Friday, 2019-12-13

*** ysandeep has joined #oooq00:35
*** saneax has quit IRC00:51
*** ysandeep has quit IRC01:01
*** ysandeep has joined #oooq01:03
*** ysandeep has quit IRC02:55
*** ysandeep has joined #oooq03:13
*** aakarsh has joined #oooq03:31
*** ysandeep has quit IRC03:36
*** skramaja has quit IRC03:44
*** bhagyashris has joined #oooq03:54
*** rlandy has quit IRC04:07
*** udesale has joined #oooq04:29
*** ykarel has joined #oooq04:37
*** epoojad1 has joined #oooq04:53
*** raukadah is now known as chkumar|rover05:08
*** holser has joined #oooq05:29
*** surpatil has joined #oooq05:35
*** holser has quit IRC05:46
*** skramaja has joined #oooq06:00
*** soniya29 has joined #oooq06:07
*** udesale has quit IRC06:25
*** marios|ruck has joined #oooq06:31
*** jbadiapa has joined #oooq06:43
*** jfrancoa has joined #oooq06:51
chkumar|rovermarios|ruck: Good morning07:15
chkumar|rovermarios|ruck: master and train RHEL8 promotion https://bugs.launchpad.net/tripleo/+bug/185627807:15
openstackLaunchpad bug 1856278 in tripleo "RHEL8 scenario 1 standalone deployment failed with The following containers failed validations and were not started: collectd" for master and train" [Critical,Confirmed]07:15
chkumar|roverbug07:15
marios|ruckchkumar|rover: o/ men07:20
marios|ruckman i am very slow today07:20
marios|rucki have some bug :(07:20
marios|ruckchkumar|rover: will check in  abit thanks07:21
chkumar|rovermarios|ruck: for tempest working on proposing patch07:23
chkumar|roveractual fix07:23
*** soniya29 has quit IRC07:39
*** soniya29 has joined #oooq07:40
*** tesseract has joined #oooq07:43
*** saneax has joined #oooq07:44
*** ykarel is now known as ykarel|lunch07:47
marios|ruckchkumar|rover: train promoting ... would be nice to get master gnna check it in a minute. but rhel is totally f***d really red and blocked on that image build :/07:59
*** jtomasek has joined #oooq08:03
*** amoralej|off is now known as amoralej08:27
chkumar|rovermarios|ruck: ah sweet :-)08:27
marios|ruckchkumar|rover: the rocky fs2 upload worked too it promoted last night08:29
marios|ruckso apart from rhel we are pretty good08:29
chkumar|rovermarios|ruck: on rhel8 standalone jobs, we have just one failure on master and train08:29
marios|ruckchkumar|rover: yeah but image build08:30
marios|ruckblocks everything08:30
chkumar|rovermarios|ruck: reproducing it08:30
chkumar|rover*trying to reproduce it08:30
marios|ruckchkumar|rover: ack good luck08:30
marios|ruckmatbu: o/08:40
*** tosky has joined #oooq08:40
marios|ruckmatbu: did you file a bug for tripleo-ci-centos-7-containerized-undercloud-upgrades train & 2019-12-12 09:11:39.923379 | primary | TASK [validate-services : Fails if we find failed systemd units] ***************08:40
marios|ruckmatbu: can't see one there ... https://bugs.launchpad.net/tripleo/+bugs?orderby=-date_last_updated&start=008:42
marios|ruckmatbu: going to file one now then08:43
*** ykarel|lunch is now known as ykarel08:45
*** jpena|off is now known as jpena08:54
marios|ruckchkumar|rover: matbu: fyi https://bugs.launchpad.net/tripleo/+bug/1856288  The train tripleo-ci-centos-7-containerized-undercloud-upgrades often fails on validate-services ironic_pxe_tftp healthcheck08:55
openstackLaunchpad bug 1856288 in tripleo "The train tripleo-ci-centos-7-containerized-undercloud-upgrades often fails on validate-services ironic_pxe_tftp healthcheck" [Critical,Triaged]08:55
chkumar|roveraye sir08:56
marios|ruckchkumar|rover: matbu is already working there so just fyi for now08:58
marios|ruckchkumar|rover:         * https://review.opendev.org/#/c/698663/3 i think is part of that08:58
matbumarios|ruck: ack you can assigned it to me, im working on it09:01
marios|ruckthanks matbu09:01
*** holser has joined #oooq09:18
*** derekh has joined #oooq09:30
marios|rucksshnaidm|off: chkumar|rover: we should bring this discussion to next scrum https://review.rdoproject.org/r/#/c/24073/1/zuul.d/upstream.yaml see if anyone has better ideas09:45
chkumar|roversouds good to me09:46
chkumar|rovermarios|ruck: weshay I tried to reproduce the overcloud build image locally, but not able to reproduce09:49
chkumar|rovertime to hold the node09:49
marios|ruckchkumar|rover: ack09:49
marios|ruckchkumar|rover: can you just point to a review or what do we need to request a hold?09:49
chkumar|roverneed to send a review then ask the sf-ops to hold the node once we recheck the job again09:50
marios|ruckchkumar|rover: ack thats what i mean thanks (only did it once and long time ago)09:50
chkumar|rovermarios|ruck: please send a review up09:51
chkumar|roverwith overcloud build image job09:51
marios|ruckack chkumar|rover maybe we can use that sec09:51
marios|ruckhttps://review.rdoproject.org/r/#/c/2391909:51
marios|ruck?09:51
marios|ruckchkumar|rover: or has to be just one job09:51
chkumar|rovermarios|ruck: good candidate09:52
chkumar|rovertime to jump on sf-ops09:52
marios|ruckchkumar|rover: ack09:52
marios|ruckchkumar|rover: going to get some food and medicine will be biab10:13
chkumar|rovermarios|ruck: sure,10:13
*** marios|ruck has quit IRC10:18
*** marios|ruck has joined #oooq10:43
*** yolanda has quit IRC10:53
*** ykarel is now known as ykarel|afk10:57
marios|ruckchkumar|rover: time on the logs weird? am looking in the held node now but the build.log has 2019-12-13 10:17:02.041 | but local machine says 06:00:41 EST 201911:00
marios|ruckchkumar|rover: fyi the ram thing you noticed yesterday is not fatal its a warning e.g. i see it in a green log http://logs.rdoproject.org/openstack-periodic-master/opendev.org/openstack/tripleo-ci/master/periodic-tripleo-rhel-8-buildimage-overcloud-full-master/ae9503e/build.log11:04
chkumar|rovermarios|ruck: yes11:04
chkumar|rovermarios|ruck: few more things discovered11:04
chkumar|rovermarios|ruck: on RDO cloud, rhel-8.0 image with selinux premissive available and there is working finer11:05
marios|ruckchkumar|rover: hmm we can test that with semanage i think we have a var for that11:05
chkumar|rovermarios|ruck: on zuul job, rhel-8.1 image is used selinux enforcing11:05
chkumar|roverwhere it is failing11:05
marios|ruckchkumar|rover: nice i hope that's it man would be great11:05
marios|ruckfriday 13th we promote all the things!11:06
chkumar|roverjpena: Hello11:06
marios|ruckchkumar|rover: indeed is enforcing in that node11:06
chkumar|roverjpena: is it possible to change the selinux permissive in rhel8 nodepool image11:07
chkumar|rover*to permissive11:07
chkumar|rovermarios|ruck: I am trying to reproduce the same on rdocloud vm11:07
marios|ruckchkumar|rover: ./roles/common/defaults/main.yml:42:selinux_enforcing: false11:07
jpenachkumar|rover: by default? I don't think it's a good idea. I'd rather set it as part of the job11:07
marios|ruckchkumar|rover: maybe we can just set that in featureset ^11:07
marios|ruckjpena: yeah we suspect it is cause of https://bugs.launchpad.net/tripleo/+bug/1853028 chkumar|rover++ noticed it11:07
openstackLaunchpad bug 1853028 in tripleo "Build overcloud image for rhel8 fails sometimes on in_target.d/post-install.d/51-enable-network-service" [Critical,Triaged]11:07
marios|ruckjpena: s/suspect/really hope11:08
marios|ruckjpena: ack we can use a var for testing it11:08
pandamarios|ruck: anything you want me to do to test the queens manifest ?11:08
marios|ruckpanda: hmm not really checked that yet panda i posted the patch yesterday but didn't check results will do in a bit11:08
marios|ruckpanda: it should be trying to pull _manifest there11:08
marios|ruckhttps://review.opendev.org/#/c/687267/ & posted queens tht w/depends-on @ https://review.opendev.org/698747 panda11:09
marios|ruckpanda: i will have a look at that stuff in a while ^11:09
marios|ruckpanda: in particular the updated v3/4 @ https://review.opendev.org/#/c/687267 happened late yesterday not sure it gives us what we want ... originally it was for standalone but no such thing for Q11:09
marios|ruckchkumar|rover: gonna update the review so we can rerun?11:10
marios|ruckchkumar|rover: hmm let me check is it in allowed overrides?11:10
chkumar|rovermarios|ruck: can we send a different review11:10
marios|ruckchkumar|rover: just to test11:10
chkumar|roverjpena: if we recheck the review does the node gets deleted?11:11
chkumar|roveri mean holded node11:11
marios|ruckchkumar|rover: we can't use featureset_override https://github.com/openstack/tripleo-ci/blob/master/roles/run-test/tasks/main.yaml#L8 will have to update the fs for that11:11
marios|ruckchkumar|rover: i suspect it will and we get a new one jpena ?11:11
jpenachkumar|rover: do you mean the existing held node? No, it will stay around11:11
marios|ruckah cool11:11
chkumar|roverjpena: ok11:11
jpenaand yes, a new node will be spawned for the new job11:12
chkumar|rovermarios|ruck: we can send a review on tripleo ci and we can add depend on which var here, may be it will works?11:12
marios|ruckjpena:11:12
marios|ruckoh11:12
marios|ruckjpena: so we need to hold the new node?11:12
jpenamarios|ruck: if you want, yes11:12
marios|ruckjpena: and we can release this one then11:12
jpenaack11:12
marios|ruckchkumar|rover: k ?11:12
jpenalet me know when I can delete it11:12
marios|ruckjpena: thanks will do11:12
chkumar|rovermarios|ruck: jpena works for me11:12
marios|ruckchkumar|rover: are you doing somethign or should i post it11:12
marios|ruckchkumar|rover: we do need a new patch11:12
marios|ruckchkumar|rover: cos we need to add into featureset or somewhere11:13
marios|ruckchkumar|rover: 13:11 < marios|ruck> chkumar|rover: we can't use featureset_override https://github.com/openstack/tripleo-ci/blob/master/roles/run-test/tasks/main.yaml#L8 will have to update the fs11:13
chkumar|rovermarios|ruck: let me send a patch11:13
marios|ruckchkumar|rover: OK11:13
marios|ruckchkumar|rover: let  me know to add it immediately at https://review.rdoproject.org/r/23919 and we can request another hold11:14
marios|ruckand release this one11:14
marios|ruckchkumar|rover: another tempest related on scen4 master pipeline still runnign but that one failed posted testproject in a sec11:17
chkumar|rovermarios|ruck: https://review.opendev.org/#/c/698876/11:19
chkumar|roverI also need to change the flag somewhere na?11:19
marios|ruckchkumar|rover: oh you added it to featureset_override?11:19
marios|ruckchkumar|rover: ok fine lets just use it for the test. i though you would update the featureset but fine11:20
marios|ruckchkumar|rover: is fine i can just pass it with job vars11:20
marios|ruckre 13:19 < chkumar|rover> I also need to change the flag somewhere na?11:20
chkumar|rovermarios|ruck: yes passing it with job var it will work11:21
chkumar|rovermarios|ruck: rhel8 overcloud image not featureset is used11:21
marios|ruckchkumar|rover: k done v6 https://review.rdoproject.org/r/#/c/23919/6/.zuul.yaml11:23
marios|ruckjpena: can we please have that? just posted it we need periodic-tripleo-rhel-8-buildimage-overcloud-full-master please thanks11:23
marios|ruckchkumar|rover: look ok?11:23
jpenamarios|ruck: ok, I'll release the old node and hold the new one11:24
marios|ruckjpena: fine for me chkumar|rover you need the old one still?11:24
chkumar|rovermarios|ruck: looks ok but it is not going to work http://codesearch.openstack.org/?q=selinux_enforcing&i=nope&files=&repos=11:24
chkumar|roveras that var used in tripleo-quickstart only11:24
marios|ruckchkumar|rover: :/11:24
marios|ruckchkumar|rover: so it has to be on the guest11:24
marios|ruckimage i mean virt-customize type thing11:25
*** yolanda has joined #oooq11:25
marios|ruckchkumar|rover: maybe we can use tripleo-heat-templates SELinuxMode11:26
chkumar|rovermarios|ruck: sorry but we need to change the selinux mode to permissive of the job vm which is used and there playbook runs11:26
marios|ruckchkumar|rover: hmmm only in /docker-baremetal-ansible.yaml11:27
marios|ruckchkumar|rover: well we just need a setenforce 0 before image build11:27
marios|ruckchkumar|rover: we could just hack it11:27
marios|ruckchkumar|rover: like11:27
marios|ruckchkumar|rover: (for testing the theory i mean at least)11:28
chkumar|rovermarios|ruck: ok let me do that in the same review11:28
marios|ruckchkumar|rover: here chandan https://github.com/openstack/tripleo-ci/blob/master/roles/oooci-build-images/templates/build-images.sh.j211:28
marios|ruckchkumar|rover: wdyt?11:28
marios|ruckchkumar|rover: we can just add setenforce to test it11:28
marios|ruckjpena: sorry11:29
marios|ruckjpena: please don't hold the new node11:29
chkumar|rovermarios|ruck: that will alos work11:29
marios|ruckjpena: i made a mistake11:29
jpenaok, no worries11:29
marios|ruckjpena: we will update and promise last time11:29
marios|ruckjpena: sorry for the noise thank so much for your support11:29
jpenashould I remove the old node?11:30
marios|ruckchkumar|rover: should i post something  for that?11:30
marios|ruckjpena: yes for me i don't need it chkumar|rover do you need the old node?11:30
chkumar|rovermarios|ruck: me neither11:30
marios|ruckchkumar|rover: k posting a change to build template for testing11:31
marios|ruckchkumar|rover: we don't need to hold node11:31
marios|ruckchkumar|rover: lets just run it see what happens11:31
marios|ruckwe can ping javier again later11:31
marios|ruckchkumar|rover: ?11:31
marios|ruckchkumar|rover: K?11:31
chkumar|roverok11:32
*** epoojad1 is now known as epoojad1|afk11:39
chkumar|roverbrb11:40
*** ykarel|afk is now known as ykarel11:42
*** epoojad1|afk has quit IRC11:44
*** bhagyashris has quit IRC11:52
marios|ruckbrb11:58
*** rfolco has joined #oooq11:58
*** dtantsur|afk is now known as dtantsur12:02
chkumar|rovermarios|ruck: weshay I reproduced the issue12:05
marios|ruckchkumar|rover: image build you mean ?12:05
chkumar|rovermarios|ruck: yes, in rdo cloud12:05
chkumar|rovermarios|ruck: change the selinux mode from permissive to enforcing12:05
marios|ruckchkumar|rover: perfect12:05
marios|ruckchkumar|rover: so thats it then12:05
marios|ruckchkumar|rover: still running there https://review.rdoproject.org/zuul/stream/69a9ff6b8dd34e56a2803eb378da85f7?logfile=console.log12:06
chkumar|rovermarios|ruck: http://paste.openstack.org/show/787545/12:06
chkumar|rovermarios|ruck: let it finish then we can update the patch12:07
marios|ruckchkumar|rover:++ you found it mate12:08
chkumar|rovermarios|ruck: we did it together and thanks to jpena also :-)12:08
marios|ruckchkumar|rover: maybe we can get some good logs for slaweq fs 1 too at https://review.rdoproject.org/r/#/c/23919/12:08
chkumar|rovermarios|ruck: yes12:09
*** EmilienM has quit IRC12:12
*** EmilienM has joined #oooq12:12
*** EmilienM is now known as EvilienM12:17
chkumar|rovermarios|ruck: it passed12:18
marios|ruckchkumar|rover: \o/12:18
marios|ruckchkumar|rover: so what do we do ? add conditional into the build script for rhel8 and setenforce?12:22
marios|ruckchkumar|rover: virtcustomize the image?12:22
marios|rucki know weshay will want it on the image by default12:22
marios|ruckat least he has done in the past for this kind of thing like rhui for example12:23
*** jpena is now known as jpena|lunch12:29
chkumar|roverkopecmartin: soniya29 surpatil please have a look at agenda https://hackmd.io/fIOKlEBHQfeTZjZmrUaEYQ?view12:32
chkumar|rovermarios|ruck: sorry coming back to the above question12:36
marios|ruckchkumar|rover: np updated the https://bugs.launchpad.net/tripleo/+bug/1853028/comments/1312:37
openstackLaunchpad bug 1853028 in tripleo "Build overcloud image for rhel8 fails sometimes on in_target.d/post-install.d/51-enable-network-service" [Critical,Triaged]12:37
marios|ruckchkumar|rover: and trello12:37
chkumar|rovermarios|ruck: adding condition or a task in build images role itself will help12:37
chkumar|rovermarios|ruck: may be in future if the same role will be used in downstream then we need to just flip the selinux switch12:37
marios|ruckchkumar|rover: yeah but still hacky sure we can just update that https://review.opendev.org/69888312:37
marios|ruckchkumar|rover: but not sure it will be acceptable12:37
chkumar|rovermarios|ruck: we will get it merged12:38
chkumar|roveras it unblocks our ci12:38
marios|ruckchkumar|rover: ok let the run finish before i update it12:38
marios|ruckchkumar|rover: so we get the fs1 logs for slaweq12:38
chkumar|rovermarios|ruck: yup12:38
* marios|ruck looks at clock12:38
marios|ruckdon't know if we'll see the rhel8 promote but maybe we can set it up for happening later12:38
marios|ruckchkumar|rover: ^12:38
chkumar|rovermarios|ruck: if we get one successful run, it will get promoted12:39
marios|ruck;)12:39
marios|ruckchkumar|rover: but we also need fs112:40
marios|ruckchkumar|rover: right? image build and fs1 i think for rhel missing12:40
* marios|ruck checks promoter logs12:40
marios|ruckchkumar|rover: hm also standalone12:41
marios|ruckchkumar|rover: but is still running master pipeline now12:41
marios|ruckchkumar|rover: there though fs1,scen1 and buildimage missing successful jobs: [u'periodic-tripleo-ci-rhel-8-scenario001-standalone-master', u'periodic-tripleo-ci-rhel-8-ovb-3ctlr_1comp-featureset001-master', u'periodic-tripleo-rhel-8-buildimage-overcloud-full-master']12:41
marios|ruckhttp://38.145.34.55/redhat8_master.log12:41
marios|ruckchkumar|rover: hmmm hope that scenario 1 is not a new blocker :/12:42
marios|ruckchkumar|rover: it failed on current run for scen1 too12:42
marios|ruckchecking ...12:42
marios|ruckfax                 * http://logs.rdoproject.org/openstack-periodic-master/opendev.org/openstack/tripleo-ci/master/periodic-tripleo-ci-rhel-8-scenario001-standalone-master/879e792/logs/undercloud/home/zuul/standalone_deploy.log.txt.gz12:42
marios|ruck                * 2019-12-13 10:06:45 |   File "/usr/lib/python3.6/site-packages/tripleoclient/v1/tripleo_deploy.py", line 1323, in _standalone_deploy12:42
marios|ruck        2019-12-13 10:06:45 |     raise exceptions.DeploymentError('Deployment failed')12:42
chkumar|rovermarios|ruck: https://review.opendev.org/#/c/698889/12:43
* marios|ruck hug chkumar|rover 12:43
chkumar|roverwill fix collectd issue12:43
* chkumar|rover hugs back marios|ruck12:43
marios|rucksweet12:43
* chkumar|rover looks for panda12:43
marios|ruckso we wait for current run to finish12:43
marios|ruckthen post testproject/rdo and done12:43
marios|ruckit should promote12:43
chkumar|roveryup12:43
marios|ruckwe have a way to build image12:44
* marios|ruck starts defrosting mojito12:44
*** ykarel is now known as ykarel|away12:48
chkumar|roverzbr: can we close this issue https://github.com/containers/libpod/issues/4580 since it is fixed12:51
chkumar|rover?12:51
pandaaaawwww12:51
zbrchkumar|rover: not really, we did build the rpm by disabling the gpgme! so mainly by shipping less of podman :D12:57
zbri think but is still valid, but I will  update it to make it clear12:57
chkumar|roverzbr: sure12:57
*** akahat has joined #oooq12:59
zbrchkumar|rover: thanks for reminding, i updated it.13:00
*** ykarel_ has joined #oooq13:00
chkumar|roverweshay: kopecmartin zbr soniya29 surpatil meeting time13:01
marios|ruckman some patches just DO NOT want to merge aweeeooooooooo recheck! https://review.opendev.org/#/c/695878/1 https://review.opendev.org/#/c/696872/113:01
marios|ruckrecheck13:01
marios|ruckrecheck13:01
marios|ruckeach time a different job!13:02
marios|ruckcan whoever hex those please unhex them so they go through whatever i did to you i am sorry13:03
*** ykarel|away has quit IRC13:03
marios|ruckpanda: 2019-12-12 17:27:43 | 2019-12-12 17:27:43,302 ERROR: Failed running docker push for 192.168.24.1:8787/tripleoqueens/centos-binary-neutron-openvswitch-agent:dc9da4e8d8269365a7af28aeebeaa7579382a132_77cf1e91_manifest13:06
marios|ruckhttps://52a5c01fbc2e5a1c1c79-24704190c22ac93ed09ee07fafdd15be.ssl.cf1.rackcdn.com/698747/1/check/tripleo-ci-centos-7-containers-multinode/3ed648d/logs/undercloud/home/zuul/overcloud_prep_containers.log.txt.gz13:06
marios|ruckpanda: want to jump in a call?13:07
marios|ruckpanda: i miss you13:07
marios|rucki promise i won't hug you i have virus13:07
marios|ruckwell promise is a strong work13:08
marios|rucklike i'll take it under very serious consideration anyway13:09
marios|rucks/work/word 15:08 < marios|ruck> well promise is a strong work13:09
chkumar|rovermarios|ruck: I think rhel8 fs01 train job the issue is same I need to find the selinux part and fire up the job http://logs.rdoproject.org/openstack-periodic-latest-released/opendev.org/openstack/tripleo-ci/master/periodic-tripleo-ci-rhel-8-scenario001-standalone-train/493b7a4/job-output.txt13:21
marios|ruckchkumar|rover: ack maybe the first thing you posted or just update the featureset13:21
chkumar|roversorry wrong log13:22
chkumar|rovermarios|ruck: I am looking first whether is enforcing or permissive13:25
marios|ruckchkumar|rover: ack we have it in /extra i believe13:25
chkumar|roverfrom here it is permissive http://logs.rdoproject.org/openstack-periodic-latest-released/opendev.org/openstack/tripleo-ci/master/periodic-tripleo-ci-rhel-8-ovb-3ctlr_1comp-featureset001-train/e2b80c8/logs/undercloud/var/log/extra/selinux.txt.gz13:25
marios|ruckchkumar|rover: k13:25
*** cgoncalves has joined #oooq13:31
beaglesk so I have a bit of a mystery and could use some advice on how to proceed. For a couple of weeks now, I've been trying to get the octavia deployment on queens and rocky to pick up a downloaded amphora image.13:34
*** ykarel_ has quit IRC13:34
beaglesI've been hacking away using a few different patches trying to figure out what's going on: https://review.opendev.org/#/c/692579/13:34
beagleshttps://review.opendev.org/#/c/692857/13:34
beaglesand https://review.opendev.org/#/c/692858/13:35
beagles(the first drives the test scenario to try out the other two)13:35
*** rlandy has joined #oooq13:35
beaglesyou can see in the journals on the undercloud that the image is downloaded and does in fact exist there.. but the ansible run by mistral via workflow tasks does not see it13:36
beagless/does/cannot/13:36
marios|ruckbeagles: tempest fail?13:36
beaglesafaict the permissions are okay13:36
marios|ruckbeagles: https://review.opendev.org/#/c/692579/ i see TASK [validate-tempest : Execute tempest fails in https://38fc547d6dc8718028ea-253b0a19be2181811797fa6cbd0c2b8d.ssl.cf5.rackcdn.com/692579/12/check/tripleo-ci-centos-7-scenario010-multinode-oooq-container/2cf8cd5/job-output.txt13:36
*** jpena|lunch is now known as jpena13:37
beaglesmarios|ruck, the stat task in the ansible returns false so it doesn't get loaded into glance13:37
beaglesmarios|ruck, so the tempest fail is expected because they cannot run13:37
marios|ruckbeagles: ack i see13:37
marios|ruckscrolling up13:37
beaglesmarios|ruck, I've loaded up a bunch of 'find' tasks etc to get some insight on what is there in the /usr/share on the undercloud (we expect there to be  a /usr/share/openstack-octavia-images dir with a qcow in it)13:38
beaglesmarios|ruck, but the dir isn't there - afaict this works ok in a local environment (tried it twice successfully) so I'm kind of stuck13:38
beagles(it would be so much better if it also didn't work "at home")13:39
beaglesfwiw: the same ansible works fine for stein and train13:39
beaglesbut they are pretty different in how the octavia tasks are run - queens is mistral workbook "workflow tasks"  - rocky also doesn't work right but that's probably for different reasons (containerized undercloud with the ansible being run *in* the container instead of on the host IIRC)13:41
marios|ruckbeagles: not sure if that is relevant but errors in mistral log like  HeatAction.stacks.get failed: ERROR: The Stack (overcloud) could not be found.13:41
marios|ruckbeagles: have you tried asking someone from mistral team to check it13:42
marios|ruckbeagles: or if there are some known queens issue/differeences13:42
chkumar|roverweshay: thank you for the yesterday guidance on dib, it helpes, we got the issue and is selinux13:43
beaglesmarios|ruck, the heat stack action get might actually end up being relevant but the problem here is that the file check "stat" operation returns false . and AFAICT the file that I *thought* should be there (and indeed seemed to be there) isn't13:43
weshaychkumar|rover, k.. cool :)13:43
weshaymarios|ruck, chkumar|rover /me works on 8.113:43
chkumar|roverweshay: rhel8 nodepool image is 8.113:43
weshaychkumar|rover, aye.. but the qcow2 file is 8.013:44
beaglesmarios|ruck, summary - quickstart triggers a role during undercloud install that lives in oooq-extras that downloads an amphora image onto a location in the undercloud13:44
chkumar|roverweshay: if we build it with enforcing mode the error gets reproduced13:44
chkumar|roverrhel8 nodepool image has selinux enforcing13:44
beaglesmarios|ruck, this file gets picked up by workflow|external tasks during the overcloud deployment if it exists13:44
chkumar|roverin rdocloud image has selinux permissive that's why it worked13:44
chkumar|roverit was a good exercise13:45
*** Goneri has joined #oooq13:45
pandamarios|ruck: lunch.13:45
marios|ruckbeagles: could it be permissions? or selinux?13:46
marios|ruckbeagles: like it can't see the file13:46
beaglesmarios|ruck, the download seems to work but the stat in the workbook isn't working ... what it looks like to me (but how can it be) that the amphora image is being put somewhere on the undercloud other than where it says it is, or the directory is being removed before the overcloud deploy somehow or the mistral process cannot see certain directories13:46
weshaychkumar|rover, marios|ruck just FYI.. we'll need to make setting selinux a param.. because we'll use this same job internally to build images https://review.opendev.org/#/c/698883/1/roles/oooci-build-images/templates/build-images.sh.j213:46
beaglesmarios|ruck, could be...13:46
marios|ruckbeagles: not selinux https://38fc547d6dc8718028ea-253b0a19be2181811797fa6cbd0c2b8d.ssl.cf5.rackcdn.com/692579/12/check/tripleo-ci-centos-7-scenario010-multinode-oooq-container/2cf8cd5/logs/undercloud/var/log/extra/selinux.txt.gz13:46
beaglesmarios|ruck, right13:47
*** soniya29 has quit IRC13:47
pandamarios|ruck: I can chat in 3013:48
marios|ruckpanda: yes13:49
marios|ruckbeagles: commented there fyi so you can point to it and not repeat if you ask anyone else to check https://review.opendev.org/#/c/692579/1213:49
beaglesmarios|ruck, ah good idea thanks13:50
marios|ruckbeagles: not sure i can have another look see if something reveals itself to me its in my reviews list now anyway13:50
beaglesmarios|ruck, k thanks!13:50
marios|ruckbeagles: are you sure the download is completed?13:51
marios|ruckbeagles: i mean is it a sequencing thing13:51
beaglesmarios|ruck, hrmm.. think so because it's a get_url in the undercloud install13:51
beaglesmarios|ruck, so should be done long before we get to that stage of the deploy13:52
marios|ruckweshay: chkumar|rover: ack so what should we do with that? https://review.opendev.org/#/c/698883/1/roles/oooci-build-images/templates/build-images.sh.j2 and the bug should we just tidy that up for merge weshay ? with conditional? not sure that is acceptable but also not sure what else to do short of virtcustomize the image13:52
beaglesmarios|ruck, well.. the undercloud install playbook not the undercloud install itself sorry13:52
marios|ruckbeagles: ack sorry just thinking out loud/13:52
beaglesmarios|ruck, I stuck a stat after the download to make sure it actually did work... was starting to get paranoid13:53
marios|ruckbeagles: ack ;)13:53
beaglesmarios|ruck, all that stuff happens around here https://zuul.opendev.org/t/openstack/build/2cf8cd5dc0f540b59b82e726c95a7f04/log/job-output.txt#581213:54
marios|ruckpanda: sent you invite for call in half hour13:54
beaglesmarios|ruck, is there anyway that the /usr/share path might be getting "sanitized" after that playbook is run?13:54
marios|ruckpanda: RSVP its the polite thing to do13:54
marios|ruckbeagles: not that i know of13:55
beaglesmarios|ruck, I guess it would prob have to happen in overcloud prep role13:55
beaglesor playbook rather13:55
chkumar|rovermarios|ruck: heading home14:01
marios|ruckchkumar|rover: ack14:01
*** skramaja has quit IRC14:13
*** epoojad1 has joined #oooq14:13
marios|ruckchkumar|rover: fyi master very green only scen4 failed but we got it with https://review.rdoproject.org/r/24149 now just waiting on fs2014:14
*** amoralej is now known as amoralej|lunch14:15
marios|ruckjpena: still around?14:16
jpenamarios|ruck: yep14:16
rlandyjpena: hi ... wrt the naming of the downstream component dlrn .../rhel8-osp17/component/14:16
marios|ruckjpena: do you think we can hold node https://review.rdoproject.org/r/#/c/23919/ periodic-tripleo-ci-rhel-8-ovb-3ctlr_1comp-featureset001-master please14:16
marios|ruckchkumar|rover: looks like its going to timeout on image prep which is the bug we want14:16
rlandyfor rdoproject we mirror rhel and redhat, would that we possible with downstream?14:16
marios|ruckjpena: that job is currently running it should timeout in the nexthalf hour or so14:17
marios|ruckjpena: would be great if we can have the node for debu14:18
marios|ruckjpena: hmm14:18
marios|ruckjpena: how long can we hol that/?14:18
marios|ruckjpena: can it be until monday or far too much?14:18
marios|ruckjpena: cos if we can't i don't see slaweq around let me check tripelo14:18
jpenamarios|ruck: I think it won't be cleaned up automatically, so yes, it's possible14:18
marios|ruckjpena: thanks pinged slaweq also in tripleo14:20
jpenamarios|ruck: ssh zuul@38.145.34.714:20
marios|ruckjpena: ah thanks14:20
jpenarlandy: I guess you want that for the component-based pipeline?14:21
marios|ruckthanks jpena in14:21
rlandyjpena: please - as we are trying to reuse the promotion code which refers to ansible_distribution14:21
jpenarlandy: so with https://review.rdoproject.org/r/23596 that should be covered14:22
jpenaI just need to fix one little thing in the internal playbooks, and we'll have the URL too14:22
jpenathe API will be there once it's enabled (it's not at the moment)14:22
rlandyjpena: great - thanks14:23
*** TrevorV has joined #oooq14:29
marios|ruckpanda: joining now14:30
jpenarlandy: I have enabled the redhat8-* paths in the PoC VM14:33
marios|ruck*** new promoter test Queens manifest revisit - some dig updated v3/4 @ https://review.opendev.org/#/c/687267/ & posted queens tht w/depends-on @ https://review.opendev.org/69874714:36
marios|ruckpanda: ^^14:36
*** dmellado has quit IRC14:36
chkumar|rovermarios|ruck: good, then we are on track14:37
rlandyjpena: yep - I see the new dirs - thanks14:37
marios|ruckweshay: panda: ran the manifest push on queens yesterday14:38
marios|ruckand it pushed fine \o/14:38
marios|ruckpanda:++14:38
weshaymarios|ruck, panda++14:39
*** dmellado has joined #oooq14:39
weshaymarios|ruck, panda so.. then that begs the question...  when are we comfortable updating the promoter server to the latest code?14:39
weshaymarios|ruck, panda or.. maybe turning off the old promoter and running from the new one.. for a bit14:40
pandaweshay: better to run the new code on the old server14:40
weshayso we have a clean and ready known good working14:40
marios|ruckweshay: NOT NOW14:40
weshaymarios|ruck, if not now when14:40
marios|ruckif anyone touches the promoter i will 100% resort to violence14:40
weshaylolz14:40
marios|ruckweshay: we are chasing rhel8 promotion14:40
* weshay touches it..14:40
marios|ruckweshay: and master14:41
* weshay runs14:41
weshaymarios|ruck, ok..14:41
marios|ruckhttps://www.youtube.com/watch?v=h1PfrmCGFnk14:41
pandaweshay: but I'm quite confident at this point taht it would work, what it's left is being sure that jobs can use the _manifest14:41
weshaypanda, anything other steps you can think of or shall we set you free to continue thinking about and working on tests for the promoter?14:42
pandaweshay: but yes, not on friday14:42
marios|ruckpanda: well yeah it should be pretty safe as we discssed yesterday just would rather not disturb it right now we are waiting on rhel centos master14:42
weshayI guess next all the managers will be away.. so the cats can play14:42
marios|rucklike it just push _manifest but replace/remove nothing and no-one uses that _manifest anyway14:42
weshayrfolco, this is pretty good timing.. considering the ppc guys have their stuff working now14:43
pandaweshay: no, at this point, only the transition is left, then it's all improving tests so next iterations will be easier to test promoter code14:43
weshaypanda, ok looking forward to hearing your thoughts on the next iteration of tests14:45
chkumar|rovermarios|ruck: on monday, I will try to make fs01 train rhel8 green14:46
marios|ruckchkumar|rover: we should get master promotion14:47
marios|ruckchkumar|rover: centos14:47
marios|ruckat least14:47
marios|ruckchkumar|rover: maybe rhel centos monday14:47
chkumar|rovermarios|ruck: it is waiting on collectd issue14:47
chkumar|roverwe will go green on centos14:47
marios|ruckchkumar|rover: scenario1 right?14:47
chkumar|rovermarios|ruck: yes14:47
*** derekh has quit IRC14:48
marios|ruckchkumar|rover: so it reported just now https://review.rdoproject.org/zuul/builds?pipeline=openstack-periodic-master14:48
rlandyrfolco: weshay: for when component and product release differ: https://review.rdoproject.org/r/#/c/24153/14:48
*** amoralej|lunch is now known as amoralej14:48
marios|ruckchkumar|rover: damn also fs20 posting now14:48
marios|ruckchkumar|rover: https://review.rdoproject.org/r/2414914:49
marios|ruckchkumar|rover: we already have green scen4 centos from last run of that ^ so if new one passes centos master promotes14:50
chkumar|rovermarios|ruck: cool14:52
chkumar|rovermarios|ruck: for scenario 4 manila tempest bugged tbarron aready14:52
*** aakarsh has quit IRC14:52
marios|ruckchkumar|rover: where? didn't see that14:52
chkumar|rovermarios|ruck: http://logs.rdoproject.org/openstack-periodic-master/opendev.org/openstack/tripleo-ci/master/periodic-tripleo-ci-centos-7-scenario004-standalone-master/5049a8c/logs/undercloud/var/log/tempest/tempest_run.log.txt.gz14:53
marios|ruckchkumar|rover: yeah but rerun green14:53
chkumar|rovermarios|ruck: it is from recent posted log14:53
marios|ruckchkumar|rover: periodic-tripleo-ci-centos-7-scenario004-standalone-master SUCCESS in 1h 10m 20s https://review.rdoproject.org/r/#/c/24149/ v114:53
* marios|ruck feel like crap14:54
chkumar|rovermarios|ruck: ah, ok14:54
chkumar|rovermarios|ruck: new patch for scenario 1 https://review.opendev.org/#/c/69891414:59
weshayrlandy, comment added15:06
rlandyweshay: updated15:09
weshayrlandy, k. .sorry same thing here https://review.rdoproject.org/r/#/c/24153/3/roles/dlrn-report/tasks/dlrn-vars-setup.yml15:09
weshayand we'll merge15:09
rlandyyeah - sorry - saw that15:10
weshaynp15:10
* weshay is paranoid about newbies15:10
weshaypanda, oh.. before you go for the day.. do we have enough data on quay to open a ticket?15:11
* marios|ruck feeling crap going to call it in a bit15:13
weshaymarios|ruck, get out of here15:13
chkumar|rovermarios|ruck: today was a good day, we fixed one of the long issue, let's celeberate15:14
marios|ruckweshay: yeah in a bit gonna tidy up/status etc15:15
marios|ruckchkumar|rover: \o/15:15
marios|ruckchkumar|rover: another day in paradise ;)15:15
chkumar|rovermarios|ruck: :-)15:15
*** ykarel_ has joined #oooq15:16
rlandyweshay: rfolco: https://review.rdoproject.org/r/#/c/24153/ updated15:20
rlandythere is more work to do to get tripleo-ci-base-promote-component-to-current-tripleo to downstream compatible15:20
rlandybut will do that in a diff patch15:21
rfolcorlandy, just fyi I am working on the promote-hash role that can be reused for any promotion job15:21
rlandyrfolco: yeah - ok - I can remove that file15:22
rlandyif you like15:22
rlandyrfolco: just trying to keep things consistent15:23
rfolcorlandy, me too, keep what your are doing, when I get the patch ready we remove dups15:23
pandaweshay: not much else than "we are failing to push 130 images in a row, with 500 errors, and time outs."15:24
marios|ruckjpena: am about to leave not sure if slaweq had time yet15:24
marios|ruckjpena: maybe ping him in tripleo when you are leaving?15:24
marios|ruckjpena: otherwise id say just kill it15:24
jpenamarios|ruck: no worries, let's keep the vm until monday15:24
marios|ruckjpena: don't wait around for that15:24
marios|ruckjpena: ok thanks15:24
* marios|ruck shutdown sequence15:25
weshaypanda, if you can give me the raw data.. I'll handle the ticket15:25
weshaypanda, we can just email support@quay.io w/ cc me and Emilien15:26
*** rfolco is now known as rfolco|doctor15:27
*** marios|ruck is now known as marios|ruck|out15:35
EvilienMEvilien you mean15:35
*** ykarel_ is now known as ykarel|away15:39
*** marios|ruck|out has quit IRC15:44
*** epoojad1 has quit IRC15:45
*** ykarel|away has quit IRC15:46
*** jbadiapa has quit IRC15:48
*** surpatil has quit IRC15:49
chkumar|roversee ya people, have a nice weekend ahead15:53
*** chkumar|rover is now known as raukadah15:53
*** akahat has quit IRC15:54
rlandymigi: hi - so I think I have the right strings now for downstream - just struggling to get nodes to work16:03
rlandyjobs die with node_failures16:03
migirlandy: which driver are you using ?16:04
*** ykarel|away has joined #oooq16:04
rlandydriver? rhel-816:04
rlandynode type16:04
migiso it's openstack16:05
rlandyhttps://sf.hosted.upshift.rdu2.redhat.com/zuul/t/tripleo-ci-internal/status16:06
rlandyyou can see the test job ^^16:06
rlandypinged admins16:06
migirlandy: but this job is running16:12
*** fmount has quit IRC16:13
*** fmount has joined #oooq16:14
migirlandy: I can see in the logs that the upshift got issues with creating IP "No more IP addresses available on network "16:15
migirlandy: so it may be problem with upshift or some left overs from nodepool leaking16:16
migirlandy: interesting that for that tenant current usage is "Floating IPs   Allocated 5 of 400"16:16
migirlandy: maybe worth rechecking the jobs as the upshift tenant looks fine imo now16:17
rlandymigi: see #sf-ops16:19
migirlandy: http://pastebin.test.redhat.com/82186316:19
*** ykarel|away has quit IRC16:44
*** tesseract has quit IRC16:53
raukadahkopecmartin: please keep an eye on this patch https://review.opendev.org/#/c/698589/ needed for tripleo ci tempest train mess17:00
weshayraukadah, fyi https://review.rdoproject.org/r/#/c/24157/2/zuul.d/tripleoci.yaml17:12
weshayraukadah, not sure if we can test that from as it's in config17:12
raukadahweshay: what about moving that it to rdo-jobs and just call the job which uses it17:13
raukadahweshay: I will take a look on that, rlandy and I discussed about making config less bulky, may be we can do something in new year17:14
weshayraukadah, you mean creating a net new overcloud image build job in rdo-jobs that overrides the vars?17:14
raukadahweshay: I mean define the job in rdo-jobs and keep the base in config which needs secrets and rest in rdo-jobs test stuff in rdo-jobs itself17:15
weshayraukadah, +1 from me17:16
raukadahby modifying projects.yaml17:16
raukadahweshay: by the way rdo cloud rhel8 nodepool image is running on 8.117:17
raukadahwe need to talk and fix selinux issue there17:17
weshayraukadah, any tripleo job should be permissive17:19
weshayraukadah, I'll try to follow up the zuul patch and get it in rdo-jobs17:19
raukadahweshay: may be we missed something while creating rhel8 jobs , it would be a good exercise while working on centos817:20
raukadahweshay: great, will look on that on monday!17:20
raukadahfrom rhel8 I learned so manythings17:20
weshay+117:21
raukadahthanks to rlandy panda sshnaidm|off and marios17:21
raukadahweshay: you played with systemd c header files? I need to do some work on podman side17:24
raukadahpanda: rlandy ^^17:25
raukadahmy main usecase is to find out systemd is there not by calling command but from header files17:26
weshayraukadah, I have not17:34
*** dtantsur is now known as dtantsur|afk17:40
EvilienMpanda: thx for the email17:43
pandaEvilienM: I'm afraid all the updates will be sent to me, I'll see if I can add your emails too, but at least you have the internal ticket number17:45
raukadahweshay: np, once find the solution, will learn together17:48
*** amoralej is now known as amoralej|off17:51
*** jpena is now known as jpena|off17:53
*** holser has quit IRC18:09
mjturekweshay: so baha and I are hitting a new error. Suddenly skydive-base is matching and failing to build with no output as to why :(18:11
mjturekweshay do you know where we find the regex that matches containers?18:13
weshaymjturek, it's all in the kolla config18:14
mjturekcool thanks18:14
mjtureklike should be in kolla-build.conf18:15
weshayhttps://review.opendev.org/#/c/693390/2/docker/nova/nova-base/Dockerfile.j2 for example18:17
weshaymjturek, I think skydive though should be excluded from the build18:17
weshayand maybe why ur hitting an error18:18
mjturekweshay it definitely should for poer18:18
mjturekpower*18:18
weshayoh..18:18
*** soniya29 has joined #oooq18:18
* mjturek has a stuck w key18:18
weshayhttp://logs.rdoproject.org/openstack-periodic-master/opendev.org/openstack/tripleo-ci/master/periodic-tripleo-centos-7-master-containers-build-push/fd6135a/logs/containers-successfully-built.log.txt.gz18:19
weshayya.. I guess we do build it18:19
weshay-agent analyzer and base18:20
weshayI have to meet a buddy for lunch.. bbiab18:20
weshaysorry to bail18:20
mjtureknp ttyl!18:21
rlandyweshay: is there any way to work out the dlrn end point of we don;t know if off hand?18:36
*** yolanda has quit IRC18:43
*** jfrancoa has quit IRC18:48
*** saneax has quit IRC18:48
*** tosky has quit IRC18:57
rlandyraukadah: ^^ if you are around, do you know the dlrn api for downstream?19:09
*** soniya29 has quit IRC19:11
EvilienMpanda: no prob19:21
EvilienMthanks for taking care of it19:21
rlandynvm found it19:21
*** jtomasek has quit IRC19:31
*** Goneri has quit IRC20:27
*** rfolco|doctor is now known as rfolco20:31
weshayrlandy, the dlrn_api client in the prod dlrn server is throwing 500's too20:51
weshayhttp://promoter.rdoproject.org/centos7_train.log20:52
*** EvilienM is now known as EmilienM20:57
*** TrevorV has quit IRC20:58
rlandyweshay; ok  - will chat with infar guys on monday21:10
weshayaye21:10
*** holser has joined #oooq21:10
*** rlandy has quit IRC21:10
*** holser has quit IRC21:35
*** holser has joined #oooq22:10
*** rfolco has quit IRC22:19
*** holser has quit IRC22:26
*** Goneri has joined #oooq22:34
*** rfolco has joined #oooq22:50
*** rfolco has quit IRC22:55
*** rfolco has joined #oooq23:49

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