, that it works like this - vars {x: 1, y: 1} on base and host-vars {x: 2} on child, vars would be for specific host(s) {x:2, y:1} | 12:39 |
| ykarel | yes ^ true, but in your example you changing x itself | 12:39 |
| ykarel | x == devstack_services | 12:40 |
| gryf | ok, so what will happen, if bas have vars {x: 1, y: 1} and childe have host-vars {z: 100}? | 12:41 |
| gryf | s/bas/base/ | 12:41 |
| ykarel | all should merge | 12:41 |
| ykarel | so for that particualar host x,y,z all available | 12:41 |
| gryf | that was my thinking. | 12:41 |
| gryf | for some reason it's not what I've observed :/ | 12:42 |
| ykarel | so in you case, parent has devstack_services: x,y,z, and host_vars have devstack_services: w,x | 12:43 |
| fungi | maybe ask in #openstack-qa (the fine folks who maintain the devstack and tempest jobs) and see if they have some concrete examples you can look at | 12:43 |
| * gryf double checking all inheritance chain. | 12:43 | |
| gryf | fungi, ack, thank you. | 12:43 |
| ykarel | yes ^ better | 12:43 |
| *** elodilles is now known as elodilles_afk | 12:48 | |
| *** elodilles_afk is now known as elodilles | 14:34 | |
| *** ykarel_ is now known as ykarel | 14:56 | |
| *** ykarel is now known as ykarel|away | 15:08 | |
| *** rlandy is now known as rlandy|ruck | 15:31 | |
| clarkb | zbr: fungi suggested an improvement to that change a whikle back that didn't seem to get any response. Fwiw I don't think its critical, wouldn't urls just work as is in the commit messages? | 15:49 |
| clarkb | for improvements like that I think it really helps to explain the why in addition to the what. We have no context for why something like that might be disireable and that naturally makes it a lesser priority | 15:50 |
| zbr | today I did a lot of cleanup among CR i raised or I was involved with. | 15:51 |
| zbr | if it does not look interesting just ignore. | 15:52 |
| *** rpittau is now known as rpittau|afk | 16:02 | |
| *** sshnaidm is now known as sshnaidm|afk | 20:10 | |
| *** rlandy|ruck is now known as rlandy|ruck|biab | 23:17 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!