Wednesday, 2018-06-27

*** yolanda_ has quit IRC00:16
hubbotFAILING CHECK JOBS on stable/queens: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/56722400:52
hubbotFAILING CHECK JOBS on master: gate-tripleo-ci-centos-7-ovb-3ctlr_1comp-featureset001-ocata-branch, gate-tripleo-ci-centos-7-ovb-3ctlr_1comp-featureset001-master, gate-tripleo-ci-centos-7-ovb-3ctlr_1comp-featureset001-pike-branch @ https://review.openstack.org/560445, stable/queens: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/56722402:52
*** jaganathan has joined #oooq03:02
*** noama has joined #oooq03:06
*** hamzy has joined #oooq03:47
*** udesale has joined #oooq03:50
*** noama has quit IRC03:50
*** agopi has quit IRC03:50
*** ykarel has joined #oooq04:27
*** skramaja has joined #oooq04:34
*** noama has joined #oooq04:44
hubbotFAILING CHECK JOBS on master: gate-tripleo-ci-centos-7-ovb-3ctlr_1comp-featureset001-ocata-branch, gate-tripleo-ci-centos-7-ovb-3ctlr_1comp-featureset001-master, gate-tripleo-ci-centos-7-ovb-3ctlr_1comp-featureset001-pike-branch @ https://review.openstack.org/560445, stable/queens: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/56722404:52
*** ratailor has joined #oooq05:12
*** waleedm has joined #oooq05:28
*** quiquell|off is now known as quiquell|rover05:36
*** yolanda_ has joined #oooq05:47
*** saneax has joined #oooq05:58
*** yolanda__ has joined #oooq06:06
*** yolanda_ has quit IRC06:09
*** yolanda__ has quit IRC06:14
*** ccamacho has joined #oooq06:22
*** dmellado has joined #oooq06:29
*** yolanda__ has joined #oooq06:30
*** bogdando has joined #oooq06:35
quiquell|roverSome cores to unblock promotions https://review.openstack.org/#/c/578292/06:41
*** quiquell|rover is now known as quique|rover|afk06:42
*** kopecmartin has joined #oooq06:42
mariosquique|rover|afk: not sure i count but voted anyway seems straight forward enough :)06:44
ykarel+2 always counted :)06:46
*** yolanda__ has quit IRC06:50
mariosykarel: ;)06:50
hubbotFAILING CHECK JOBS on master: gate-tripleo-ci-centos-7-ovb-3ctlr_1comp-featureset001-ocata-branch, gate-tripleo-ci-centos-7-ovb-3ctlr_1comp-featureset001-master, gate-tripleo-ci-centos-7-ovb-3ctlr_1comp-featureset001-pike-branch @ https://review.openstack.org/560445, stable/queens: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/56722406:52
*** yolanda__ has joined #oooq06:53
*** yolanda__ is now known as yolanda06:53
*** udesale has quit IRC06:54
*** udesale has joined #oooq06:55
*** udesale has quit IRC06:59
*** udesale has joined #oooq07:00
*** tesseract has joined #oooq07:07
*** quique|rover|afk is now known as quiquell|rover07:08
quiquell|roverBack in business07:08
*** jaosorior has quit IRC07:23
*** zoli is now known as zoli|wfh07:28
*** zoli|wfh is now known as zoli07:28
*** brault has quit IRC07:30
*** brault has joined #oooq07:30
*** tosky has joined #oooq07:40
*** amoralej|off is now known as amoralej07:42
*** noama has quit IRC07:53
*** ykarel is now known as ykarel|lunch08:16
*** ykarel|lunch has quit IRC08:20
*** ykarel|lunch has joined #oooq08:21
*** dtantsur|afk is now known as dtantsur08:22
*** ccamacho has quit IRC08:27
*** ccamacho has joined #oooq08:27
*** holser_ has joined #oooq08:30
ajoquiquell|rover: marios ykarel|lunch  Do you know an easy way to bump nova::cpu_allocation_ratio from the quickstart side?08:38
ajoI'm investigating where I'd set that on a normal tripleo config, but, then I have to find out how to do it in quickstart08:40
ykarel|lunchajo, are you looking for something similar08:43
ykarel|lunchControllerExtraConfig:08:43
*** holser_ has quit IRC08:44
ykarel|lunch     nova::cpu_allocation_ratio: somtihng08:44
*** ykarel|lunch is now known as ykarel08:45
mariosajo: yeah i think you need to add some env file with your extraconfig override then add that to quickstart, like add _extra_args in your config08:45
mariosajo: eg. here https://github.com/openstack/tripleo-quickstart/blob/98fcc9d14d5b46a3bd456c55e8bab8c48d282427/config/general_config/featureset048.yml#L2908:46
mariosif you were using this for -c featureset08:46
mariosajo: but i've only very recently shifted focus to ci :) so quiquell|rover may well know of better ways08:46
mariosymmv08:46
mariost&c apply08:46
mariosetc08:46
*** holser_ has joined #oooq08:47
* quiquell|rover is with the fireman hat now, give me a few08:47
mariosajo: does it help/make sense?08:48
ajomarios: thanks looking :D08:49
ajomarios: yeah, makes sense, but I may need the extra yaml file to reference there uploaded to the undercloud, right?08:50
mariosajo: yeah heat needs to find it on the local host along with any other env files/templates08:50
hubbotFAILING CHECK JOBS on master: gate-tripleo-ci-centos-7-ovb-3ctlr_1comp-featureset001-ocata-branch, gate-tripleo-ci-centos-7-ovb-3ctlr_1comp-featureset001-master, gate-tripleo-ci-centos-7-ovb-3ctlr_1comp-featureset001-pike-branch @ https://review.openstack.org/560445, stable/queens: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/56722408:52
*** d0ugal has quit IRC08:54
*** holser_ has quit IRC09:16
*** holser_ has joined #oooq09:17
quiquell|rovercores: to fix linter problem09:23
quiquell|roverhttps://review.openstack.org/57833609:23
quiquell|roverhttps://review.openstack.org/57833909:24
quiquell|roverhttps://review.openstack.org/57834009:24
*** holser_ has quit IRC09:27
*** holser_ has joined #oooq09:28
*** holser_ has quit IRC09:30
*** holser_ has joined #oooq09:30
*** holser_ has quit IRC09:32
*** ykarel has quit IRC09:35
*** ykarel has joined #oooq09:39
*** holser_ has joined #oooq09:48
*** holser_ has quit IRC09:55
*** holser_ has joined #oooq09:56
sshnaidmquiquell|rover, what does mean 'yaml upgraded'?10:02
quiquell|roversshnaidm: Need to investigate10:07
*** gkadam__ has quit IRC10:16
*** gkadam__ has joined #oooq10:16
*** jaosorior has joined #oooq10:18
*** zoli is now known as zoli|lunch10:29
*** jaganathan has quit IRC10:46
hubbotFAILING CHECK JOBS on stable/queens: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/56722410:52
quiquell|roversshnaidm: https://review.rdoproject.org/r/#/c/14473/11:10
sshnaidmquiquell|rover, left comments there11:12
sshnaidmquiquell|rover, I don't understand this patch.. bluejeans?11:15
quiquell|roversshnaidm: Sure give me a sec11:16
quiquell|roversshnaidm: Calling you11:18
sshnaidmquiquell|rover, https://review.rdoproject.org/grafana/dashboard/db/tripleo-ci?orgId=111:20
*** udesale has quit IRC11:24
*** zoli|lunch is now known as zoli11:26
*** zoli is now known as zoli|wfh11:26
*** zoli|wfh is now known as zoli11:26
*** amoralej is now known as amoralej|lunch11:37
*** jfrancoa has joined #oooq11:48
sshnaidmquiquell|rover, http://38.145.34.41/11:52
*** atoth has joined #oooq11:55
*** jfrancoa has quit IRC12:04
*** dtantsur is now known as dtantsur|brb12:06
sshnaidmrfolco, hi12:14
rfolcoo/12:14
sshnaidmrfolco, any news about dependencies in internal SF?12:14
rfolcosshnaidm, nhicher fixed the gerrit user account... but still not fetching code. I did a test upstream and it works well....12:15
rfolcosshnaidm, there must be something missing...12:15
sshnaidmrfolco, ack12:16
rfolcoI'm doing some tests sshnaidm, will ping nhicher12:17
sshnaidmrfolco, anything that we can help him with it..?12:17
rfolcosshnaidm, not sure... I'll work on it, don't worry12:17
sshnaidmrfolco, ok, ping me if you need help with it12:18
rfolcosshnaidm, I will. Thanks.12:18
*** ratailor has quit IRC12:19
*** ykarel has quit IRC12:25
*** rlandy has joined #oooq12:29
*** trown|outtypewww is now known as trown12:30
rlandymyoung|off: ping re: virtblade email12:30
rlandyweshay|ruck: ^^ I am not using any of the virtblades anymore are you?12:31
rlandyone still says doc server12:32
*** ykarel has joined #oooq12:39
*** ykarel has quit IRC12:39
*** ykarel has joined #oooq12:40
quiquell|roversshnaidm ++12:41
quiquell|rover++ sshnaidm12:41
quiquell|rover++sshnaidm12:41
quiquell|roverhubbot dumn it12:41
hubbotquiquell|rover: Error: "dumn" is not a valid command.12:41
myoung|offrlandy: aye, i'll respond to email in a bit, omw to doctor.  We are using virtblade servers for at least 3 things now...doc server, an IRC bouncer used by a few folks in RelDel and myself, as well as an rdo2 symlink promoter arxcruz put in place a year ago.  We will need to migrate these to $otherPlaces12:42
*** zoli is now known as zoli|afk12:42
arxcruznot my fault!12:42
arxcruz:)12:43
myoung|offrlandy: I can't help but think there's a 4th thing, was going to log in to them to ensure we are in the clear prior to responding12:43
rlandymyoung|off, ok - we will talk later12:43
*** myoung|off is now known as myoung12:47
myoungarxcruz: didn't mean it that way :)  was part of sprint we did on internal CI12:48
hubbotFAILING CHECK JOBS on master: openstack-tox-linters @ https://review.openstack.org/560445, stable/queens: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/56722412:52
weshay|rucksshnaidm, quiquell|rover I'm still interested in giving the full tempest run fs20 more resources than the average job.. a larger flavor12:54
weshay|ruckthoughts?12:54
weshay|ruckI was previously nacked on that12:55
sshnaidmweshay|ruck, it should involve te-broker12:55
quiquell|roverweshay|ruck: Idempotency is activated ther e?12:55
weshay|ruckquiquell|rover, not sure if that merged.. but we're testing it in 27 so we could turn it back off in 2012:55
weshay|ruckquiquell|rover, that should only take about 5min12:55
quiquell|roverweshay|ruck: ack12:56
weshay|rucksshnaidm, ok.. can you help me understand how to use the te-broker to do that12:56
*** holser_ has quit IRC12:58
weshay|ruckpanda|off, 1-112:59
panda|offweshay|ruck: I'm in your room13:00
*** panda|off is now known as panda13:00
*** quiquell|rover is now known as quique|rover|lch13:01
quique|rover|lchweshay|ruck: I am testing the ARA + OC now, didn't have time at my morning13:01
sshnaidmweshay|ruck, if shortly, you need to hack this line: https://github.com/openstack-infra/tripleo-ci/blob/b229760043e50fbcb491cc423ed1241177f8710f/scripts/te-broker/create-env#L5113:03
sshnaidmweshay|ruck, but it's not configurable, and you can pass it only with testenv-client by adding new hackish option here: https://github.com/openstack-infra/tripleo-ci/blob/6369985369e9f2ace8f47ce4f3be621cb52e9cb6/testenv-client#L17913:04
sshnaidmweshay|ruck, and then passing it from there: https://github.com/openstack-infra/tripleo-ci/blob/1edb87d06888fb397ee39b2145f24d78affa7215/toci_gate_test-oooq.sh#L28513:05
sshnaidmweshay|ruck, I can do a POC, but only thinking about these hacks makes my cry..13:06
*** agopi has joined #oooq13:07
rlandyquique|rover|lch: sorry - forgot to add you to this request before ... https://redhat.service-now.com/pnt?id=ticket&table=x_redha_pnt_devops_table&sys_id=deede177db321f00a9e306e2ca961993 - for rhos-13 gates13:10
*** holser_ has joined #oooq13:12
*** ykarel is now known as ykarel|afk13:17
*** bogdando has quit IRC13:18
*** amoralej|lunch is now known as amoralej13:18
*** bogdando has joined #oooq13:21
*** ykarel|afk has quit IRC13:23
weshay|ruckarxcruz, will need a few minutes13:24
arxcruzweshay|ruck: you're the boss, boss :)13:24
*** zoli|afk is now known as zoli|wfh13:32
*** zoli|wfh is now known as zoli13:32
*** dtantsur|brb is now known as dtantsur13:35
*** quique|rover|lch is now known as quiquell|rover13:35
quiquell|roverrlandy: ack13:36
rlandyquiquell|rover: we're going nowhere without a reasonable sized node to deploy on13:36
trownsshnaidm: want to pull the trigger on https://review.openstack.org/558121 ... seems good to do on a wednesday :)13:39
sshnaidmtrown, yay \o/13:40
*** waleedm has quit IRC13:40
sshnaidmtrown, nobody will mention among all other problems :D13:40
trownsshnaidm: haha13:40
arxcruzweshay|ruck: let me know when13:45
weshay|ruckarxcruz, ready13:47
*** skramaja has quit IRC13:50
*** udesale has joined #oooq14:04
mariostrown: sorry binarya sounds awful :/14:07
trown:P14:07
panda*drumroll*14:18
trownwait for it14:19
quiquell|roverwhat's happening ?14:25
quiquell|roverI am always open for a beer14:26
panda*drumroll* intensifies14:26
* marios pokerface14:27
mariosso excited14:27
marios(no idea whats going on )14:27
quiquell|rovermarios: What are we waitting for ?14:27
quiquell|roverDamn...14:27
* marios pokerface intensifies14:27
quiquell|roverpanda: give me a clue14:27
pandano, I'm superstitious14:30
mariosso whats this ci escalation status call14:31
mariosand should we join it?14:31
mariosweshay|ruck: panda ? ^14:31
mariosmyoung: ?14:31
quiquell|rovermarios: is about ruck/rover talking about state of things14:31
weshay|ruckmarios, no.. that is ruck/rover14:31
mariosack14:31
mariosi might lurk then thanks14:31
weshay|ruckquiquell|rover, you on? I'm still in a meeting14:31
quiquell|roverI will connect now14:32
myoungmarios: what weshay|ruck said :)14:34
weshay|rucksshnaidm, panda, trown, rlandy https://review.rdoproject.org/r/#/c/14496/14:34
mariosthanks myoung14:35
pandaweshay|ruck: I am not very familiar with those jobs. They seem more on the packaging, DLRN and repos side14:39
pandaweshay|ruck: but yay, zuulv3 migration started!14:41
weshay|ruck:)14:42
panda*VICTORY TRUMPET*14:43
pandaso, the first drumroll was when overcloud deployed correctly, second was wehn tempest passed, now the test job is green. All welcome our new parent job "multinode"14:44
panda*applauses*14:44
pandaI'll rewrite the patch to test all the other job with the new configuration now14:45
trownnice14:48
trownguess we also need to split it into smaller mergable patches14:48
pandatrown: yes, I think I'll start a new patch14:49
*** hamzy has quit IRC14:49
hubbotFAILING CHECK JOBS on master: openstack-tox-linters @ https://review.openstack.org/560445, stable/queens: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/56722414:52
*** quiquell|rover is now known as quiquell|off14:52
*** hamzy has joined #oooq14:52
*** hamzy has quit IRC14:53
sshnaidmweshay|ruck, not sure I understand what happens in this patch: https://review.rdoproject.org/r/#/c/1449614:54
*** hamzy has joined #oooq14:54
weshay|rucksshnaidm, ok.. paul is in #rdo14:55
weshay|ruckask him14:55
*** waleedm has joined #oooq15:02
*** ledo_ has joined #oooq15:02
mariospanda: myoung /anyone else please i have a question. say reproducer failed for me like http://pastebin.test.redhat.com/609556 looks like running this https://github.com/openstack/tripleo-quickstart/blob/master/roles/repo-setup/tasks/setup_repos.yml15:06
* myoung looks15:06
mariospanda: myoung where are the logs for it so i can see more? i am on the undercloud15:06
mariospanda: myoung like there is no repo_setup_log generated15:06
*** waleedm has quit IRC15:07
myoungmarios: rlandy and I were just looking at an ovb job ironically at this same log.  an example would be https://thirdparty.logs.rdoproject.org/jenkins-tq-gate-rhos-13-ci-rhos-ovb-featureset001-82/undercloud/home/stack/repo_setup.log.txt.gz15:08
myoungwhich is generated by the tee --> log file of this jinja-template-generated .sh --> https://thirdparty.logs.rdoproject.org/jenkins-tq-gate-rhos-13-ci-rhos-ovb-featureset001-82/undercloud/home/stack/repo_setup.sh.txt.gz15:09
myoungmarios: do you have an tmate or link to job logs?  HTH15:09
mariosmyoung: i see... sec (in opt/stack right?015:09
rlandythis is a diff failures15:09
rlandyrhos-13 failed undercloud-setup15:09
mariosmyoung: so i don't have the usual logs15:09
myoungmarios: the actual log filename used is a variable and can be reset (https://github.com/openstack/tripleo-quickstart/blob/master/roles/repo-setup/defaults/main.yml#L4) but defaults to repo_setup.log15:09
myoungrlandy: diff fail, but an example of the same repo_setup.log with the details (unrelated root cause/issue)15:10
mariosmyoung: because i am running         /opt/stack/tripleo-ci/toci_gate_test-oooq.sh 2>&1 | tee console.log15:10
myoungrlandy, marios: i just had that log open in a tab as example :)15:10
mariosmyoung: thanks much for pointers appreciated15:10
rlandyweshay|ruck: arxcruz: myoung: we need to respond regarding the virtblade decommissioning email. I have nothing to keep on those machines15:19
pandatrown: one patch for the base + nodeset, one patch for the new run playbook, and I was thinking one patch per job category. One for undercloud jobs for example, to not migrate one job at a time ..15:19
rlandymyoung mentioned the tripleo-env doc15:19
trownpanda: makes sense15:20
rlandyand the upgrades of internal links15:20
rlandymyoung: pls list what we need to move15:20
myoungrlandy: from the email it looks like "we plan to retire the virtblade chassis by the end of next month." - so we need to migrate things by end of July15:22
*** kopecmartin has quit IRC15:24
myoungsshnaidm: are we still running a sova instance on virtblade14?15:25
*** gkadam__ has quit IRC15:26
myoungrlandy: virtblade13 is set up as a squid proxy, however the jobs that used to need it (osp need_images puddles) no longer run so dont' think  we need it anymore15:26
*** gkadam__ has joined #oooq15:26
rlandymyoung: will you be back by the end of JUly?15:26
rlandymyoung: I have a request in to reconfigure the networking of old env 1 to be independent hosts15:27
myoungrlandy: virtblade15 is afaik atilla's box that was running hubbot, i think that's since been migrated15:27
rlandyas such we will have more hardware available to us15:27
myoungrlandy: virtblade16 unknown, "temp test box"15:27
rlandybut I don;t have an ETA on that15:27
myoungrlandy: virtblade11 is running doc server VM, an quassel instance used by myself and RelDel folk.15:28
weshay|rucksshnaidm, http://logs.openstack.org/50/578150/1/check/tripleo-ci-centos-7-containers-multinode/5c81f43/logs/15:28
myoungrlandy: I'm back monday 16-July15:28
myoungrlandy: would prefer to migrate things then, i'm rapidly running out of time prior to PTO, hopping on a plane tonight15:29
myoungrlandy: I responded to Yash's email confirming the july 31 date15:35
rlandythanks15:35
rlandyI asked the same question15:35
rlandybased on the outcome of this ticket:15:35
myoungrlandy: ahhh lol just saw15:35
rlandyhttps://redhat.service-now.com/pnt?id=ticket&table=x_redha_pnt_devops_table&sys_id=6bd1c3f3db7e5f003306abc5ca96190d15:36
rlandy^^ we get that hardware,15:36
rlandywe can move whatever15:36
pandarfolco: addedd reviews to https://trello.com/c/tO42ubRD/824-create-new-base-jobs, please comment with the results of your tests15:38
rfolcopanda, I have added a comment already with my undercloud job test15:39
pandarfolco: that's cheating15:39
rfolcopanda, it works with swap15:39
pandarfolco: comment on the reviews, not only on the card15:39
pandarfolco: if youi have something to add to the changes15:40
myoungtrown: I've only hit one issue so far with libvirt reproducer etherpad... getting "zuul-hosts.yaml did not meet yaml requirements" when attempting to load inventory using yaml.py plugin15:40
rfolcopanda, ack15:40
*** bogdando has quit IRC15:41
pandamyoung: copying an pasting the yaml file created some problems for me with wierd characters, maybe it's what's happening with your yaml15:42
myoungtrown: I'm running ansible 2.5.5 on CentOS 7.5.1804, is that the same version you've been running?15:42
myoungtrown: ya retracing steps, I think there's some gunk from a copy paste as well15:42
pandamyoung: he's running 2.4.0, I already asked him once :)15:42
trownmyoung: ya that seems more likely... but I have ansible 2.4 on my dev box15:42
myoungpanda, trown: also some oddities around the yaml plugin and etherpad steps, got past them when the yaml inventory plugin is enabled it really doesn't like non-yaml inventories, even though extensions are set for the plugin.  I'll retest on 2.415:43
trownmyoung: updating it now though, so we will see15:43
myoungthe first steps were failing, had to disable yaml plugin, then re-enable to grok the zuul inventory15:44
trownmyoung: fwiw... I didnt have to do anything wrt the yaml plugin15:44
myoungtrown: I started with clean/new centos 7.5, and ansible 2.5.5, i might be hitting $newAnsibleVersionFun15:44
* marios almost hometime15:45
* rfolco almost lunchtime15:45
mariosi updated v4/5 @ https://review.openstack.org/#/c/578081/ weshay|ruck is that what you had in mind? fyi15:45
* myoung almost ptotime15:46
* panda almost kickbantime15:46
mariosthanks trown for checking https://review.openstack.org/578366 i think i will abandon that tomorrow morning and update the commit message at https://review.openstack.org/#/c/57808115:46
rfolcopanda wins15:46
mariosbut imo my biggest contribution today is this http://paste.openstack.org/show/724413/ i cant stop doing it15:46
mariosendless hours of fun15:47
*** saneax has quit IRC15:47
mariosmyoung: have a nice time15:48
mariosits totally fine15:48
marioswe'll stay here and work15:48
mariostrown: rlandy sshnaidm thanks for checking @ https://review.openstack.org/#/c/578023/15:49
mariossshnaidm: fwiw i did cycle back to your review request but i was too late i owe you one15:49
mariossshnaidm: for the record i'm usually in review mode in the morning, and then on request later depending on urgency (or i note it for next morning)15:49
mariossshnaidm: i think we have exact same TZ15:49
marioso/15:49
trownmyoung: hmm odd... I just updated centos7 and I still have ansible2.4 ... what repo are you using?15:49
weshay|ruckmarios, that looks like a nice add to the reproducer.. yes I think that does what I was thinking but need to test.. also looking for a similiar check around https://github.com/openstack/tripleo-quickstart/blob/master/quickstart.sh#L136-L16215:50
sshnaidmmarios, ack :)15:50
myoungtrown: ansible.noarch                          2.5.5-1.el7                    @epel15:50
weshay|ruckmarios, let me know if you have questions around that15:50
trownmyoung: eww... epel15:51
pandalol15:51
mariosweshay|ruck: ack noted will check tomrrow grateful for any more thoughts you might have today on the review if you get chance15:51
* myoung will retest with old(er) ansible15:51
weshay|ruckk.. thank you!15:51
weshay|ruckrlandy, aye re: virtblades.. on my todo15:52
mariosgimmeFeatureset 03715:52
mariosits fantastic15:52
trownwith epel on centos you might as well be on fedora15:52
myoungtrown: epel aside, should we then be targeting 2.4.1 (https://github.com/openstack/tripleo-quickstart/blob/master/requirements.txt#L3 source of truth?).15:56
myoungI've been doing some ansible module  exploration and it's handy to be running newer things, as devenv for ansible modules is to run ansible from git.  If it's "invalid" to be ahead w.r.t. ansible version makes total sense and I'll retest15:57
trownmyoung: to me epel is a bigger deal than being "ahead" on ansible... epel is totally untested15:58
trownwhereas base centos is just rhel15:59
* myoung loads less bleeding edge boot and will retest15:59
trownaka tested out the wazzoo15:59
*** myoung is now known as myoung|biaf15:59
pandamyoung|biaf: we target what's in cento-7 without epel, epel has been scrapped out of every possible place15:59
pandatrown: added some reviews for the reparenting16:00
weshay|ruckrlandy, probably makes sense to migrate those services to the dell boxes we are renicking.. they larger and we can host this stuff in vm's wdyt?16:01
rlandyweshay|ruck: ack - that is what I responded in the email16:01
rlandythe only thing is that we do not have an ETA on the renick project16:01
rlandyif we really have til the end of July, we will make it16:02
rlandyweshay|ruck - our task has been assigned to jpainter so we should be fine - no worries :)16:03
weshay|ruckrlandy++16:03
hubbotweshay|ruck: rlandy's karma is now 816:03
weshay|ruckthanks16:03
rlandypanda: I was thinking about the playbooks and release vars16:03
rlandyas one thought, what do you think about setting release based on tag/playbook ebing executed16:04
rlandybeing16:04
pandarlandy: you mean the "release" var itself ?16:05
rlandyyes16:05
pandarlandy: how do you deal with all the release-related variables ?16:06
*** myoung|biaf is now known as myoung16:06
pandarlandy: mmhh, maybe a conditional include_vars may work16:06
rlandypanda: well more a dict than a conditional16:06
rlandypanda: anyways, it's a half thought now16:07
pandarlandy: you mean something likle releases["queens"] = @queens.yml16:07
rlandyI need to experiment with that before making a formal suggestion16:07
pandarlandy: ok, I think you are in the right direction.16:08
rlandyreleases["queens"]["overcloud"]16:08
rlandyso we pass one var but it has all the info included16:08
*** waleedm has joined #oooq16:09
pandarlandy: do we need to change all the tasks to refer to this new releases[release]['overcloud'] dict that stores all the variables needed, like repo_cmd_after ?16:09
rlandyI would think so16:10
rlandyI just wanted to know if the idea was somewhat acceptable to you16:10
rlandybefore I proposed it16:10
pandarlandy: I would try to understand what's the rewrite overhead for this solution16:10
rlandyack16:10
rlandyI just think it's safer than trying to pass vars around a list of playbooks16:11
rlandypassed in zuul16:11
rlandywhere we have less control16:11
pandarlandy: yes, the releases dict was in my first releases overhaul design, but we put it aside because we thought it was too much rewrite16:11
*** ledo_ has quit IRC16:11
rlandypanda: I am worried about possible overrides if we rely on a single release var being passed around16:12
pandarlandy: override from whom ?16:13
rlandyas I understand, we will get to pass a variable in the job itself ... so you have the releases dict work you discarded, for comparison?16:17
rlandydo you16:17
pandarlandy: no, it was just an idea. I wanted to just crate a releases = {} in ansible then load all the releases files in the keys. So we would get releases['queens']['repo_cmd_after']16:19
rlandypanda: pretty much my thought16:19
pandarlandy: than each task would have to call releases[current_release]['repo_cmd_after']16:20
rlandyas considering we will be able to pass vars: in the job and that will apply to all playbooks16:20
rlandysomething like that16:20
rlandyit's ugly but safe16:20
pandarlandy: maybe now we don't have to rewrite everything, the dict idea needs refinement if we would still need to change the vars usage task per task16:21
*** apetrich has quit IRC16:21
*** apetrich has joined #oooq16:21
rlandypanda: ok - I'll refine the usage a bit - as long as you are bot completely opposed to the idea itself, it's worth the time to investigate16:21
pandarlandy: maybe using a dynamic include_vars in the a pre task taht we include before any playbook could be consider16:22
pandarlandy: we load only the variable that we need at the start of each playbook16:22
rlandywe have an example of thta16:22
rlandyweshay|ruck and I looked at it some time back16:22
rlandywhat I didn't like about that is that the task constantly runs16:22
rlandyif you add it to extras-common16:22
rlandyor the like16:22
* rlandy looks to see if if we still have that review somewhere16:23
pandarlandy: that's fine, maybe we can iunclude a role explicitly at the start of each playbook ... at this point I think anything woul be better than using multiple job inheritancein zuul.d16:24
myoungpanda: re: avoiding zuul job inheritance...why?  This seems to be one of the focal points of the zuul v3 design center.16:26
myoungpanda: do you mean "anything better *now* w.r.t. easy migration of existing job collateral?  or "no not ever" in a clean room implementation of new jobs?16:27
pandamyoung: not in general, but as a workaround to be able to pass different --extra-vars to each playbook in our list of playbooks to run16:27
rlandyweshay|ruck: do you still have that dynamic include_vars patch somewhere?16:29
rlandylike from last year maybe?16:29
weshay|rucklolz16:29
*** tosky has quit IRC16:29
weshay|ruckrlandy, I'll look16:29
rlandyI didn't find it16:29
*** tosky has joined #oooq16:29
weshay|ruckout of my 50 or so.. outgoing reviews :)16:30
weshay|ruckrlandy, I remember how to do it if needed16:31
weshay|ruckjust taking advantage of include_vars and when: statements16:31
rlandyweshay|ruck: we may need it - see discussion with panda re: release vars passed in zuul with list of playbooks16:31
rlandyone idea16:31
rlandytesting a few possibilities16:31
pandarlandy: anyway, check first how many things we would need to rewrite with the raw dict16:31
pandarlandy: then investiate the alternatives16:32
rlandypanda: ack - seems the most obvious approach if the damage is low16:32
pandarlandy: thanks for diving into this, among all the other tasks16:32
*** panda is now known as panda|bike16:33
weshay|ruckpassing vars w/ zuul would make it difficult for recreates no?16:33
*** udesale has quit IRC16:33
rlandyidk16:33
rlandydoing some investigation ahead of the next sprint16:34
panda|bikeweshay|ruck: we are currently passing variables needed for the pre-run playbook. We are trying to avoid passing too many vars16:34
*** tesseract has quit IRC16:34
rlandynot necessarily if we have pass the same vars as extra-vars to recreate16:34
rlandyalso the release var need to change per playbook run16:35
panda|bikewe will probably add just the split jobtype vars, and a few more16:35
rlandy^^ is the real issue16:35
myoungpanda|bike: i see16:36
weshay|ruckya.. I can see the jobtype being split16:37
*** gkadam__ has quit IRC16:37
weshay|rucksshnaidm, I updated that patch16:43
*** rlandy is now known as rlandy|brb16:45
hubbotFAILING CHECK JOBS on master: openstack-tox-linters @ https://review.openstack.org/560445, stable/queens: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/56722416:52
*** waleedm has quit IRC16:53
*** ssbarnea[m] has joined #oooq16:54
*** zoli is now known as zoli|gone17:02
*** zoli|gone is now known as zoli17:02
*** myoung is now known as myoung|lunch17:02
*** dtantsur is now known as dtantsur|afk17:04
*** rlandy|brb is now known as rlandy17:07
*** trown is now known as trown|lunch17:16
*** amoralej is now known as amoralej|off17:29
rlandyweshay|ruck: panda|bike: lol - seek and you shall find ... https://review.openstack.org/#/c/494674/17:45
rlandyhttps://review.openstack.org/#/c/494674/7/roles/common/tasks/main.yml17:45
weshay|rucknice17:46
weshay|ruckrlandy, panda|bike ya.. I'm a fan of include_vars17:46
rlandyit only took reading through all weshay|ruck's reviews until last year August17:46
rlandyit's a reasonable idea17:46
weshay|ruckheh17:46
rlandywe will have to weigh the options up against eachother17:46
* rlandy ran out of sprint work for prepping for next sprint17:47
rlandywhat I'm thinking is we get to assign release per playbook/per role based on jobtype and the playbook running17:51
*** atoth has quit IRC18:00
*** yolanda_ has joined #oooq18:11
*** trown|lunch is now known as trown18:14
*** yolanda has quit IRC18:14
weshay|rucktrown, panda|bike https://review.openstack.org/#/c/576597/18:19
*** myoung|lunch is now known as myoung18:21
trownweshay|ruck: how are hammer/jewel getting installed in the first place?18:21
weshay|ruckI could ping kb to find out ;)18:22
weshay|ruckfak18:25
weshay|ruckthe reproducer is busted18:25
weshay|ruckdon't we have a whole team on this???18:29
weshay|ruckhttps://review.openstack.org/#/c/576834/18:29
weshay|ruckmyoung, trown ?18:29
weshay|ruckpanda|bike, ^18:29
weshay|ruckpanda--18:34
weshay|rucktrown--18:34
hubbotweshay|ruck: trown's karma is now 118:34
weshay|ruckmyoung==18:34
weshay|ruckmyoung--18:34
hubbotweshay|ruck: myoung's karma is now 118:34
weshay|ruckhttps://bugs.launchpad.net/tripleo/+bug/177897418:34
openstackLaunchpad bug 1778974 in tripleo "reproduce-quickstart.sh broken by 576834, no such file" [Critical,Triaged]18:34
* myoung makes a note in https://trello.com/c/bLWnZcWq/781-ci-job-create-libvirt-reproducer-ci-job, which was punted from last sprint and looks at the bug18:35
*** rlandy is now known as rlandy|afk18:41
rlandy|afkgoing to do airport pickup - back later18:42
weshay|ruckmyoung, why are you reverting the change twice?18:51
myoungweshay|ruck, panda|bike, trown, sshnaidm, rlandy|afk: https://review.openstack.org/#/c/57852018:51
myoungweshay|ruck: I didn't see your revert, sorry.18:52
hubbotFAILING CHECK JOBS on master: openstack-tox-linters @ https://review.openstack.org/560445, stable/queens: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/56722418:52
myoungdisregard18:52
sshnaidmweshay|ruck, myoung let's not revert, but fix it18:53
sshnaidmweshay|ruck, myoung I don't see a point to revert the whole change instead of small fix18:53
*** atoth has joined #oooq18:54
*** holser_ has quit IRC18:55
myoungI think a revert is cleaner w.r.t. git history - patch broke this is an ungated dependency...IMHO step 1:  revert patch, return to known state.  step 2: repost updated patch.  step 3: profit and have clear git history.  I've updated the revert commit message with link to bug.18:59
myoung^^ just my opinion19:00
trownrevert is very heavy handed for a very simple fix19:03
trownI am posting said simple fix19:03
trownhttps://review.openstack.org/578530 is the only thing that is truly broken... will put up a second patch to tqe to change the script name19:05
trownand the rename https://review.openstack.org/57853319:10
myoungsshnaidm, trown: "the point" is to have clarity and to go from a known-good state to known good state, particularly for ungated dependencies.  Until we address https://trello.com/c/bLWnZcWq/781-ci-job-create-libvirt-reproducer-ci-job, it's risky.  Yes a revert is heavy handed, but predictable.  patch broke thing.  revert patch.  What if there's another small thing that we missed as part of a well intentioned review?  What if there's another19:17
myoungmole after this one?  Will we remember all of these small things a year from now?    We can iterate and fix...that works too.   And I understand the desire to not re-review a massive patch.   {shrug}19:17
* myoung turns his talker off and look at the patches/fixes.19:17
trownfix is literally one line19:18
trownnot reverting a huge patch for one line19:18
myoungi know.  i'm disabling my pedantic OCD unit.19:18
trownI had that logs dir in my snapshot19:18
sshnaidmmyoung, you're right, but this fix is really trivial19:19
sshnaidmit's good reminder tough that we need to test reproducer on every patch..19:19
trownsshnaidm: can you look at https://review.openstack.org/578533 too... name of the script is also important19:20
sshnaidmtrown, yep, thanks19:21
weshay|ruckhttps://review.openstack.org/#/c/576858/6/zuul.d/multinode-jobs.yaml19:45
weshay|ruckneeds review19:45
weshay|ruckhttps://review.openstack.org/#/c/576957/2/zuul.d/multinode-jobs.yaml19:46
*** yolanda_ has quit IRC20:43
*** yolanda_ has joined #oooq20:44
hubbotFAILING CHECK JOBS on master: openstack-tox-linters @ https://review.openstack.org/560445, stable/queens: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/56722420:52
*** yolanda has joined #oooq20:53
*** yolanda_ has quit IRC20:55
*** trown is now known as trown|outtypewww20:55
*** yolanda_ has joined #oooq20:59
*** yolanda has quit IRC21:02
*** jtomasek has quit IRC21:03
*** sanjay__u has joined #oooq21:34
*** jtomasek has joined #oooq21:36
*** myoung is now known as myoung|pto21:41
*** agopi has quit IRC21:43
*** rlandy|afk is now known as rlandy22:15
*** agopi has joined #oooq22:17
hubbotFAILING CHECK JOBS on stable/queens: tripleo-ci-centos-7-scenario000-multinode-oooq-container-upgrades @ https://review.openstack.org/567224, master: openstack-tox-linters @ https://review.openstack.org/56044522:52
*** jjoyce|gone has joined #oooq23:06
*** jjoyce has quit IRC23:08
*** jjoyce has joined #oooq23:08
*** jjoyce|gone has quit IRC23:09
*** yolanda__ has joined #oooq23:11
*** yolanda_ has quit IRC23:15
*** yolanda__ has quit IRC23:18
*** yolanda has joined #oooq23:20
*** yolanda_ has joined #oooq23:21
*** yolanda has quit IRC23:24
*** yolanda__ has joined #oooq23:25
*** yolanda_ has quit IRC23:29
*** yolanda has joined #oooq23:29
*** yolanda__ has quit IRC23:31
*** tosky has quit IRC23:35
*** brault has quit IRC23:54

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!