| *** hamzy has joined #oooq | 00:25 | |
| *** rlandy has quit IRC | 00:36 | |
| hubbot1 | FAILING CHECK JOBS on master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades, legacy-tripleo-ci-centos-7-ovb-3ctlr_1comp-featureset001-master, tripleo-ci-centos-7-scenario000-multinode-oooq-container-updates @ https://review.openstack.org/604298, master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades, tripleo-quickstart-extras-gate-master-delorean-full-featureset052, (1 more message) | 00:49 |
|---|---|---|
| *** ykarel has joined #oooq | 02:18 | |
| *** ykarel has quit IRC | 02:32 | |
| hubbot1 | FAILING CHECK JOBS on master: legacy-tripleo-ci-centos-7-ovb-3ctlr_1comp-featureset001-master @ https://review.openstack.org/604298, master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades, tripleo-quickstart-extras-gate-master-delorean-full-featureset052, legacy-tripleo-ci-centos-7-ovb-3ctlr_1comp-featureset001-queens-branch @ https://review.openstack.org/560445, stable/queens: tripleo-ci- (1 more message) | 02:49 |
| *** skramaja has joined #oooq | 02:53 | |
| *** ykarel has joined #oooq | 03:13 | |
| *** sanjayu_ has joined #oooq | 03:48 | |
| *** jaganathan has joined #oooq | 04:00 | |
| hubbot1 | FAILING CHECK JOBS on master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades, tripleo-quickstart-extras-gate-master-delorean-full-featureset052 @ https://review.openstack.org/560445, stable/queens: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/567224, stable/rocky: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ (1 more message) | 04:49 |
| *** udesale has joined #oooq | 05:09 | |
| *** ykarel_ has joined #oooq | 05:11 | |
| *** ykarel has quit IRC | 05:14 | |
| *** ykarel__ has joined #oooq | 05:15 | |
| *** ykarel_ has quit IRC | 05:18 | |
| *** ykarel_ has joined #oooq | 05:20 | |
| *** ykarel__ has quit IRC | 05:23 | |
| *** ykarel__ has joined #oooq | 05:24 | |
| *** ykarel_ has quit IRC | 05:27 | |
| *** ykarel_ has joined #oooq | 05:29 | |
| *** ykarel__ has quit IRC | 05:32 | |
| *** quiquell|off is now known as quiquell | 05:41 | |
| quiquell | Good morning vietnam | 05:45 |
| *** chkumar|off is now known as chandankumar | 05:46 | |
| *** dsneddon has quit IRC | 06:23 | |
| *** verdurin has quit IRC | 06:23 | |
| *** holser_ has joined #oooq | 06:25 | |
| *** anande has joined #oooq | 06:27 | |
| *** jfrancoa has joined #oooq | 06:31 | |
| *** holser_ has quit IRC | 06:33 | |
| *** jtomasek has joined #oooq | 06:44 | |
| *** holser_ has joined #oooq | 06:49 | |
| hubbot1 | FAILING CHECK JOBS on stable/pike: tripleo-ci-centos-7-scenario001-multinode-oooq-container, tripleo-ci-centos-7-scenario004-multinode-oooq-container @ https://review.openstack.org/602248, master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades, tripleo-quickstart-extras-gate-master-delorean-full-featureset052 @ https://review.openstack.org/560445, stable/queens: tripleo-ci- (1 more message) | 06:49 |
| *** holser_ has quit IRC | 06:49 | |
| *** holser_ has joined #oooq | 06:50 | |
| *** quiquell is now known as quiquell|brb | 07:04 | |
| *** dtrainor has quit IRC | 07:09 | |
| *** ssbarnea|bkp has quit IRC | 07:16 | |
| *** gkadam has joined #oooq | 07:23 | |
| *** amoralej|off is now known as amoralej | 07:26 | |
| *** quiquell|brb is now known as quiquell | 07:35 | |
| *** bogdando has joined #oooq | 07:48 | |
| *** tosky has joined #oooq | 07:50 | |
| *** quiquell is now known as quiquell|brb | 07:55 | |
| *** holser_ has quit IRC | 07:55 | |
| *** ratailor has joined #oooq | 07:56 | |
| *** holser_ has joined #oooq | 07:58 | |
| *** quiquell|brb is now known as quiquell | 08:05 | |
| *** anande has quit IRC | 08:17 | |
| *** ykarel__ has joined #oooq | 08:28 | |
| *** ykarel_ has quit IRC | 08:31 | |
| *** dtrainor has joined #oooq | 08:31 | |
| *** anande has joined #oooq | 08:36 | |
| *** ratailor has quit IRC | 08:40 | |
| *** anande has quit IRC | 08:41 | |
| *** ratailor has joined #oooq | 08:41 | |
| *** arxcruz is now known as arxcruz|doctor | 08:49 | |
| hubbot1 | FAILING CHECK JOBS on master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/604298, stable/pike: tripleo-ci-centos-7-scenario001-multinode-oooq-container, tripleo-ci-centos-7-scenario004-multinode-oooq-container @ https://review.openstack.org/602248, master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades, tripleo-quickstart-extras-gate- (1 more message) | 08:49 |
| *** ykarel__ is now known as ykarel | 08:59 | |
| *** tosky has quit IRC | 09:02 | |
| *** tosky has joined #oooq | 09:03 | |
| *** ykarel is now known as ykarel|lunch | 09:04 | |
| marios|rover | quiquell: gonna dig into some other issues wes emailed about then look at thos post | 09:17 |
| quiquell | marios|rover: you are the rover man | 09:18 |
| quiquell | ssbarnea: you there ? | 09:18 |
| rascasoft | marios|rover, hey the containerized undercloud is supported from Rocky on, right? | 09:21 |
| rascasoft | marios|rover, I mean, the correct and expected way to deploy the undercloud from rocky on is by using containerized_undercloud: true | 09:21 |
| rascasoft | ? | 09:21 |
| marios | rascasoft: yes it is default i believe | 09:24 |
| marios|rover | rascasoft: you have to specify if you want to not use heat to deploy undercloud | 09:25 |
| rascasoft | marios, ok, let's talk about this because I'm having strange problems | 09:25 |
| marios|rover | rascasoft: well i don't think i'm your guy for that i dind't really work on that stuff but i can maybe point to soemthing? | 09:25 |
| rascasoft | marios|rover, sure, my RDOPhase2 jobs for Rocky and Master are failing since yesterday at the undercloud installation task | 09:26 |
| rascasoft | marios|rover, exactly on this error: 2018-09-28 07:39:26 | fatal: [undercloud]: FAILED! => {"attempts": 10, "changed": false, "content": "", "msg": "Status code was -1 and not [200]: Request failed: <urlopen error timed out>", "redirected": false, "status": -1, "url": "http://192.168.24.1:8787/v2/"} | 09:27 |
| rascasoft | marios|rover, while doing some verification here and there I saw that these setups are still using openstack undercloud install as installation command | 09:27 |
| rascasoft | marios|rover, I was wondering if that's correct or not, because the reason my setups are failing is that the undercloud interfaces are not available | 09:28 |
| rascasoft | marios|rover, /me doesn't want to bother you, just looking for some help on this | 09:29 |
| marios|rover | rascasoft: so there is a different command for the containeized undercloud i believe but i have to check gimme few no worries | 09:31 |
| marios|rover | https://github.com/openstack/python-tripleoclient/blob/87c2877cce670e208e26ca580731dfa39edd0a71/tripleoclient/v1/tripleo_deploy.py#L80 | 09:32 |
| marios|rover | rascasoft:you are right, it was until recently the same with openstack undercloud install | 09:32 |
| marios|rover | rascasoft: looks like the last commit there was to remove https://github.com/openstack/python-tripleoclient/blob/master/tripleoclient/v1/undercloud.py the --use-heat from here | 09:33 |
| marios|rover | rascasoft: openstack tripleo deploy https://github.com/openstack/python-tripleoclient/blob/87c2877cce670e208e26ca580731dfa39edd0a71/setup.cfg#L51 | 09:33 |
| marios|rover | rascasoft: but hold on i see the ci jobs are still using undercloud install/upgrade at http://logs.openstack.org/75/603275/6/check/tripleo-ci-centos-7-containerized-undercloud-upgrades/6ce472c/logs/undercloud/home/zuul/undercloud_upgrade.sh.txt.gz | 09:35 |
| marios|rover | rascasoft: yeah so i suspect the tripleo install/upgrade is for stand-alone deployment (?) whereas the openstack undercloud install/upgrade is by default now doing containerized | 09:37 |
| marios|rover | rascasoft: so you should be able to use the normal openstack undercloud install | 09:37 |
| *** gfidente has joined #oooq | 09:37 | |
| rascasoft | marios|rover, ok so you're saying that we're implicitly use openstack undercloud deploy even when I see openstack undercloud install | 09:38 |
| rascasoft | ? | 09:38 |
| rascasoft | marios|rover, I am really confused | 09:38 |
| marios|rover | rascasoft: i am saying that openstack undercloud install is by default doing containerized undercluod now | 09:38 |
| marios|rover | rascasoft: this might be relevant to any recent errors though landed last night https://review.openstack.org/#/c/588606/ | 09:39 |
| marios|rover | well yesterday | 09:39 |
| rascasoft | marios|rover, that patch didn't land into the repo I'm using so it should no matter | 09:42 |
| marios|rover | rascasoft: ok then the answer to if you should be using 'openstack undercloud install' is yes | 09:43 |
| marios|rover | rascasoft: regardless of the patch i mean. that just removes the previous non containerized instack-undercloud &option | 09:43 |
| rascasoft | marios|rover, ok, I still need to understand why I am failing | 09:43 |
| marios|rover | rascasoft: or try applying that patch to be sure | 09:44 |
| rascasoft | marios|rover, uhm | 09:44 |
| marios|rover | rascasoft: i mean if you apply that patch, you will be sure that the undercloud deployment is containerized not legacy instack-undercloud | 09:44 |
| marios|rover | rascasoft: but if you are using any recent version of client it should be default | 09:45 |
| marios|rover | been for a while i think | 09:45 |
| rascasoft | marios|rover, yes but the question I have in my mind is : why this was working yesterday? | 09:45 |
| marios|rover | rascasoft: don't know yet | 09:45 |
| marios|rover | rascasoft: are you sure that patch wasn't applied in your env? you could sanity check /usr/share/python2.7/site-packages tripleoclient/v1/undercloud.py and compare it to https://github.com/openstack/python-tripleoclient/blob/master/tripleoclient/v1/undercloud.py | 09:47 |
| marios|rover | rascasoft: just because it also merged yesterday so wondering if it is somehow related | 09:47 |
| *** ykarel|lunch is now known as ykarel | 09:48 | |
| *** jtomasek has quit IRC | 09:56 | |
| quiquell | marios|rover: base jobs are not already define in the Depends-On here https://review.rdoproject.org/r/#/c/16370 ? | 10:20 |
| marios|rover | quiquell: which jobs you mean? I helped with these ones but haven't done anything else there https://review.rdoproject.org/r/#/c/16370/11..12/zuul.d/LEFT_TO_TRANSLATE | 10:22 |
| marios|rover | https://review.rdoproject.org/r/#/c/16370/14/zuul.d/base.yaml ? | 10:22 |
| quiquell | marios|rover: yep, those are already defined in the Depends-On | 10:22 |
| marios|rover | defines 'base' jobs for all the jobs | 10:22 |
| marios|rover | quiquell: oh i see leme check | 10:23 |
| quiquell | marios|rover: Depends-On do the same I think | 10:23 |
| quiquell | marios|rover: I think they are supose to move to the other project and remove from this, but not sure | 10:23 |
| marios|rover | quiquell: but seems to be different repo? | 10:23 |
| quiquell | marios|rover: but the same definitions | 10:23 |
| quiquell | have to be related to trusted vs untrusted | 10:24 |
| marios|rover | quiquell: i guess rlandy will have more context for you about that one then | 10:24 |
| rascasoft | marios|rover, I think I've found the problem | 10:24 |
| marios|rover | rascasoft: nice | 10:24 |
| rascasoft | marios|rover, I'm overriding the undercloud-parameter-defaults.yaml file | 10:25 |
| marios|rover | rascasoft: do you require payment for the explanation? | 10:25 |
| marios|rover | ;) | 10:25 |
| rascasoft | marios|rover, I'm writing it down you jerk :D :D :D | 10:25 |
| marios|rover | rascasoft: oh you mean you were overriding some parameter that clashed? | 10:25 |
| marios|rover | rascasoft: :) | 10:25 |
| rascasoft | marios|rover, yeah but wait a sec | 10:25 |
| marios|rover | rascasoft: but again, why since yesterday, did some default flip or soething? | 10:25 |
| quiquell | rascasoft: manners !!! | 10:25 |
| rascasoft | marios|rover, because it's been since yesterday that I started declaring this: | 10:27 |
| rascasoft | ctlplane_masquerade: true | 10:27 |
| rascasoft | undercloud_custom_env_files: >- | 10:27 |
| rascasoft | undercloud-parameter-defaults.yaml | 10:27 |
| rascasoft | undercloud_network_environment_args: | 10:27 |
| rascasoft | "MasqueradeNetworks": { | 10:27 |
| rascasoft | "172.20.0.0/24": [ | 10:27 |
| rascasoft | "172.20.0.0/24" | 10:27 |
| rascasoft | ] | 10:27 |
| rascasoft | } | 10:27 |
| rascasoft | marios|rover, and it seemed to work initially | 10:27 |
| rascasoft | marios|rover, but just because (I realized it NOW) I was testing this on an already deployed undercloud | 10:27 |
| rascasoft | marios|rover, so interfaces were already set | 10:28 |
| rascasoft | marios|rover, I *think* that if I override undercloud_network_environment_args then I don't get any net conf | 10:28 |
| rascasoft | marios|rover, so I need to a) find another variable for those options I need b) find a way to declare a full undercloud-parameter-defaults.yaml file | 10:29 |
| rascasoft | marios|rover, tell me I am explaining myself correctly please :) | 10:29 |
| rascasoft | quiquell, marios|rover, sorry for that word, I just realized it was too strong | 10:30 |
| marios|rover | rascasoft: np at all it wasn't taken badly thanks quiquell for defending me :) np at all peace love and all that good stuff. now it shouldn't matter if you are setting the variables to the same values i.e. as long as the network config doesn't change it shouldn't make a difference to the deployment | 10:31 |
| rascasoft | marios|rover, so I need those undercloud_network_environment_args added to the defaults, not substitute them: am I doing it right? | 10:32 |
| quiquell | rascasoft: thank you jerk !!! | 10:32 |
| rascasoft | LOL | 10:32 |
| hubbot1 | FAILING CHECK JOBS on master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/604298, stable/pike: tripleo-ci-centos-7-scenario001-multinode-oooq-container, tripleo-ci-centos-7-scenario004-multinode-oooq-container @ https://review.openstack.org/602248, master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/560445, (1 more message) | 10:49 |
| quiquell | marios|rover: Can we +w this again https://review.openstack.org/#/c/594511/ ? | 10:56 |
| *** dtantsur|afk is now known as dtantsur | 10:57 | |
| *** rfolco has quit IRC | 11:07 | |
| *** ssbarnea|bkp has joined #oooq | 11:08 | |
| quiquell | ssbarnea: best here | 11:09 |
| quiquell | ssbarnea: I have added the reviews from f28 to taiga, that's right ? | 11:09 |
| ssbarnea | quiquell: perfect, as you observed, the firt two were added by me. | 11:10 |
| quiquell | ssbarnea: What else is missing there ? I see there is work for libvirt | 11:10 |
| quiquell | ssbarnea: Yep added the missing ones they are from wes, what else is missing or needed to do ? | 11:10 |
| ssbarnea | mainly testing them. because there are lots of them it would make sense to merge them in batches. | 11:11 |
| quiquell | ssbarnea: How do we test this ? vm with fedora ? | 11:12 |
| quiquell | ssbarnea: zuul new job with fedora ? | 11:12 |
| ssbarnea | quiquell: wes was working on a change to add new f28 standalone job to zuul | 11:15 |
| *** jjoyce has quit IRC | 11:15 | |
| quiquell | ssbarnea: ok, going to rename the task to cleanup the review and add the review to the task itself in the reviews section | 11:15 |
| *** ratailor has quit IRC | 11:17 | |
| *** jjoyce has joined #oooq | 11:17 | |
| quiquell | ssbarnea: ok modified them all, taiga is not that bad | 11:20 |
| ssbarnea | give me few minutes to fix something on one review | 11:21 |
| quiquell | ssbarnea: sure | 11:22 |
| *** rfolco has joined #oooq | 11:26 | |
| ssbarnea | quiquell: this shoud be ok to review and +W, https://review.openstack.org/#/c/605643 and https://review.openstack.org/#/c/605642/ - i had to fix one bug with one of them so we will need to wait for them to run. | 11:27 |
| *** udesale has quit IRC | 11:32 | |
| quiquell | marios|rover: did you have a review to remove QUICKSTART_RELEASE ? | 11:37 |
| marios | quiquell: started poking here https://tree.taiga.io/project/tripleo-ci-board/task/52 | 11:38 |
| marios | quiquell: there is v1 review but its broken (it was before i was rover ;) ) | 11:38 |
| quiquell | marios: collecting missing pieces | 11:38 |
| marios | i want to revisit it but if you want to work today go ahead | 11:38 |
| marios | if it is useful | 11:38 |
| quiquell | marios: just adding missing reviews about remomving bash | 11:38 |
| marios | https://review.openstack.org/604768 quiquell this too then | 11:39 |
| marios | https://tree.taiga.io/project/tripleo-ci-board/us/50?no-milestone=1 | 11:39 |
| ssbarnea | quiquell: have a look at https://review.openstack.org/#/c/595374/ -- aparently is not ready. | 11:39 |
| marios | tracked in taiga quiquell https://tree.taiga.io/project/tripleo-ci-board/us/50?no-milestone=1 | 11:39 |
| quiquell | marios: Humm i have this https://review.openstack.org/#/c/589876/ | 11:39 |
| ssbarnea | quiquell: just tell me if I am right, and I can find fix it. | 11:39 |
| quiquell | ssbarnea: sure give me a sec | 11:39 |
| *** agopi|brb is now known as agopi | 11:40 | |
| *** ssbarnea|bkp has quit IRC | 11:40 | |
| quiquell | marios: this is the right user storey ? | 11:40 |
| marios | quiquell: ah cool sorry i missed it | 11:40 |
| marios | quiquell: yeah so we were previously using trello | 11:40 |
| marios | quiquell: there were two cards there | 11:40 |
| quiquell | marios: ack | 11:40 |
| quiquell | marios: so I think we need this https://review.openstack.org/#/c/585649/ needed to remove UPGRADE_RELEASE | 11:41 |
| marios | quiquell: but they were moved as https://tree.taiga.io/project/tripleo-ci-board/us/50?no-milestone=1 and https://tree.taiga.io/project/tripleo-ci-board/us/49?no-milestone=1 and https://tree.taiga.io/project/tripleo-ci-board/us/48?no-milestone=1 | 11:42 |
| *** panda|off is now known as panda | 11:42 | |
| marios | quiquell: maybe easier if you see https://tree.taiga.io/project/tripleo-ci-board/us/48?no-milestone=1 # 48-49-50 | 11:42 |
| marios | quiquell: i will also checkout your review on my next reviews run | 11:42 |
| marios | added | 11:42 |
| quiquell | marios: is very old, and mix two things QUICKSTART_RELEASE and UPGRADE_RELEASE | 11:43 |
| marios | quiquell: ack, well tags was nice and clean so i made it by itself | 11:43 |
| marios | quiquell: and started working on quickstart release | 11:43 |
| marios | once we do that we can remove is_featurest and previous_release stuff | 11:43 |
| marios | from functions | 11:43 |
| quiquell | marios: Agree | 11:44 |
| *** yolanda has joined #oooq | 11:44 | |
| marios | quiquell: btw the other way i thought we should go with quickstart release and previous release stuff is the emit release file | 11:44 |
| marios | main problem not everything is using int | 11:44 |
| marios | it | 11:44 |
| marios | i just saw your change is touching there | 11:45 |
| quiquell | marios: Yep we have to make it run first, but wes -1 until we have legacy working and it's also working at RDO | 11:45 |
| quiquell | marios: make sense though | 11:45 |
| quiquell | marios: Also is not the focus of this sprint | 11:45 |
| marios | where is the ruck/ what is martin nick | 11:45 |
| quiquell | marios: Let's just remove the old workflow, life is going to be easier | 11:45 |
| quiquell | marios: I think he was on PTO today or the like | 11:46 |
| marios | quiquell: yea i remember you trying to explain it to me a long time ago i mean the tags specifically into the emit release i think | 11:46 |
| marios | quiquell: but it looks like it works ok on the jobs too anyway check the review comment there please when you have time https://review.openstack.org/604768 | 11:46 |
| agopi | thanks quiquell rfolco. Updated it https://review.openstack.org/#/c/604355/ lets see how it goes | 11:46 |
| quiquell | agopi: btw... where do we run this thing ? | 11:47 |
| marios | quiquell: heh 'long time ago' like last month but 'openstack time]' | 11:47 |
| quiquell | marios: yep :-) | 11:47 |
| quiquell | panda: Have a question about new RDO base jobs review https://review.rdoproject.org/r/#/c/16370/ | 11:48 |
| quiquell | panda: do you have a minute ? | 11:48 |
| agopi | quiquell, it runs on legacy fs53 and also https://review.rdoproject.org/r/#/c/16310/5 this will trigger a v3 job as well | 11:49 |
| agopi | i triggered both | 11:49 |
| agopi | so lets see | 11:49 |
| panda | quiquell: yes | 11:49 |
| quiquell | agopi: You are going to have the variables only at new workflow jobs | 11:49 |
| quiquell | agopi: It means upstream for now | 11:49 |
| agopi | so the legacy job won't be able to use it you mean quiquell ? | 11:50 |
| quiquell | agopi: Let's see if we can remove the oldworkflow from RDO this sprint | 11:50 |
| quiquell | agopi: nope, they are generated at run-v3.yml, we can change run.yaml, but don't know if it worth it | 11:50 |
| agopi | okay if this https://review.rdoproject.org/r/#/c/16310/5 works tbh i am fine with dumping the legacy job | 11:50 |
| quiquell | agopi: you don't need the legacy job ? | 11:51 |
| agopi | if the v3 job is perfect, legacy job won't be needed right, quiquell ? | 11:51 |
| agopi | quiquell: i've limited understanding of the differences between the two so tbh all I need is a fs53 job that triggers not only on browbeat but also couple of upstream projects :) | 11:52 |
| quiquell | agopi: if upstream is enough for your | 11:52 |
| quiquell | ssbarnea: so we are going to need a whole set of releases files for fedora ? | 11:59 |
| ssbarnea | quiquell: yep, but symlinking does wonders, even for folders. | 12:01 |
| ssbarnea | i am worries more about the number of files with content, not the symlinks. | 12:02 |
| quiquell | ssbarnea: Yep, a lot of stuff to track, are they going to be very different ? | 12:03 |
| ssbarnea | quiquell: wes mentioned me that he wants to keep version in the folder name. like centos-7, centos-8, .... | 12:03 |
| ssbarnea | i think that between 7-8 we will have signifiant number of changes, this was the idea. i am not a bif fan of it. anyway we can merge/consolidate them later if we discover that is possible. | 12:04 |
| quiquell | ssbarnea: ack | 12:04 |
| ssbarnea | i like more the approach of having conditionals inside the playbook for specific cases. | 12:05 |
| ssbarnea | quiquell: i already have an idea about how ugly it can get with this, from https://github.com/redhat-openstack/octario/tree/master/components_config -- you should check there for duplicated config and copy/paste work to update them. | 12:06 |
| quiquell | ssbarnea: Why we install requests and urllib3 from yum/dnf and not from pip in the job (adding comments there) | 12:07 |
| ssbarnea | quiquell: i was wondering the same but it was outside the scope of the change ;) | 12:08 |
| quiquell | ssbarnea: out of the scope ? | 12:08 |
| quiquell | maybe is something with the DLRN repos or the like | 12:08 |
| ssbarnea | based on my downstream experience, i would guess is because the rpm versions were outdated. | 12:08 |
| ssbarnea | quiquell: ahh, correction. installing from rpm is correct, we need to respect distro versions. | 12:09 |
| quiquell | ssbarnea: ok | 12:11 |
| chandankumar | brb after 30 mins | 12:13 |
| ssbarnea | joys of rdo cloud, had two vms there, there are still running but i am unable to connect to them... floatingips messes, aparently. | 12:17 |
| *** rfolco has quit IRC | 12:18 | |
| ssbarnea | quiquell: maybe you can explain me the difference between toci_gate_test.sh.j2 and toci_gate_test.sh and why we need almost the same code in two places. | 12:19 |
| quiquell | ssbarnea: yep, jinja-> new workflow, bash -> old workflow | 12:22 |
| *** weshay is now known as weshay_ruck | 12:23 | |
| ssbarnea | can't we use ansible to generate the old file and run it, to avoid keeping the same file in the repo twice? | 12:23 |
| quiquell | ssbarnea: nope, they are alike and if we are lucky old workflow will disappear this sprint | 12:24 |
| *** rlandy has joined #oooq | 12:24 | |
| ssbarnea | quiquell: ok. a 3 weeks RIP date is a good enough reason not to dig deeper. | 12:24 |
| quiquell | ssbarnea: we run TOCI at three places | 12:25 |
| quiquell | ssbarnea: upstream, reproducer and rdo jobs | 12:25 |
| quiquell | ssbarnea: new workflow -> upsgream | 12:25 |
| quiquell | ssbarnea: old workflow -> reproducer and rdo jobs | 12:26 |
| *** skramaja has quit IRC | 12:26 | |
| quiquell | ssbarnea: Imaging how pain is for people to introduce needed changes at CI (like upgrade guys) | 12:26 |
| quiquell | ssbarnea: is like the final technical debt | 12:26 |
| ssbarnea | quiquell: do you have 5mins? i want to show you something cool on BJ. | 12:28 |
| *** ykarel_ has joined #oooq | 12:28 | |
| quiquell | ssbarnea: sure man, I am waitting for some reviews jobs | 12:30 |
| quiquell | ssbarnea: pass me your bj | 12:30 |
| ssbarnea | https://bluejeans.com/2655417928 | 12:30 |
| *** ykarel has quit IRC | 12:31 | |
| *** ykarel_ is now known as ykarel | 12:31 | |
| panda | rlandy: going for the tox-docs error in 16370 | 12:32 |
| rlandy | panda: that was addressed I think | 12:32 |
| rlandy | https://review.rdoproject.org/r/#/c/16370/ is very outdated | 12:32 |
| rlandy | we need to get rid of nodesets in that pacth | 12:33 |
| rlandy | and I need to update the child rdoinfo jobs | 12:33 |
| rlandy | going to chat with amoralej about those jobs now | 12:33 |
| quiquell | rlandy: Did a comment in the review, don't understand why the base jobs are also define din the Depends-On | 12:33 |
| panda | rlandy: ok, I'll do it, anything else ? it contains all the translations, I prefer to update it | 12:33 |
| tosky | ping for https://review.openstack.org/#/c/590298/ | 12:33 |
| panda | quiquell: I replied | 12:33 |
| tosky | it has been waiting for a while and it should be mostly harmless for everyone but the people involved | 12:34 |
| *** agopi is now known as agopi|brb | 12:34 | |
| quiquell | rlandy, panda: Asked about the tox issue at RDO, they say the Depends-On does not work also | 12:34 |
| rlandy | panda: you can join the conversation on #sf-dfg | 12:34 |
| amoralej | rlandy, right, i owe you some reviews | 12:35 |
| amoralej | right? | 12:35 |
| panda | rlandy: anyway the tox-doc is just because there is the LEFT_TO_TRANSLATE file which is not yaml int zuul.d dir, and that is not handled by zuul-sphinx, I'll move the file esewhere, and anyway we don't need to keep it before merging it | 12:35 |
| rlandy | amoralej: well that's what I wanted to chat about ... posting link | 12:35 |
| quiquell | panda: ack, I understand it, thanks ! | 12:35 |
| rlandy | panda: sure - sounds good | 12:35 |
| rlandy | amoralej: I originally wanted to move the rdoinfo-specific piece to a pre-run file ... as in https://review.rdoproject.org/r/#/c/16412/15/playbooks/tripleo-rdo-base/pre-rdoinfo.yml | 12:36 |
| rlandy | but looking at that file, we need to pick up release | 12:36 |
| rlandy | and panda pointed out a role would be better | 12:37 |
| rlandy | in any case, iiuc, this file just overwrites the config | 12:37 |
| rlandy | with the new parent for rdoinfo jobs, can we not just pass the correct parameters? | 12:37 |
| rlandy | even if we did keep the shell script, moving it to a role and templated bash is probably better | 12:38 |
| rlandy | but I wanted to get your take first | 12:38 |
| *** agopi|brb has quit IRC | 12:39 | |
| rlandy | amoralej: ^^ make sense? lastly, I built https://review.rdoproject.org/r/#/c/16412/15/playbooks/tripleo-rdo-base/pre-rdoinfo.yml by comparing the diff of the rdoinfo jobs run.yaml as opposed to the regular tripleo jobs. possible I missed something | 12:42 |
| *** trown|outtypewww is now known as trown | 12:43 | |
| amoralej | reading it | 12:43 |
| amoralej | rlandy, | 12:48 |
| amoralej | you your plan is to have a pre-run specific for rdoinfo jobs | 12:49 |
| amoralej | and same run playbook? | 12:49 |
| hubbot1 | FAILING CHECK JOBS on master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/604298, stable/pike: tripleo-ci-centos-7-scenario001-multinode-oooq-container, tripleo-ci-centos-7-scenario004-multinode-oooq-container @ https://review.openstack.org/602248, master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/560445, (1 more message) | 12:49 |
| rlandy | amoralej: originally , yes, that was the plan. But looking at what the script does, I am not sure that would work - or is the best appoach | 12:50 |
| amoralej | so | 12:50 |
| rlandy | could this work not be called from repo-set up or the like? | 12:51 |
| amoralej | that playbook is creating the tripleo.yml | 12:51 |
| amoralej | then we need to override the release config file | 12:51 |
| amoralej | for the release we are running with it | 12:51 |
| amoralej | let me look where that's currently done | 12:51 |
| amoralej | the part creating the tripleo.yml is ok | 12:51 |
| amoralej | well | 12:51 |
| amoralej | is missing something | 12:51 |
| amoralej | i'm leaving a comment | 12:52 |
| amoralej | but we still need to move the config file | 12:52 |
| amoralej | the problem is that | 12:52 |
| amoralej | if tripleo-quickstart repo is cloned in the run part | 12:52 |
| amoralej | then we need to override the release config in the same part | 12:52 |
| amoralej | after it's cloned | 12:52 |
| amoralej | am i explaining it clear? | 12:53 |
| weshay_ruck | amoralej, can you briefly spec out what you need enabled/disabled w/ that code? | 12:53 |
| amoralej | weshay_ruck, what i do is to add a new repo | 12:54 |
| rlandy | amoralej: yes - we can do what you whatever suggest. I just didn;t want to blanket move a complex shell script - that might have bee the only option with the legacy code | 12:54 |
| amoralej | a temporary repo which is created in a parent job | 12:54 |
| amoralej | actually, that's the only difference | 12:54 |
| amoralej | with a regular oooq job | 12:54 |
| weshay_ruck | amoralej, can we create a general use role to create a temp repo? | 12:55 |
| amoralej | you mean create it or add it to configuration? | 12:55 |
| weshay_ruck | right | 12:55 |
| amoralej | i mean, create the repo is in other job | 12:55 |
| amoralej | as it has it's own validation | 12:56 |
| amoralej | but | 12:56 |
| amoralej | and that repo is used by many nested job | 12:56 |
| amoralej | so it must be a separated job | 12:56 |
| amoralej | but | 12:56 |
| amoralej | the point is how to add it to the release config files | 12:56 |
| weshay_ruck | amoralej, right.. config should be in the release config | 12:56 |
| amoralej | baseurl of those repos is different for each job | 12:56 |
| amoralej | in fact it's stored in the log server | 12:57 |
| weshay_ruck | amoralej, ok.. sounds like something we could pass via zuul param? | 12:57 |
| amoralej | we infer the url | 12:57 |
| amoralej | let me look for the code | 12:57 |
| amoralej | https://review.rdoproject.org/r/#/c/16412/15/playbooks/tripleo-rdo-base/pre-rdoinfo.yml | 12:58 |
| amoralej | in fact | 12:58 |
| amoralej | it's already a zuul parameter | 12:58 |
| amoralej | logs={{ buildset_artifacts_url }} | 12:58 |
| *** jaosorior has quit IRC | 13:00 | |
| weshay_ruck | amoralej, ok.. so if we have that url? why not just use the release file to create the repo? | 13:01 |
| weshay_ruck | amoralej, we're trying to unpack the code here | 13:01 |
| weshay_ruck | https://github.com/openstack/tripleo-quickstart/blob/master/config/release/trunk/queens.yml#L23 | 13:02 |
| weshay_ruck | amoralej, you guys stopped using release config | 13:02 |
| weshay_ruck | why is that? | 13:02 |
| amoralej | weshay_ruck, the problem is how to disable/enable repos | 13:03 |
| amoralej | depending on the job | 13:03 |
| weshay_ruck | baseurl: {{ buildset_artifacts_url }} | 13:03 |
| amoralej | i'm ok with using regular | 13:03 |
| weshay_ruck | amoralej, zuul config? | 13:03 |
| weshay_ruck | amoralej, create all the repos required | 13:04 |
| weshay_ruck | and enable/disable based on job config | 13:04 |
| amoralej | so | 13:04 |
| amoralej | can we have conditionals in the release config? | 13:04 |
| weshay_ruck | yes sir | 13:04 |
| rlandy | template it | 13:04 |
| amoralej | an example? | 13:04 |
| amoralej | but release configs are not jinja template | 13:05 |
| amoralej | ok, i see an example | 13:05 |
| rlandy | https://github.com/openstack/tripleo-quickstart/blob/master/roles/repo-setup/templates/repo_setup.sh.j2#L56 | 13:05 |
| weshay_ruck | https://github.com/openstack/tripleo-quickstart/blob/master/config/release/tripleo-ci/master.yml#L124 | 13:05 |
| amoralej | https://github.com/openstack/tripleo-quickstart/blob/master/config/release/trunk/queens.yml#L43-L45 | 13:05 |
| amoralej | i need to check how to do it | 13:06 |
| *** rfolco has joined #oooq | 13:06 | |
| amoralej | as i also need to enable or disable cloudsig repos | 13:06 |
| amoralej | testing | 13:06 |
| amoralej | mmmm | 13:06 |
| rlandy | amoralej; weshay_ruck: https://review.rdoproject.org/r/#/c/16412/16/zuul.d/tripleo-rdo-base.yaml ( see last section) - we can use the parent job or override vars in child jobs to get exact values needed | 13:06 |
| rlandy | feel free to edit that review to test | 13:07 |
| weshay_ruck | amoralej, I think an axample if a simple implementation .. is to create all repos required and enable/disable them in repo_cmd_after | 13:07 |
| weshay_ruck | I say that because the use of conditionals in the repos: config itself is difficult | 13:08 |
| amoralej | what i was thinking is in passing some parameters | 13:09 |
| amoralej | and apply conditionals in the release file | 13:09 |
| weshay_ruck | sure | 13:12 |
| weshay_ruck | amoralej, for some context.. I believe this is one of the last steps required to get all the tripleo based jobs off zuul legacy | 13:13 |
| weshay_ruck | amoralej, and we're desperate to get it done | 13:13 |
| weshay_ruck | rlandy, is that a fair statement? | 13:13 |
| amoralej | but, is this blocking y ou? | 13:13 |
| rlandy | weshay_ruck: correct | 13:13 |
| amoralej | sorry, i thought you could migrate all the rest | 13:13 |
| amoralej | and it was not blocker | 13:13 |
| weshay_ruck | I'm not 100% certain | 13:13 |
| weshay_ruck | tbh | 13:13 |
| rlandy | amoralej: we could do the least complicated change | 13:13 |
| rlandy | and just mass move the script | 13:13 |
| *** agopi|brb has joined #oooq | 13:14 | |
| amoralej | yes, let's start with that | 13:14 |
| rlandy | but we thought it was better to take advantage of zuul_vars | 13:14 |
| *** agopi|brb is now known as agopi|afk | 13:14 | |
| amoralej | and i'll check if i can move to the proper release file | 13:14 |
| weshay_ruck | panda, shall we sync? | 13:14 |
| *** chem has quit IRC | 13:15 | |
| rlandy | amoralej: thanks - we can merge other pieces in the mean time - but we would like to the get the rdoinfo jobs included as soon as we can | 13:15 |
| *** chem has joined #oooq | 13:15 | |
| rlandy | I'm leaving the base rdoino job in the review | 13:16 |
| rlandy | we can merge that w/o issue if it's not referenced anywhere | 13:16 |
| *** zul has quit IRC | 13:16 | |
| *** zul has joined #oooq | 13:17 | |
| panda | weshay_ruck: ok | 13:18 |
| rlandy | panda: +1 on https://review.rdoproject.org/r/#/c/16370/ - I'm just going to edit the rdoinfo jobs to reparent them | 13:19 |
| weshay_ruck | panda, ok.. /me jumps in | 13:20 |
| *** amoralej is now known as amoralej|lunch | 13:23 | |
| *** arxcruz|doctor is now known as arxcruz | 13:27 | |
| *** toure|gone is now known as toure | 13:30 | |
| weshay_ruck | arxcruz, chandankumar can you guys please look at the unfified tempest fole epic and user story I created | 13:33 |
| weshay_ruck | in taiga | 13:33 |
| weshay_ruck | panda, ^ | 13:33 |
| weshay_ruck | arxcruz, chandankumar what needs to happen is that we need multiple user stories | 13:33 |
| weshay_ruck | created there | 13:33 |
| weshay_ruck | such that the work you guys can complete in this sprint | 13:33 |
| weshay_ruck | is in one user story | 13:34 |
| weshay_ruck | and the other user stories are left out of this sprint | 13:34 |
| arxcruz | weshay_ruck: ok, i'll check it | 13:36 |
| weshay_ruck | panda, https://review.rdoproject.org/grafana/dashboard/db/tripleo-ci?orgId=1 | 13:39 |
| weshay_ruck | panda, http://dashboard-ci.tripleo.org/d/cEEjGFFmz/cockpit?orgId=1 | 13:39 |
| weshay_ruck | panda, sorry.. this one http://dashboard-ci.tripleo.org/d/poOr-d0mk/ansible-exploration?orgId=1 | 13:39 |
| *** holser_ has quit IRC | 13:42 | |
| weshay_ruck | panda, https://review.openstack.org/#/c/605756/ | 13:46 |
| chandankumar | weshay_ruck: arxcruz panda Since we are moving to integration to unified role, Can we also start moving bugcheck to a seperate repo? | 13:48 |
| chandankumar | it will make the integration easier | 13:48 |
| arxcruz | chandankumar: yes, but i don't think this can be done right now, we have other priorities | 13:48 |
| arxcruz | chandankumar: but i agree to move it | 13:48 |
| weshay_ruck | chandankumar, sure | 13:49 |
| chandankumar | arxcruz: ack | 13:49 |
| rlandy | panda; when you are out of meeting ... can you explain why we have two files ... LEFT_TO_TRANSLATE and TO_TRANSLATE - what would go in the second? | 13:49 |
| *** amoralej|lunch is now known as amoralej | 13:49 | |
| *** agopi|afk is now known as agopi | 13:53 | |
| panda | rlandy: I wanted to keep track of what we already translated without losing sight of what was there to translate | 14:03 |
| rlandy | panda: hmm, so the TO-TRANSLATE file should contain what is already translated? | 14:05 |
| rlandy | the names are not clear to me | 14:05 |
| amoralej | rlandy, in the last PS you removed the pre for rdoinfo | 14:07 |
| amoralej | i understood you would keep it with the script as-is | 14:08 |
| panda | rlandy: these were meant mostly for me to understand what I needed to do. So, TO_TRANSLATE contains the name of all the legacy job that needs translation. LEFT_TO_TRANSLATE was initially a copy of TO_TRANSLATE, but I removed the name once the job was translated. The translated jobs contain a comment above indicating which legacy job they translate. Now LEFT_TO_TRANSLATE is empty because all the jobs were | 14:08 |
| panda | translated. Probably the rdoinf and -tht are not translated correctly | 14:08 |
| rlandy | amoralej: yeah - because I thought we would go with editing the repos as we discussed | 14:08 |
| rlandy | but as an intermediate step, we can out that back | 14:08 |
| amoralej | yeah, that was my understanding | 14:08 |
| amoralej | anyway, in rdoinfo those are non-voting | 14:08 |
| amoralej | and i was thinking in re-doing them anyway | 14:09 |
| rlandy | amoralej: either way, we can include the intermediate step to make sure the jobs work with legacy | 14:10 |
| rlandy | probably using a role is better | 14:10 |
| rlandy | than what I first did | 14:10 |
| amoralej | why is better? | 14:10 |
| amoralej | i'm not saying it's not | 14:11 |
| rlandy | we need to refer to zuul vars | 14:11 |
| amoralej | just trying to understand | 14:11 |
| panda | amoralej: because it becomes more testable | 14:11 |
| amoralej | ok | 14:11 |
| rlandy | I changed release in https://review.rdoproject.org/r/#/c/16412/15/playbooks/tripleo-rdo-base/pre-rdoinfo.yml | 14:11 |
| panda | amoralej: to test a playbook in rdo-infra/config we have to merge it each time | 14:11 |
| rlandy | but there is still ZUUL_REF | 14:11 |
| amoralej | ah, ok | 14:11 |
| rlandy | and ZUUL_PIPELINE used | 14:12 |
| panda | amoralej: if we make a role in rdo-jobs instead, it can be imported and reused in rdo-infra/config and we can eventually test it without merging it | 14:12 |
| rlandy | the question is - where (at what point) exactly do we call this role | 14:13 |
| rlandy | panda: ^^ | 14:13 |
| rlandy | it needs to happen after tripleo-quickstart is cloned, iiuc from above | 14:13 |
| rlandy | we could call it from run-v3 | 14:14 |
| rlandy | or in the parent job directly | 14:15 |
| *** chem has quit IRC | 14:16 | |
| panda | rlandy: in the parent job would be better, a playbook that calls just this role, otherwise we have to create a conditional to call the role or not in run-v3 itself | 14:16 |
| *** chem has joined #oooq | 14:16 | |
| rlandy | panda: right ... considering above comment ... <amoralej> if tripleo-quickstart repo is cloned in the run part | 14:17 |
| rlandy | tripleo-quickstart should be cloned before pre, right? | 14:17 |
| rlandy | as in http://logs.openstack.org/23/605523/1/check/tripleo-ci-centos-7-containers-multinode/7abdddc/job-output.txt.gz#_2018-09-28_00_19_32_613114 | 14:18 |
| rlandy | if it has to be in the same part as run-v3, a conditional is not a big deal | 14:19 |
| rlandy | as it can be defined in the parent job | 14:19 |
| arxcruz | panda: do you have time to spread the word of our lord taiga ? | 14:19 |
| panda | always be praised | 14:20 |
| *** bogdando has quit IRC | 14:20 | |
| panda | rlandy: if tripleo-quickstart is in required-projects it should be cloned by zuul even before starting any pre playbook | 14:20 |
| *** holser_ has joined #oooq | 14:21 | |
| rlandy | amoralej: ^^ considering above, would a pre playbook to call the rdoinfo script be acceptable? | 14:21 |
| panda | arxcruz: always be praised | 14:21 |
| * rlandy puts back role and playbook | 14:22 | |
| *** vkapalav has joined #oooq | 14:27 | |
| amoralej | rlandy, i'd say so | 14:29 |
| amoralej | looks fine to use pre-run | 14:29 |
| rlandy | ok - working on that role now | 14:29 |
| *** agopi is now known as agopi|afk | 14:29 | |
| *** quiquell is now known as quiquell|off | 14:31 | |
| weshay_ruck | panda, rfolco based on the conversation I just had w/ rfolco I think I can say the following | 14:32 |
| weshay_ruck | expect to see in the next sprint an epic around building a f28 pipeline to build images and containers.. I expect that pipeline to fail | 14:32 |
| weshay_ruck | limited amount of jobs, resources however we need to get the job going to build containers | 14:33 |
| weshay_ruck | panda, as rfolco frees up over the next couple weeks, I've asked that he starts a rhel7 / master pipeline in the internal sf | 14:33 |
| weshay_ruck | again w/ limited jobs / resources | 14:34 |
| weshay_ruck | this could prove very handy once we have rhel 8 in hand | 14:34 |
| weshay_ruck | and could also prove to be a thing we could move to the internal sf in RHOS to permantly build and publish rdo on rhel containers | 14:34 |
| *** gfidente is now known as gfidenteN00b | 14:46 | |
| amoralej | rlandy, just note the PHASE variable must be a parameter too | 14:48 |
| rlandy | amoralej: yep - got that in my current working version | 14:48 |
| rlandy | will define phase var in the zuul job | 14:49 |
| rlandy | per job | 14:49 |
| rlandy | also ... | 14:49 |
| hubbot1 | FAILING CHECK JOBS on master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/604298, stable/pike: tripleo-ci-centos-7-scenario001-multinode-oooq-container, tripleo-ci-centos-7-scenario004-multinode-oooq-container @ https://review.openstack.org/602248, master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/560445, (1 more message) | 14:49 |
| rlandy | the other $ZUUL_* vars are to set out old logs | 14:49 |
| rlandy | we are overriding those with v3 jobs so can probably skip ... https://github.com/rdo-infra/rdo-jobs/blob/master/playbooks/legacy/rdoinfo-tripleo-ocata-release-centos-7-multinode-1ctlr-featureset005/run.yaml#L122 | 14:50 |
| *** chem has quit IRC | 15:16 | |
| *** chem has joined #oooq | 15:17 | |
| rlandy | panda: amoralej: pls review https://review.rdoproject.org/r/#/c/16370/ ... | 15:20 |
| rlandy | playbooks/legacy/pre-rdoinfo.yaml | 15:20 |
| rlandy | roles/tripleo-ci-rdoinfo/tasks/main.yaml | 15:20 |
| rlandy | zuul.d/base.yaml | 15:20 |
| rlandy | if you agree, I will edit the child jobs to overwrite phase where needed | 15:21 |
| *** dtrainor has quit IRC | 15:21 | |
| *** dtrainor has joined #oooq | 15:21 | |
| rlandy | and will edit https://review.rdoproject.org/r/#/c/16412/15/zuul.d/tripleo-rdo-base.yaml to match | 15:21 |
| *** ykarel is now known as ykarel|away | 15:35 | |
| panda | rlandy: added on comment, but wanted to understand before saving it. The base of this job will in in rdo-infra/config in the end ? | 15:36 |
| rlandy | panda: ack - making that edit now ... just wanted to get a review on the first one | 15:37 |
| *** chem has quit IRC | 15:37 | |
| panda | rlandy: ok, so the playbooks here is only for testing ? | 15:37 |
| panda | playbooks/legacy/pre-rdoinfo.yaml | 15:37 |
| rlandy | panda: ack same as legacy/pre.yaml | 15:38 |
| rlandy | also duplicated | 15:38 |
| rlandy | panda:sec - let me make the edit to config and you can review | 15:38 |
| panda | rlandy: ok, saved the comment, it's not related to config | 15:39 |
| *** chem has joined #oooq | 15:40 | |
| *** zul has quit IRC | 15:41 | |
| rlandy | panda: when I do the include_role from config, | 15:49 |
| rlandy | what do I need to add for the role to be picked up | 15:50 |
| rlandy | in rdo-jobs repo | 15:50 |
| panda | rlandy: nothing, the role should be picked up. in rdo-infra/config job configuration instead you have to specify roles: - zuul: rdo-jobs | 15:50 |
| panda | roles in this case is zuul configuration, not ansible module | 15:51 |
| *** holser_ has quit IRC | 15:52 | |
| *** jfrancoa has quit IRC | 15:55 | |
| rlandy | panda: pls review https://review.rdoproject.org/r/#/c/16412/ | 15:55 |
| rlandy | amoralej: making an edit for your comment ... where was that copy done in the legacy jobs? | 15:57 |
| weshay_ruck | panda, you still with me? | 16:03 |
| rlandy | ie: I am not sure which dir it picks up from | 16:06 |
| rlandy | oh I see ... nvm | 16:07 |
| *** sanjayu_ has quit IRC | 16:15 | |
| rlandy | panda: amoralej: https://review.rdoproject.org/r/#/c/16370/19/roles/tripleo-ci-rdoinfo/tasks/main.yaml is updated - pls review. I think this should move to its own commit though - otherwise we will have a problem with the merge order of the reviews | 16:23 |
| amoralej | rlandy, i'm not sure if sudo is needed | 16:26 |
| amoralej | other than that | 16:27 |
| amoralej | looks fine | 16:27 |
| *** zul has joined #oooq | 16:27 | |
| rlandy | amoralej: ok - thanks - we'll test it out | 16:28 |
| *** ykarel_ has joined #oooq | 16:29 | |
| *** ykarel|away has quit IRC | 16:30 | |
| *** ykarel__ has joined #oooq | 16:31 | |
| *** ykarel_ has quit IRC | 16:31 | |
| *** gkadam has quit IRC | 16:42 | |
| rlandy | test job in queue | 16:44 |
| *** dtantsur is now known as dtantsur|afk | 16:45 | |
| weshay_ruck | rlandy++ | 16:45 |
| hubbot1 | weshay_ruck: rlandy's karma is now 23 | 16:45 |
| rlandy | let's see what it does first | 16:46 |
| hubbot1 | FAILING CHECK JOBS on master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/604298, stable/pike: tripleo-ci-centos-7-scenario001-multinode-oooq-container, tripleo-ci-centos-7-scenario004-multinode-oooq-container @ https://review.openstack.org/602248, master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/560445, (1 more message) | 16:49 |
| rlandy | well, that didn;t exactly work ... | 16:51 |
| rlandy | 2018-09-28 16:48:16.307926 | PLAY [primary] | 16:51 |
| rlandy | 2018-09-28 16:48:16.345123 | | 16:51 |
| rlandy | 2018-09-28 16:48:16.345293 | TASK [Include rdoinfo role] | 16:51 |
| rlandy | 2018-09-28 16:48:16.382948 | primary | ok | 16:51 |
| rlandy | failed in the role and didn;t report out | 16:51 |
| rlandy | 'buildset_artifacts_url' is undefined | 16:53 |
| panda | mmhhh. | 16:53 |
| rlandy | ^^ where is that defined?? | 16:54 |
| panda | I asked in sf-dfg once. but the only one that can answer is paul apparently, if there's a problem including rdo-jobs in the configuration for rdo-infra/config in any way. Because it seems they are clearly avoiding this depdendncy | 16:55 |
| rlandy | https://github.com/rdo-infra/rdo-jobs/blob/master/roles/set-artifact-path-for-children/tasks/main.yaml | 16:57 |
| rlandy | ^^ I think we need to call that | 16:58 |
| rlandy | and rdoinfo job probably only run after dlrn build of sorts | 16:58 |
| *** ykarel__ has quit IRC | 17:00 | |
| weshay_ruck | ssbarnea, https://review.openstack.org/#/c/605642/ | 17:02 |
| weshay_ruck | ssbarnea, you on that? | 17:02 |
| *** jaganathan has quit IRC | 17:06 | |
| rlandy | panda: looks like we need a dlrn change to get buildset_artifacts_url | 17:06 |
| rlandy | amoralej: ^^ can we fake that out by defining it? | 17:06 |
| rlandy | for our test job? | 17:07 |
| rlandy | https://review.rdoproject.org/r/#/c/16567/ | 17:07 |
| amoralej | buildset_artifacts_url is created | 17:07 |
| amoralej | by something in zuul | 17:07 |
| amoralej | rlandy, let me try a different test review | 17:08 |
| amoralej | i think we need to use rdoinfo | 17:08 |
| *** gfidenteN00b has quit IRC | 17:10 | |
| *** trown is now known as trown|lunch | 17:11 | |
| amoralej | rlandy, i just realized of other issue | 17:15 |
| amoralej | let me check | 17:15 |
| amoralej | rlandy, https://review.rdoproject.org/r/#/c/16418/ | 17:15 |
| amoralej | that's test review | 17:15 |
| rlandy | amoralej: thanks - will watch that | 17:17 |
| amoralej | one of the things we do in the script is to create not_required file | 17:18 |
| amoralej | for jobs not needed | 17:18 |
| amoralej | and the task that actually runs oooq checks if it exist | 17:18 |
| amoralej | https://github.com/rdo-infra/rdo-jobs/blob/master/playbooks/legacy/rdoinfo-tripleo-ocata-testing-centos-7-multinode-1ctlr-featureset005/run.yaml#L306 | 17:18 |
| rlandy | hmmm ... I think I left that out | 17:18 |
| rlandy | because I didn;t create the logs dir | 17:19 |
| amoralej | the point is | 17:19 |
| amoralej | a change in rdoinfo can affect diferent releases | 17:20 |
| amoralej | but we don't want to run jobs for all of them in every review | 17:20 |
| amoralej | so, the initial script has some logic to dectect when a job is needed | 17:20 |
| rlandy | the not_required file is created - I left out the check | 17:21 |
| rlandy | adding that | 17:22 |
| *** jtomasek has joined #oooq | 17:24 | |
| ssbarnea | wesweshay_ruck (IRC) : yep, but i seen a weird error where the build failed blaming a missing file that was supposed to be there (symlink). i am away while moving hw into garage. | 17:26 |
| rlandy | roles/tripleo-ci-rdoinfo/tasks/main.yaml is updated | 17:27 |
| rlandy | amoralej: ^^ can I recheck the test job? | 17:28 |
| amoralej | yeah | 17:28 |
| amoralej | but it seems it's not using the .zuul.yaml | 17:28 |
| rlandy | hmm ... | 17:30 |
| amoralej | yeah | 17:30 |
| amoralej | did i do something wront? | 17:30 |
| amoralej | wrong? | 17:30 |
| amoralej | or do repos need something special? | 17:30 |
| amoralej | it has to be in testproject? | 17:33 |
| rlandy | testproject is setup to pickup .zuul.yaml | 17:33 |
| rlandy | I am not sure how else we can test it | 17:34 |
| rlandy | I must just recheck that I have not left out anything else required | 17:34 |
| * rlandy runs diff | 17:34 | |
| amoralej | ok | 17:34 |
| amoralej | let me try in your review | 17:35 |
| amoralej | rlandy, i also realized of a problems in job definition with branches | 17:39 |
| amoralej | i just left a comment | 17:39 |
| ssbarnea | weshay_ruck: have a look at http://logs.openstack.org/42/605642/3/check/tripleo-ci-centos-7-undercloud-containers/3f97ae6/job-output.txt.gz#_2018-09-28_12_18_42_736511 -- and if you check the change this is a symlink. I checked and ansible has no problem loading yaml files from symlinks. I suspect that the CR was not takes with its sister one by the job. this could explain the failure. | 17:39 |
| rlandy | looking | 17:39 |
| rlandy | fixing | 17:40 |
| rlandy | amoralej: https://review.rdoproject.org/r/#/c/16370/21/zuul.d/multinode-jobs.yaml is updated to fix the branches ... rdoinfo is added to the parent job ... see additional required-projects .. https://review.rdoproject.org/r/#/c/16370/21/zuul.d/base.yaml | 17:46 |
| *** dsneddon has joined #oooq | 17:47 | |
| amoralej | thanks rlandy | 17:48 |
| *** dsneddon has joined #oooq | 17:49 | |
| *** agopi|afk is now known as agopi | 17:51 | |
| *** chem has quit IRC | 18:18 | |
| *** chem has joined #oooq | 18:18 | |
| amoralej | rlandy, https://logs.rdoproject.org/67/16567/2/check/rdoinfo-tripleo-ocata-testing-centos-7-multinode-1ctlr-featureset005/22ea4a5/job-output.txt.gz | 18:32 |
| amoralej | it seems it's not actually running the role | 18:32 |
| amoralej | in pre-run | 18:33 |
| rlandy | it's failing out | 18:33 |
| rlandy | and not reporting | 18:33 |
| rlandy | it's a bug we've seen before | 18:33 |
| rlandy | was talking with nhicher about that | 18:33 |
| rlandy | it's probably hitting undefined variable | 18:35 |
| rlandy | worst case, I can pull it out of the role and include it directly in the playbook until we see it pass | 18:36 |
| rlandy | 2018-09-28 18:05:19.316112 | TASK [Include rdoinfo role] | 18:36 |
| rlandy | 2018-09-28 18:05:19.355498 | primary | ok | 18:36 |
| rlandy | ^^ bug in zuul | 18:36 |
| *** trown|lunch is now known as trown | 18:37 | |
| rlandy | I'm waiting to see if legacy picks up the var | 18:45 |
| rlandy | if it does, I'll move the role to a playbook | 18:45 |
| rlandy | panda: ^^ do you remember what we did previously when the role failure was not picked up by zuul | 18:46 |
| rlandy | thinking of reverting to putting the code directly in a playbook until we see it pass | 18:46 |
| hubbot1 | FAILING CHECK JOBS on master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/604298, stable/pike: tripleo-ci-centos-7-scenario001-multinode-oooq-container, tripleo-ci-centos-7-scenario004-multinode-oooq-container @ https://review.openstack.org/602248, master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/560445, (1 more message) | 18:49 |
| panda | rlandy: I think it was a bug on zuul, the role is not imported properly, I think because the repo is not included in the general configuration | 19:06 |
| rlandy | panda: 'general configuration'? | 19:12 |
| rlandy | I remember the zuul bug but not how we fixed it :( | 19:13 |
| *** chem has quit IRC | 19:19 | |
| *** chem has joined #oooq | 19:19 | |
| panda | rlandy: we didn't, I think we had to include the repo in another part of the configuration. Let me look at the logs | 19:21 |
| panda | rlandy: where's the test patch | 19:22 |
| rlandy | https://review.rdoproject.org/r/#/c/16567/ | 19:23 |
| rlandy | and https://review.rdoproject.org/r/#/c/16418/ | 19:23 |
| rlandy | same as before ... | 19:24 |
| rlandy | https://logs.rdoproject.org/67/16567/2/check/rdoinfo-tripleo-ocata-testing-centos-7-multinode-1ctlr-featureset005/22ea4a5/job-output.txt.gz | 19:24 |
| rlandy | ^^ panda | 19:24 |
| rlandy | the role passes with ok but is never run | 19:24 |
| panda | rlandy: you can't depend-on on the patch in rdo-infra/config, it will not be take into consideration | 19:27 |
| panda | rlandy: bah nevermind, it the patch in rdo-job | 19:27 |
| panda | rlandy: that's legit | 19:28 |
| rlandy | panda: if you mean ... https://review.rdoproject.org/r/#/c/16370/ - I never added that | 19:28 |
| panda | rlandy: added a comment on https://review.rdoproject.org/api/16370 | 19:32 |
| rlandy | looking | 19:33 |
| rlandy | panda: didn;t help :( | 19:51 |
| rlandy | same ok pass with no run | 19:51 |
| * rlandy is ready to put the code back in the playbook | 19:52 | |
| rlandy | at least to see it run | 19:52 |
| weshay_ruck | FYI.. folks look for the new gui option in SF | 20:02 |
| weshay_ruck | rdo -sf | 20:02 |
| weshay_ruck | panda, I'm in mumble I think | 20:08 |
| panda | weshay_ruck: I see you | 20:12 |
| panda | rlandy: let me try one more thing first | 20:17 |
| rlandy | panda: I tried adding a fuller path | 20:18 |
| rlandy | rdo-infra/rdo-jobs | 20:18 |
| rlandy | ugh - no show :( | 20:22 |
| rlandy | panda: any success? otherwise this can move back to playbook | 20:23 |
| panda | rlandy: give me two minutes | 20:25 |
| rlandy | panda: you can have 20 - need to run out for a few | 20:26 |
| *** rlandy is now known as rlandy|brb | 20:27 | |
| *** agopi is now known as agopi|brb | 20:37 | |
| *** agopi|brb has quit IRC | 20:41 | |
| weshay_ruck | anyone have a sec to bounce an idea around for https://bugs.launchpad.net/tripleo/+bug/1795043 | 20:42 |
| openstack | Launchpad bug 1795043 in tripleo "designate's named is conflicting with unbound in CI scenario003" [High,In progress] - Assigned to Ben Nemec (bnemec) | 20:42 |
| hubbot1 | FAILING CHECK JOBS on master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/604298, stable/pike: tripleo-ci-centos-7-scenario001-multinode-oooq-container, tripleo-ci-centos-7-scenario004-multinode-oooq-container @ https://review.openstack.org/602248, master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/560445, (2 more messages) | 20:49 |
| *** rlandy|brb is now known as rlandy | 20:49 | |
| rlandy | panda: anything new? | 20:54 |
| rlandy | I tried just rdo-jobs btw | 20:55 |
| *** panda has quit IRC | 20:57 | |
| *** panda has joined #oooq | 20:58 | |
| *** rfolco has quit IRC | 21:18 | |
| *** dtrainor_ has joined #oooq | 21:22 | |
| *** dsneddon has quit IRC | 21:22 | |
| *** dsneddon has joined #oooq | 21:23 | |
| panda | rlandy: got the role included | 21:23 |
| panda | rlandy: but it's not looking good anyway | 21:24 |
| panda | rlandy: the role is failing an there is no log to debug | 21:24 |
| rlandy | panda: well that's a start | 21:24 |
| rlandy | what did you change? | 21:24 |
| *** dtrainor has quit IRC | 21:24 | |
| rlandy | where is the failure? | 21:25 |
| panda | rlandy: I'm discussing with nhicher in sf-dfg, but apparently my first suggestion of using include_role wasn't a good idea. include_role is not supported in this case. | 21:25 |
| rlandy | 2018-09-28 21:25:07.174432 | PLAY [primary] | 21:25 |
| rlandy | 2018-09-28 21:25:07.206027 | | 21:25 |
| rlandy | 2018-09-28 21:25:07.206192 | TASK [tripleo-ci-rdoinfo : Create temporary repos] | 21:25 |
| rlandy | 2018-09-28 21:25:07.729426 | primary | MODULE FAILURE | 21:25 |
| rlandy | 2018-09-28 21:25:07.755077 | | 21:25 |
| panda | yes | 21:26 |
| rlandy | ok - I'll reverse the include_role in the other review | 21:28 |
| *** agopi|brb has joined #oooq | 21:28 | |
| panda | I'm not sure if using the playbook directly instead will show more errors | 21:30 |
| rlandy | panda: I can try | 21:32 |
| rlandy | really this is meant to be a easy transition step | 21:32 |
| rlandy | but I am learning that there is no such thing | 21:33 |
| rlandy | perhaps the full translation is easier | 21:33 |
| rlandy | let me try the other test job | 21:33 |
| rlandy | https://review.rdoproject.org/r/#/c/16418/ | 21:34 |
| rlandy | let's see what that does | 21:34 |
| rlandy | I think the playbook may give us a more exact error | 21:34 |
| panda | rlandy: I see some other problems, I see two rdoinfo jobs with the same name | 21:34 |
| rlandy | step of failure | 21:34 |
| rlandy | possible - where? | 21:34 |
| panda | rlandy: all the rdoinfo jobs in multinode | 21:35 |
| panda | rlandy: with one difference: one version has a phase: variable | 21:35 |
| panda | the other version has none | 21:35 |
| rlandy | correct | 21:35 |
| rlandy | the other is release | 21:35 |
| rlandy | taken from the parent job | 21:35 |
| panda | ah | 21:35 |
| panda | one is testing | 21:35 |
| panda | the other is release | 21:35 |
| rlandy | that is right | 21:35 |
| panda | ok, just tired eyes then | 21:36 |
| rlandy | I removed the branches per amoralej comment | 21:36 |
| rlandy | they shoudl all be master | 21:36 |
| rlandy | panda; it's late for you | 21:36 |
| rlandy | thanks for your help | 21:36 |
| panda | another thing that I want to try to see if logging improves is move the playbook to run from pre-run | 21:36 |
| rlandy | I'll try the playbook if I get nothing from the current testjob run | 21:36 |
| rlandy | probably on sunday | 21:36 |
| rlandy | the logging with role i terrible | 21:37 |
| rlandy | that is a failure of zuul | 21:37 |
| rlandy | if you want to try the playbook, go ahead but I'd like to see the current test job run | 21:37 |
| panda | rlandy: ping me when you have the results | 21:39 |
| rlandy | k - job is queued atm | 21:39 |
| rlandy | should start shortly | 21:39 |
| *** vkapalav has quit IRC | 21:42 | |
| panda | rlandy: are we sure we can run arbitrary bash code in the ansible executor directly ? | 21:45 |
| rlandy | panda:same deal | 21:46 |
| rlandy | panda: you can go ahead and try the playbook | 21:46 |
| panda | rlandy: I mean, the role is trying to run a shell module on the executor itself | 21:46 |
| rlandy | really idk - I would love to do the real translation | 21:47 |
| panda | I know that te executor is pretty much locked down | 21:47 |
| panda | rlandy: what's the real translation for you ? | 21:47 |
| rlandy | maybe that is the issue | 21:47 |
| rlandy | this role adds a new repo file | 21:47 |
| rlandy | this could be done in repo-setup | 21:47 |
| rlandy | with the proper vars | 21:47 |
| rlandy | but requires work - that amoralej would do | 21:48 |
| panda | no wait, the playbook says host: primary | 21:48 |
| rlandy | if this is becoming too much hassle, we should go for the real solution right away | 21:48 |
| panda | bah, I'm starting to make no sense | 21:48 |
| rlandy | playbook? or forget it? | 21:49 |
| rlandy | what's with the tht tests/ | 21:49 |
| rlandy | what needs to change to those? | 21:50 |
| panda | rlandy: I'm not sure, there are some jobs that have a suffix -tht | 21:50 |
| panda | I did not check what they do different from the others | 21:50 |
| rlandy | diff from other tests? | 21:50 |
| panda | rlandy: yes | 21:50 |
| rlandy | ok - I'll check | 21:50 |
| panda | rlandy: do you have the test results, you want me to test the playbook debugging ? | 21:51 |
| rlandy | panda: go ahead | 21:51 |
| rlandy | the test results were the same | 21:51 |
| panda | oh joy | 21:52 |
| panda | rlandy: nothing works. Even calling the tasks in the playbook direclty without role makes zuul saying just MODULE_FAILURE | 22:07 |
| panda | rlandy: https://logs.rdoproject.org/67/16567/2/check/rdoinfo-tripleo-ocata-testing-centos-7-multinode-1ctlr-featureset005/faf1197/job-output.txt.gz#_2018-09-28_21_58_37_604977 | 22:08 |
| panda | we need to convert this shell script into something debuggable | 22:09 |
| *** toure is now known as toure|gone | 22:24 | |
| panda | rlandy: I did some testing locally, seems the shell module fails on the line "curl -o $CREPOS_FILE -sf "$logs/repos/changed_repos.txt" || exit 1" | 22:25 |
| panda | basically it exits 1 | 22:25 |
| panda | but it's a nightmare debugging it like this. | 22:25 |
| panda | I'm done, calling it a day | 22:25 |
| panda | this task is practically collecting all the technical debt left behind for the last 7 months ... | 22:26 |
| *** panda is now known as panda|off | 22:26 | |
| rlandy | panda|off: I hear you | 22:27 |
| rlandy | will pick this up on sunday | 22:27 |
| *** rlandy has quit IRC | 22:27 | |
| *** tosky has quit IRC | 22:42 | |
| hubbot1 | FAILING CHECK JOBS on master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/604298, stable/pike: tripleo-ci-centos-7-scenario001-multinode-oooq-container, tripleo-ci-centos-7-scenario004-multinode-oooq-container @ https://review.openstack.org/602248, master: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades, tripleo-ci-centos-7-scenario009 (2 more messages) | 22:49 |
| *** dtrainor__ has joined #oooq | 23:17 | |
| *** dtrainor_ has quit IRC | 23:20 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!