Monday, 2016-08-01

*** rossella_s has quit IRC00:00
*** rossella_s has joined #openstack-infra00:00
*** baoli has joined #openstack-infra00:01
openstackgerritYAMAMOTO Takashi proposed openstack-infra/project-config: networking-midonet: Add periodic jobs for stable branches
*** markvoelker has joined #openstack-infra00:12
*** amitgandhinz has joined #openstack-infra00:13
*** sdake has joined #openstack-infra00:13
*** dingyichen has quit IRC00:16
*** tonytan4ever has joined #openstack-infra00:16
*** dingyichen has joined #openstack-infra00:16
*** markvoelker has quit IRC00:17
*** sdake has quit IRC00:17
*** amitgandhinz has quit IRC00:17
*** achanda has quit IRC00:20
*** tonytan4ever has quit IRC00:21
*** kzaitsev_mb has joined #openstack-infra00:23
*** chlong has joined #openstack-infra00:26
openstackgerritCraige McWhirter proposed openstack-infra/puppet-phabricator: Patches Required to Deliver Pholio
*** kzaitsev_mb has quit IRC00:28
*** chlong has quit IRC00:30
*** baoli has quit IRC00:40
*** chlong has joined #openstack-infra00:41
*** baoli has joined #openstack-infra00:45
*** tqtran has joined #openstack-infra00:46
*** thorst has joined #openstack-infra00:50
*** tqtran has quit IRC00:51
ianwgotta love a bit of yak shaving debugging the debugger before you can debug the problem00:51
*** thorst has quit IRC00:59
*** baoli has quit IRC00:59
*** esberglu has joined #openstack-infra01:02
*** hichihara has joined #openstack-infra01:02
*** julim has joined #openstack-infra01:05
*** esberglu has quit IRC01:06
*** zhurong has joined #openstack-infra01:09
prometheanfireanyone mind looking at this diskimage builder change?
*** chlong has quit IRC01:10
prometheanfirealso, a fix for install-puppet on gentoo in system-config ...01:12
prometheanfirethat one is a one line change01:12
*** markvoelker has joined #openstack-infra01:13
prometheanfireI'll leave the project-config one out of it for now :P01:13
*** amitgandhinz has joined #openstack-infra01:13
*** markvoelker has quit IRC01:17
*** tonytan4ever has joined #openstack-infra01:18
*** amitgandhinz has quit IRC01:18
*** tonytan4ever has quit IRC01:22
*** yanyanhu has joined #openstack-infra01:22
*** chlong has joined #openstack-infra01:24
*** kzaitsev_mb has joined #openstack-infra01:24
*** chlong has quit IRC01:28
*** kzaitsev_mb has quit IRC01:30
*** chlong has joined #openstack-infra01:30
*** lucasagomes has quit IRC01:32
*** mriedem has quit IRC01:38
*** lucasagomes has joined #openstack-infra01:39
*** rkukura has joined #openstack-infra01:40
openstackgerritYanyan Hu proposed openstack-infra/project-config: Add integration test job for Senlin
*** gongysh has joined #openstack-infra01:44
*** aysyd has joined #openstack-infra01:44
*** aysyd has quit IRC01:47
*** gildub has joined #openstack-infra01:48
*** sputnik13 has quit IRC02:01
*** thorst has joined #openstack-infra02:04
*** gongysh has quit IRC02:07
openstackgerritCraige McWhirter proposed openstack-infra/puppet-phabricator: Patches Required to Deliver Pholio
*** amitgandhinz has joined #openstack-infra02:14
*** thorst has quit IRC02:18
*** amitgandhinz has quit IRC02:19
*** Liuqing has joined #openstack-infra02:32
*** liusheng has joined #openstack-infra02:39
prometheanfireianw: hope that explains it02:41
*** zhurong_ has joined #openstack-infra02:43
*** Liuqing_ has joined #openstack-infra02:43
*** Liuqing has quit IRC02:46
*** zhurong has quit IRC02:46
*** tonytan4ever has joined #openstack-infra02:52
*** gongysh has joined #openstack-infra02:59
*** baoli has joined #openstack-infra03:01
*** bhunter71 has quit IRC03:05
*** baoli has quit IRC03:06
*** yuanying has quit IRC03:08
*** roxanaghe has joined #openstack-infra03:10
*** amotoki has joined #openstack-infra03:11
*** amitgandhinz has joined #openstack-infra03:15
*** thorst has joined #openstack-infra03:19
*** amitgandhinz has quit IRC03:19
*** bhunter71 has joined #openstack-infra03:20
*** Krenair has quit IRC03:24
*** Krenair has joined #openstack-infra03:27
*** thorst has quit IRC03:27
*** rbuzatu has joined #openstack-infra03:27
*** amotoki has quit IRC03:28
*** karthikp_ has joined #openstack-infra03:29
*** Teresita-Warrior has joined #openstack-infra03:31
*** karthikp_ has quit IRC03:32
*** coolsvap has joined #openstack-infra03:34
*** karthikp_ has joined #openstack-infra03:37
*** amitkqed has quit IRC03:38
*** amitkqed has joined #openstack-infra03:38
*** EricGonczer_ has quit IRC03:41
*** tqtran has joined #openstack-infra03:47
*** rbuzatu has quit IRC03:50
*** Teresita-Warrior has quit IRC03:52
*** tqtran has quit IRC03:52
*** Teresita-Warrior has joined #openstack-infra03:52
*** rbuzatu has joined #openstack-infra03:52
*** achanda has joined #openstack-infra03:54
*** roxanaghe has quit IRC03:54
*** amotoki has joined #openstack-infra03:57
*** Teresita-Warrior has quit IRC03:57
*** gongysh has quit IRC03:58
*** achanda has quit IRC04:00
*** ljjjustin has quit IRC04:03
*** amotoki_ has joined #openstack-infra04:06
*** ljjjustin has joined #openstack-infra04:06
*** amotoki has quit IRC04:09
*** chlong has quit IRC04:15
*** amitgandhinz has joined #openstack-infra04:16
*** amotoki_ has quit IRC04:17
*** yamahata has joined #openstack-infra04:18
*** roxanaghe has joined #openstack-infra04:19
*** amitgandhinz has quit IRC04:20
*** thorst has joined #openstack-infra04:27
*** amotoki has joined #openstack-infra04:31
*** chlong has joined #openstack-infra04:32
*** thorst has quit IRC04:33
openstackgerritCraige McWhirter proposed openstack-infra/infra-specs: Initial commit of the Pholio spec
*** gouthamr has quit IRC04:35
*** bhunter71 has quit IRC04:37
*** bhunter71 has joined #openstack-infra04:38
*** rbuzatu has quit IRC04:42
openstackgerritCraige McWhirter proposed openstack-infra/infra-specs: Initial commit of the Pholio spec
*** bhunter71 has quit IRC04:47
*** abregman has joined #openstack-infra04:47
*** psachin has joined #openstack-infra04:48
*** psachin has quit IRC04:48
*** rbuzatu has joined #openstack-infra04:49
*** achanda has joined #openstack-infra04:57
*** Jeffrey4l has joined #openstack-infra04:57
openstackgerritChristian Fetzer proposed openstack-infra/jenkins-job-builder: gitlab trigger: Update Gitlab to >=1.3.0
openstackgerritYanyan Hu proposed openstack-infra/project-config: Add integration test job for Senlin
*** Srinu has quit IRC05:03
*** ianychoi has joined #openstack-infra05:10
*** psachin has joined #openstack-infra05:13
*** amitgandhinz has joined #openstack-infra05:16
openstackgerritSteve Baker proposed openstack-infra/project-config: Configure tempest in gate-heat-dsvm-functional-*
*** roxanaghe has quit IRC05:20
*** amitgandhinz has quit IRC05:21
*** pahuang has quit IRC05:22
*** tonytan4ever has quit IRC05:25
*** rbuzatu has quit IRC05:25
*** karthikp_ has quit IRC05:26
*** e0ne has joined #openstack-infra05:27
*** nwkarsten has joined #openstack-infra05:28
*** thorst has joined #openstack-infra05:29
*** nwkarsten has quit IRC05:32
*** thorst has quit IRC05:37
*** swat30 has quit IRC05:37
*** e0ne has quit IRC05:41
*** e0ne has joined #openstack-infra05:44
*** e0ne has quit IRC05:47
*** tqtran has joined #openstack-infra05:49
*** e0ne has joined #openstack-infra05:51
*** e0ne has quit IRC05:51
*** tqtran has quit IRC05:53
*** javeriak has joined #openstack-infra06:16
*** amitgandhinz has joined #openstack-infra06:17
*** snarwade has joined #openstack-infra06:18
*** snarwade has quit IRC06:18
*** snarwade has joined #openstack-infra06:18
*** ramishra has joined #openstack-infra06:20
*** tqtran has joined #openstack-infra06:20
*** roxanaghe has joined #openstack-infra06:21
*** amitgandhinz has quit IRC06:22
*** tqtran has quit IRC06:24
*** roxanaghe has quit IRC06:25
*** tonytan4ever has joined #openstack-infra06:26
*** tkelsey has joined #openstack-infra06:29
*** tonytan4ever has quit IRC06:31
*** pgadiya has joined #openstack-infra06:33
*** tkelsey has quit IRC06:33
*** Cormite has joined #openstack-infra06:34
*** ihrachys has joined #openstack-infra06:36
*** rbuzatu has joined #openstack-infra06:41
*** slaweq has joined #openstack-infra06:43
*** Guest38666 has quit IRC06:44
*** tesseract- has joined #openstack-infra06:44
*** camunoz has quit IRC06:45
*** nmagnezi has joined #openstack-infra06:47
*** daemontool_ has joined #openstack-infra06:53
*** javeriak has quit IRC06:54
*** ramishra has quit IRC06:54
*** ihrachys has quit IRC06:55
*** chlong has quit IRC06:59
yolandagood morning06:59
*** Na3iL has joined #openstack-infra06:59
*** javeriak has joined #openstack-infra07:02
*** dimtruck is now known as zz_dimtruck07:04
*** jed56 has joined #openstack-infra07:04
*** zz_dimtruck is now known as dimtruck07:05
*** ramishra has joined #openstack-infra07:06
*** slagle has quit IRC07:06
*** ihrachys has joined #openstack-infra07:06
*** pgadiya has quit IRC07:07
*** bhavik has joined #openstack-infra07:08
*** vsaienko has quit IRC07:08
*** psanchez has quit IRC07:09
*** stevebaker has quit IRC07:09
*** flyingtt has joined #openstack-infra07:10
*** ihrachys_ has joined #openstack-infra07:11
*** chlong has joined #openstack-infra07:11
*** ihrachys has quit IRC07:12
*** Cormite has quit IRC07:14
*** dimtruck is now known as zz_dimtruck07:16
*** ccamacho has joined #openstack-infra07:17
openstackgerritMerged openstack-infra/project-config: Add new project vmware-nsxlib
openstackgerritYanyan Hu proposed openstack-infra/project-config: Add integration test job for Senlin
*** amitgandhinz has joined #openstack-infra07:18
openstackgerritMerged openstack-infra/project-config: Make ironicclient functional job voting
*** pgadiya has joined #openstack-infra07:19
AJaegermorning, yolanda !07:19
snarwadehello there, can I use if-else condition in JJB ?07:19
*** thorst has joined #openstack-infra07:20
*** amitgandhinz has quit IRC07:22
*** rbuzatu has quit IRC07:23
*** rbuzatu has joined #openstack-infra07:24
AJaegersnarwade: that question is too generic to give a yes or now answer ;(07:25
*** mikelk has joined #openstack-infra07:26
AJaegerJJB is YAML and has no if-else. You can use bash in scripts and those can use if-else.07:26
*** tonytan4ever has joined #openstack-infra07:27
openstackgerritYanyan Hu proposed openstack-infra/project-config: Add integration test job for Senlin
*** thorst has quit IRC07:29
*** claudiub has joined #openstack-infra07:30
openstackgerritafazekas proposed openstack/os-testr: Construct a list of test cases instead of passing a regexp
*** tonytan4ever has quit IRC07:32
*** vincentll has joined #openstack-infra07:33
*** Hal has joined #openstack-infra07:34
*** chlong has quit IRC07:34
*** zz_dimtruck is now known as dimtruck07:34
*** Hal is now known as Guest4053807:34
*** jlanoux has joined #openstack-infra07:35
*** Sukhdev has joined #openstack-infra07:37
*** pahuang has joined #openstack-infra07:37
*** ihrachys_ has quit IRC07:41
*** abregman is now known as abregman|sync07:41
*** dimtruck is now known as zz_dimtruck07:44
*** Sukhdev has quit IRC07:46
*** ihrachys has joined #openstack-infra07:47
*** kzaitsev_mb has joined #openstack-infra07:49
*** ihrachys has quit IRC07:50
*** Cormite has joined #openstack-infra07:53
*** gildub has quit IRC07:54
openstackgerritMerged openstack-infra/project-config: Move {name}-publish-specs job + others to ubuntu-xenial
openstackgerritMerged openstack-infra/project-config: Swap neutron/nova-net multinode full jobs in nova check queue
openstackgerritMerged openstack-infra/project-config: Set up stable/newton temporary ACLs
openstackgerritMerged openstack-infra/project-config: Keep localrc in CLI Sahara job
*** ihrachys has joined #openstack-infra07:56
*** e0ne has joined #openstack-infra07:57
*** abregman|sync is now known as abregman07:58
*** zzzeek has quit IRC08:00
*** rossella_s has quit IRC08:00
*** zzzeek has joined #openstack-infra08:00
*** rossella_s has joined #openstack-infra08:00
*** jlanoux has quit IRC08:01
*** jaosorior has joined #openstack-infra08:03
*** achanda_ has joined #openstack-infra08:05
*** achanda has quit IRC08:07
*** e0ne has quit IRC08:07
*** jlanoux has joined #openstack-infra08:08
*** yuanying has joined #openstack-infra08:08
*** roxanaghe has joined #openstack-infra08:09
*** shardy has joined #openstack-infra08:11
openstackgerritOpenStack Proposal Bot proposed openstack-infra/project-config: Normalize projects.yaml
*** rbuzatu has joined #openstack-infra08:13
*** roxanaghe has quit IRC08:14
*** e0ne has joined #openstack-infra08:15
*** asettle has joined #openstack-infra08:18
*** amitgandhinz has joined #openstack-infra08:19
*** tkelsey has joined #openstack-infra08:20
*** amitgandhinz has quit IRC08:23
*** ggnel_t has joined #openstack-infra08:25
*** yamamoto has quit IRC08:26
*** thorst has joined #openstack-infra08:26
*** e0ne has quit IRC08:28
*** dingyichen has quit IRC08:30
*** ihrachys has quit IRC08:31
*** javeriak has quit IRC08:31
*** thorst has quit IRC08:33
*** tkelsey has quit IRC08:33
apuimedolifeless: dims: please, review
openstackgerritTatiana Ovchinnikova proposed openstack-infra/project-config: Add Django 1.9 and 1.10 tests to Manila UI
*** e0ne has joined #openstack-infra08:40
AJaegerapuimedo: please discuss on #openstack-release channel any releases08:40
*** ramishra has quit IRC08:41
*** vincentll has quit IRC08:43
apuimedoAJaeger: sorry. I was not aware of that channel :/08:43
apuimedogoing there now. Thanks08:43
*** vincentll has joined #openstack-infra08:44
openstackgerritMerged openstack-infra/project-config: Normalize projects.yaml
*** ramishra has joined #openstack-infra08:47
*** chem` has joined #openstack-infra08:48
*** pgadiya has quit IRC08:49
*** vincentll has quit IRC08:50
*** abregman is now known as abregman|mtg08:51
*** achanda_ has quit IRC08:52
*** vincentll has joined #openstack-infra08:52
*** r-mibu has quit IRC08:56
*** r-mibu has joined #openstack-infra08:56
*** roxanaghe has joined #openstack-infra09:00
*** javeriak has joined #openstack-infra09:01
*** degorenko has joined #openstack-infra09:03
snarwadeAJaeger, thanks09:04
*** pgadiya has joined #openstack-infra09:04
*** pgadiya has quit IRC09:05
*** roxanaghe has quit IRC09:06
*** pgadiya has joined #openstack-infra09:06
*** ramishra has quit IRC09:06
*** zz_dimtruck is now known as dimtruck09:07
*** ramishra has joined #openstack-infra09:07
*** sambetts|afk is now known as sambetts09:11
*** eric_lopez has joined #openstack-infra09:13
*** hichihara has quit IRC09:13
*** matrohon has joined #openstack-infra09:13
*** kzaitsev_mb has quit IRC09:14
*** elo has quit IRC09:16
*** Goneri has joined #openstack-infra09:16
*** dimtruck is now known as zz_dimtruck09:17
*** HeOS has joined #openstack-infra09:18
*** acoles_ is now known as acoles09:19
*** amitgandhinz has joined #openstack-infra09:19
openstackgerritSam Betts proposed openstack-infra/project-config: Ensure IRONIC_BUILD_DEPLOY_RAMDISK is always explict
*** dtantsur|pto is now known as dtantsur09:23
*** hparekh has quit IRC09:24
*** amitgandhinz has quit IRC09:24
*** acoles is now known as acoles_09:25
*** cbader has quit IRC09:25
*** cbader has joined #openstack-infra09:26
*** asettle has quit IRC09:26
*** asettle has joined #openstack-infra09:27
*** tonytan4ever has joined #openstack-infra09:28
*** thorst has joined #openstack-infra09:31
flyingttHi, all, how to filter the statistics info between two dates with CLI stackalytics-processor?09:32
*** notnownikki has joined #openstack-infra09:32
flyingttI know the stackalytics's API has already supported this feature.09:32
*** tonytan4ever has quit IRC09:33
*** acoles_ is now known as acoles09:36
*** hparekh has joined #openstack-infra09:37
*** thorst has quit IRC09:38
*** matrohon has quit IRC09:39
*** ramishra has quit IRC09:40
*** ramishra has joined #openstack-infra09:41
*** acoles is now known as acoles_09:50
*** achanda has joined #openstack-infra09:52
*** yamamoto has joined #openstack-infra09:53
*** sshnaidm|afk is now known as sshnaidm09:54
*** rodrigods has quit IRC09:55
*** acoles_ is now known as acoles09:55
*** rodrigods has joined #openstack-infra09:55
*** tosky has joined #openstack-infra09:55
*** yamamoto has quit IRC09:57
*** achanda has quit IRC09:58
openstackgerritGiulio Fidente proposed openstack-infra/tripleo-ci: Add 'ceph status' and tripleo-overcloud-passwords to host_info
*** Na3iL has quit IRC10:02
ianwpabelanger: interesting findings (to me) in about ntpd.  basically we need the ntpdate service (which doesn't call ntpdate) to pre-set the time to be sure to avoid long timeouts of ntp-wait.  or, switch to chronyd10:03 bug 1361382 in ntp "ntp-wait hangs after boot for a long time, unless ntpd is restarted" [Unspecified,Closed: notabug] - Assigned to mlichvar10:03
zigoianw: FYI, you can use "ntpstat" (which I maintain in Debian) to check for the accuracy of the clock.10:05
*** rbuzatu has quit IRC10:05
zigoVery useful.10:05
*** zhurong_ has quit IRC10:07
ianwzigo: cool ... yeah in this case i don't think it would help with the timeouts, it seems like it's expected behavior.  i think ntpd just isn't good for short-lived vms10:12
*** yamamoto has joined #openstack-infra10:13
*** Liuqing_ has quit IRC10:13
*** ramishra has quit IRC10:15
*** yamamoto has quit IRC10:16
*** yamamoto has joined #openstack-infra10:18
*** amitgandhinz has joined #openstack-infra10:20
AJaegerianw, if you still have time: could you review my changes,, and , please?10:21
openstackgerritAlistair Coles proposed openstack-infra/project-config: Add swift job tox-func-in-process-encryption
*** yanyanhu has quit IRC10:22
*** yamamoto has quit IRC10:24
*** amitgandhinz has quit IRC10:25
*** ramishra has joined #openstack-infra10:26
openstackgerritVasyl Saienko proposed openstack-infra/project-config: Add networking-generic-switch-dsvm-temepst job.
*** vsaienko has joined #openstack-infra10:28
vsaienkoinfra team please review, the patch adds a non-voting job for networking-generic-switch, that launch tempest tests, we are going to switch to tempest based tests instead of shell script.10:29
*** rbuzatu has joined #openstack-infra10:30
openstackgerritJesse Pretorius (odyssey4me) proposed openstack-infra/project-config: Implement Xenial/CentOS jobs for OpenStack-Ansible
*** abregman|mtg is now known as abregman10:33
*** john-davidge has joined #openstack-infra10:36
*** thorst has joined #openstack-infra10:36
*** john-dav_ has joined #openstack-infra10:38
*** john-davidge has quit IRC10:41
*** thorst has quit IRC10:44
*** oanson has joined #openstack-infra10:46
*** rbuzatu has quit IRC10:48
*** yamamoto has joined #openstack-infra10:54
openstackgerritafazekas proposed openstack/os-testr: Remove construct_regex function
*** bhavik has quit IRC10:57
*** javeriak has quit IRC10:59
*** ramishra has quit IRC11:01
*** ramishra has joined #openstack-infra11:04
*** ramishra has quit IRC11:04
*** ramishra has joined #openstack-infra11:05
*** zz_dimtruck is now known as dimtruck11:08
*** john-dav_ is now known as john-davidge11:11
*** Goneri has quit IRC11:11
*** rhallisey has joined #openstack-infra11:13
openstackgerritMerged openstack-infra/project-config: Move doc team jobs to Xenial
*** Goneri has joined #openstack-infra11:16
*** dimtruck is now known as zz_dimtruck11:18
*** yamamoto has quit IRC11:19
*** yamamoto has joined #openstack-infra11:19
*** jcoufal has joined #openstack-infra11:20
*** rbuzatu has joined #openstack-infra11:20
*** amitgandhinz has joined #openstack-infra11:21
*** sdague has joined #openstack-infra11:24
AJaegerianw: thanks for reviews. Regarding nodepool: that's AFAIK just the name, the jobs are all renamed - I just checked. But perhaps I missed something...11:25
*** amitgandhinz has quit IRC11:25
*** gildub has joined #openstack-infra11:26
*** mat128|pto is now known as mat12811:27
*** jkilpatr has joined #openstack-infra11:29
*** pblaho has joined #openstack-infra11:30
*** slagle has joined #openstack-infra11:31
openstackgerritafazekas proposed openstack/os-testr: Remove construct_regex function
openstackgerritafazekas proposed openstack/os-testr: Construct a list of test cases instead of passing a regexp
*** jcoufal has quit IRC11:35
AJaegerpabelanger: could you double check later ianw's comment on - and review as well, please? Thanks!11:35
*** jcoufal has joined #openstack-infra11:35
*** javeriak has joined #openstack-infra11:37
*** dprince has joined #openstack-infra11:38
*** edmondsw has joined #openstack-infra11:40
*** aviau has quit IRC11:43
*** aviau has joined #openstack-infra11:43
*** weshay has joined #openstack-infra11:44
*** yamamoto has quit IRC11:46
*** yamamoto has joined #openstack-infra11:46
openstackgerritafazekas proposed openstack/os-testr: Remove construct_regex function
openstackgerritafazekas proposed openstack/os-testr: Construct a list of test cases instead of passing a regexp
*** zhenguo has quit IRC11:51
*** chem` is now known as chem11:53
openstackgerritOleksii Zamiatin proposed openstack-infra/project-config: Make pika jobs voting
*** ldnunes has joined #openstack-infra11:54
*** tpsilva has joined #openstack-infra11:54
openstackgerritJuan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: TEST IRONIC SSL - DO NOT MERGE
*** javeriak has quit IRC11:55
openstackgerritVasyl Saienko proposed openstack-infra/project-config: Add networking-generic-switch-dsvm-temepst job.
*** javeriak has joined #openstack-infra11:55
*** achanda has joined #openstack-infra11:56
*** asettle has quit IRC11:56
*** yottatsa has joined #openstack-infra11:57
*** yottatsa has quit IRC11:58
*** yottatsa has joined #openstack-infra11:58
*** pt_15 has joined #openstack-infra12:00
*** itisha has joined #openstack-infra12:00
openstackgerritMartin Mágr proposed openstack-infra/tripleo-ci: Add opstools candidate repo
*** achanda has quit IRC12:01
*** sdake_ has joined #openstack-infra12:02
*** asettle has joined #openstack-infra12:02
*** psilvad has joined #openstack-infra12:03
*** Ravikiran_K has joined #openstack-infra12:03
*** aysyd has joined #openstack-infra12:04
*** amotoki has quit IRC12:05
*** raildo has joined #openstack-infra12:05
*** thorst has joined #openstack-infra12:05
*** asettle has quit IRC12:06
*** asettle has joined #openstack-infra12:07
*** jcoufal_ has joined #openstack-infra12:07
openstackgerritafazekas proposed openstack/os-testr: Allow to specifiy black/white/regex at the same time
*** jcoufal has quit IRC12:10
*** javeriak has quit IRC12:10
*** nwkarsten has joined #openstack-infra12:11
*** akshai has joined #openstack-infra12:12
*** kgiusti has joined #openstack-infra12:15
*** pradk has joined #openstack-infra12:17
*** rbuzatu has quit IRC12:19
*** gouthamr has joined #openstack-infra12:19
*** amotoki has joined #openstack-infra12:21
openstackgerritsandro.gattuso proposed openstack-infra/jenkins-job-builder: Wrapper pre-scm-buildstep: added support for failOnError parameter
*** amitgandhinz has joined #openstack-infra12:22
*** rbuzatu has joined #openstack-infra12:23
*** amitgandhinz has quit IRC12:26
*** rvasilets has joined #openstack-infra12:26
*** yottatsa has quit IRC12:27
*** markvoelker has joined #openstack-infra12:27
*** rfolco_ has joined #openstack-infra12:27
*** amotoki has quit IRC12:33
*** scottda_vacation is now known as scottda12:33
rvasiletsAJaeger, Hi, earlier Rally have tried to switch there jobs to keystone v3 by default. See the patch By for now we've got still v2 by default proof Possibly this is because of Could you help to fix that issue? Thank you12:33
openstackgerritSam Betts proposed openstack-infra/project-config: Ensure IRONIC_BUILD_DEPLOY_RAMDISK is always explict
AJaegerrvasilets: Sorry, cannot help on that. You might want to see what's done elsewhere with keystone v3 like  in
AJaegerwrong link, let me give you somethign better12:37
*** ttx has quit IRC12:38
odyssey4meany idea who reviews patches to bindep? we need some review love for
*** rlandy has joined #openstack-infra12:39
AJaegerodyssey4me: all the infra-roots do. Let me share my +/-1 love with it ;)12:39
rvasiletsAJaeger, Thx, as I understand I need to do the same work for Rally. Right?12:40
*** aysyd has quit IRC12:40
*** julim has quit IRC12:40
odyssey4meAJaeger thanks, this is causing an issue with one of our patches which is trying to address SSL SNI failures12:41
AJaegerrvasilets: I don't know all the details, you might need to check with devstack experts...12:42
*** amotoki has joined #openstack-infra12:43
rvasiletsAJaeger, Sure, thank you12:43
*** berendt has joined #openstack-infra12:43
*** sileht has quit IRC12:44
*** sileht has joined #openstack-infra12:46
*** javeriak has joined #openstack-infra12:46
*** rbuzatu has quit IRC12:46
AJaegerodyssey4me: once that change is in, we need a bindep release as well. infra-root can help you with both.12:47
odyssey4methanks AJaeger12:47
*** amotoki has quit IRC12:48
*** sdake_ is now known as sdake12:48
*** Liuqing has joined #openstack-infra12:52
*** jsavak has joined #openstack-infra12:57
*** mikelk has quit IRC12:59
*** oanson has quit IRC12:59
*** mtanino has joined #openstack-infra13:00
*** pgadiya has quit IRC13:01
*** abregman is now known as abregman|mtg13:02
*** mikelk has joined #openstack-infra13:02
*** ccamacho is now known as ccamacho|lunch13:05
*** xyang1 has joined #openstack-infra13:07
*** Liuqing has quit IRC13:07
*** amitgandhinz has joined #openstack-infra13:07
*** baoli has joined #openstack-infra13:08
*** zz_dimtruck is now known as dimtruck13:09
*** mdrabe has joined #openstack-infra13:09
*** baoli_ has joined #openstack-infra13:10
*** bryan_att has joined #openstack-infra13:10
*** krtaylor has joined #openstack-infra13:12
*** EricGonczer_ has joined #openstack-infra13:13
*** matrohon has joined #openstack-infra13:13
*** Guest40538 has quit IRC13:13
*** amotoki has joined #openstack-infra13:13
*** baoli has quit IRC13:14
*** Hal has joined #openstack-infra13:15
*** matt-borland has joined #openstack-infra13:15
*** tonytan4ever has joined #openstack-infra13:15
*** Hal is now known as Guest918813:15
*** abregman|mtg is now known as abregman13:16
*** yottatsa has joined #openstack-infra13:16
*** dimtruck is now known as zz_dimtruck13:18
*** tonytan4ever has quit IRC13:19
*** amitgandhinz has quit IRC13:20
*** amitgandhinz has joined #openstack-infra13:20
*** _ari_ has joined #openstack-infra13:21
*** matrohon has quit IRC13:21
robcresswellAnyone know if there happens to be an openstack link shortener? Gerrit links are a pain but I know some people dislike link shorteners.13:24
*** julim has joined #openstack-infra13:24
robcresswellWould be nice to have a or something :)13:24
*** julim has quit IRC13:26
*** zz_dimtruck is now known as dimtruck13:28
openstackgerritMerged openstack-infra/tripleo-ci: Test overcloud deletion in periodic job
*** julim has joined #openstack-infra13:28
*** nwkarsten has quit IRC13:29
*** nwkarsten has joined #openstack-infra13:29
*** EricGonczer_ has quit IRC13:31
*** Guest9188 has quit IRC13:32
*** yamamoto has quit IRC13:32
openstackgerritafazekas proposed openstack/os-testr: Allow to specifiy black/white list at the same time
openstackgerritafazekas proposed openstack/os-testr: Add --black-regex/-B option
openstackgerritJuan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: Implement a new job for undercloud testing only
openstackgerritJuan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: Use delorean-current for puppet modules
openstackgerritJuan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: WIP - Implement undercloud upgrade job - Mitaka -> Newton
*** nwkarsten has quit IRC13:34
rcarrillocruzrobcresswell: that sounds like a nice toy project for infra :D13:34
*** tonytan4ever has joined #openstack-infra13:34
*** yamamoto has joined #openstack-infra13:34
*** ayoung has joined #openstack-infra13:36
robcresswellrcarrillocruz: I was just curious if something happened to exist already. Wasn't trying to get someone to do it :)13:37
*** matrohon has joined #openstack-infra13:37
AJaegerrobcresswell: I'm not aware of any13:37
*** zul_ has quit IRC13:38
*** Na3iL has joined #openstack-infra13:39
*** ccamacho|lunch is now known as ccamacho13:39
*** tonytan_brb has joined #openstack-infra13:39
rcarrillocruzrobcresswell: feel free to add it to the infra agenda if you like13:39
*** javeriak has quit IRC13:39
*** yamamoto has quit IRC13:40
*** tonytan4ever has quit IRC13:41
*** zhurong has joined #openstack-infra13:43
*** mriedem has joined #openstack-infra13:43
*** zhurong has quit IRC13:43
*** zul has joined #openstack-infra13:43
mtreinishfungi, rcarrillocruz, yolanda: if you get a sec: is almost ready to start deploying. Just need to land the germqtt changes (then I'll push a tag) and the system config change should be good to go13:43
*** yfried has joined #openstack-infra13:45
*** matrohon has quit IRC13:46
*** yfried has left #openstack-infra13:47
*** rbuzatu has joined #openstack-infra13:47
*** bhunter71 has joined #openstack-infra13:47
*** nwkarste_ has joined #openstack-infra13:48
yolandamtreinish, i'll take a look later13:48
fungirobcresswell: we've discussed it, consensus seemed to be that as long as it can only shorten links for specific domains under our control it's probably safe from abuse by spammers13:49
*** nwkarste_ has quit IRC13:49
openstackgerritKien Ha proposed openstack-infra/jenkins-job-builder: gitlab trigger: Support new "trigger-open-merge-request-push" options
mtreinishyolanda: thanks13:49
*** nwkarste_ has joined #openstack-infra13:49
fungirobcresswell: so basically an openstack community link shortener that can only shorten links for subdomain sites13:50
AJaegermorning fungi! May I push some reviews on your queue for constraints, please?13:51
fungiAJaeger: i haven't really caught up yet this morning, but i can try13:52
*** rbuzatu has quit IRC13:52
AJaegerfungi, - venv usage. Needs +A and we should tell dhellmann about it.13:52
fungiahh, right, that one13:52
*** mmandel has joined #openstack-infra13:52
AJaegerfungi: I can self-A, just wanted to get your go-ahead...13:52
AJaegeryolanda asked for you to ACK13:52
yolandai didn't have time to babysit, and i prefered to wait for fungi13:54
openstackgerritKien Ha proposed openstack-infra/jenkins-job-builder: gitlab trigger: Support new "trigger-open-merge-request-push" options
AJaegeryolanda: that's perfectly fine ;)13:55
*** mriedem is now known as mriedem_meeting13:56
*** dimtruck is now known as zz_dimtruck13:57
*** mestery has quit IRC13:58
openstackgerritMerged openstack-infra/germqtt: Add MQTT auth support
*** achanda has joined #openstack-infra13:59
*** zhurong has joined #openstack-infra13:59
openstackgerritKien Ha proposed openstack-infra/jenkins-job-builder: gitlab trigger: Support new "trigger-open-merge-request-push" options
*** mestery has joined #openstack-infra14:00
openstackgerritMatt Borland proposed openstack-infra/project-config: WIP: Splitting current/deprecated integ. test
*** rbrndt has joined #openstack-infra14:01
openstackgerritMatt Borland proposed openstack-infra/project-config: Splitting current/deprecated integ. test
fungiyep, i'm happy to approve, just trying to dig my way out of the morning pile of new information so don't want people to block on me to get stuff done14:02
*** baoli_ has quit IRC14:02
*** matrohon has joined #openstack-infra14:02
AJaegerthanks, fungi. No need to do it *now*.14:02
*** baoli has joined #openstack-infra14:03
AJaegerfungi, FYI: I'm pushing constraints since glance_store in liberty has failures in docs jobs that we should be able to fix with constraints.14:03
*** jamesmcarthur has quit IRC14:03
*** ildikov has quit IRC14:03
*** sridhar_ram has quit IRC14:04
*** mattfarina has quit IRC14:04
*** fkautz has quit IRC14:04
*** dougwig has quit IRC14:04
*** jpmaxman has quit IRC14:04
*** xgerman has quit IRC14:04
*** NobodyCam has quit IRC14:04
*** mgkwill has quit IRC14:04
*** petems has quit IRC14:04
*** amit213 has quit IRC14:04
*** monester has quit IRC14:04
fungithe openssh 7.3 release notes are a fun read14:04
*** ameade has quit IRC14:04
*** cargonza has quit IRC14:04
*** rpodolyaka has quit IRC14:04
*** tpeoples has quit IRC14:04
*** ogelbukh has quit IRC14:04
*** thiagolib has quit IRC14:04
*** samyewell has quit IRC14:04
*** csatari has quit IRC14:04
*** geekinutah has quit IRC14:04
*** puiterwijk|cld has quit IRC14:04
*** DuncanT has quit IRC14:04
*** rackertom has quit IRC14:04
*** vdrok has quit IRC14:04
*** wewe0901 has quit IRC14:04
*** simonmcc has quit IRC14:04
*** tpsilva has quit IRC14:04
*** jraim has quit IRC14:04
*** jaranovich has quit IRC14:04
*** bryan_att has quit IRC14:04
*** erlon has quit IRC14:04
*** jed56 has quit IRC14:04
*** dham1 has quit IRC14:04
*** zaro has quit IRC14:04
*** teran has quit IRC14:04
*** maximov has quit IRC14:04
*** zhiyan has quit IRC14:04
*** auggy has quit IRC14:04
*** briancurtin has quit IRC14:04
*** scottda has quit IRC14:04
*** kozhukalov has quit IRC14:04
*** betherly has quit IRC14:04
*** tdurakov has quit IRC14:04
*** kgalanov has quit IRC14:04
*** boris-42 has quit IRC14:04
*** edwarnicke has quit IRC14:04
*** zubchick has quit IRC14:04
*** madhuvishy has quit IRC14:04
*** fmccrthy has quit IRC14:04
*** itisha has quit IRC14:04
*** asaprykin has quit IRC14:04
*** kong has quit IRC14:04
*** lindsayk has quit IRC14:04
*** serverascode has quit IRC14:04
*** Ng has quit IRC14:04
*** rajinir has quit IRC14:04
*** derekjhyang has quit IRC14:04
*** coolsvap has quit IRC14:04
*** igorbelikov has quit IRC14:04
*** patrickeast has quit IRC14:04
*** ggherdov- has quit IRC14:04
AJaegerAnd that needs also my third patch: - and that's the real tricky one IMHO. So, first drink your coffee before you review that one, please14:04
fungioh that's a neat idea... they publish their sha256 checksums in a base64-encoded form instead of hex14:05
*** achanda has quit IRC14:05
openstackgerritMerged openstack-infra/project-config: Use constraints for venv
*** matrohon has quit IRC14:06
*** madhuvishy has joined #openstack-infra14:09
*** vdrok has joined #openstack-infra14:09
*** rajinir has joined #openstack-infra14:09
*** wewe0901 has joined #openstack-infra14:09
*** NobodyCam has joined #openstack-infra14:09
*** maximov has joined #openstack-infra14:09
*** teran has joined #openstack-infra14:09
*** cargonza has joined #openstack-infra14:09
*** dham1 has joined #openstack-infra14:09
mtreinishfungi: I guess 7.3 means no more 1026 character passwords. :) I'm also gonna miss the possible break in attempt messages14:10
*** Ng has joined #openstack-infra14:10
*** puiterwijk|cld has joined #openstack-infra14:10
*** ajmiller has joined #openstack-infra14:11
fungimtreinish: also no more <1024-bit rsa keys14:11
fungifor auth14:11
fungiand no more ssh v1 protocol14:11
mtreinishfungi: well those were future things right? (well v1 is compile time disabled by default)14:12
fungithis has all been a long time coming. nobody should be surprised14:12
*** bhavik1 has joined #openstack-infra14:12
fungialso it'll take a while to get into releases of distros other than openbsd14:13
*** zz_dimtruck is now known as dimtruck14:13
mtreinishheh, arch already has the 7.3 package up. So I'll see what issues it causes as soon as the mirror I'm using pulls it in14:13
*** ildikov has joined #openstack-infra14:14
*** samyewell has joined #openstack-infra14:14
AJaegerwow, they're fast14:14
*** dougwig has joined #openstack-infra14:14
*** jaranovich has joined #openstack-infra14:14
*** petems has joined #openstack-infra14:14
*** tonytan4ever has joined #openstack-infra14:14
fungiis arch only rolling release or do they do point in time releases too?14:14
*** burgerk has joined #openstack-infra14:15
*** nmagnezi has quit IRC14:15
mtreinishfungi: only rolling14:15
*** Goneri has quit IRC14:15
zigofungi: Hi there!14:16
zigofungi: Do you have any idea how I can leverage pabelanger's new debian repo, which was merged here?
*** roxanaghe has joined #openstack-infra14:17
*** Goneri has joined #openstack-infra14:17
*** xarses has quit IRC14:17
zigofungi: How can I publish there?14:17
*** tonytan_brb has quit IRC14:17
fungizigo: i'm guessing he has separate jobs proposed for that or something. i haven't been following closely14:18
*** fkautz has joined #openstack-infra14:18
*** csatari has joined #openstack-infra14:18
*** asaprykin has joined #openstack-infra14:18
*** itisha has joined #openstack-infra14:18
*** kong has joined #openstack-infra14:18
*** zaro has joined #openstack-infra14:18
*** geekinutah has joined #openstack-infra14:18
*** mattfarina has joined #openstack-infra14:18
*** lindsayk has joined #openstack-infra14:18
*** mgkwill has joined #openstack-infra14:18
*** bryan_att has joined #openstack-infra14:18
*** erlon has joined #openstack-infra14:18
*** coolsvap has joined #openstack-infra14:18
*** boris-42 has joined #openstack-infra14:18
*** patrickeast has joined #openstack-infra14:18
zigofungi: I'll wait until he shows up today then.14:18
*** edwarnicke has joined #openstack-infra14:18
*** DuncanT has joined #openstack-infra14:18
*** xgerman has joined #openstack-infra14:18
*** rackertom has joined #openstack-infra14:18
AJaegerzigo, AFAIK: should merge first - and then you just copy that file to the right place.14:18
*** briancurtin has joined #openstack-infra14:18
*** jamesmcarthur has joined #openstack-infra14:18
*** thiagolib has joined #openstack-infra14:18
*** jpmaxman has joined #openstack-infra14:18
*** fmccrthy has joined #openstack-infra14:18
*** ogelbukh has joined #openstack-infra14:18
*** zubchick has joined #openstack-infra14:18
*** monester has joined #openstack-infra14:18
*** ggherdov- has joined #openstack-infra14:18
*** kgalanov has joined #openstack-infra14:18
*** amit213 has joined #openstack-infra14:18
*** jed56 has joined #openstack-infra14:18
*** auggy has joined #openstack-infra14:18
*** kozhukalov has joined #openstack-infra14:18
*** igorbelikov has joined #openstack-infra14:19
*** tpeoples has joined #openstack-infra14:19
*** rpodolyaka has joined #openstack-infra14:19
*** scottda has joined #openstack-infra14:19
*** tdurakov has joined #openstack-infra14:19
*** jraim has joined #openstack-infra14:19
*** sridhar_ram has joined #openstack-infra14:19
*** betherly has joined #openstack-infra14:19
fungiAJaeger: i think he's asking about how to put packages into the repo, not how to install packages from it14:19
*** ameade has joined #openstack-infra14:19
*** zhiyan has joined #openstack-infra14:19
*** simonmcc has joined #openstack-infra14:19
*** ttx has joined #openstack-infra14:19
*** petems has quit IRC14:19
*** serverascode has joined #openstack-infra14:19
AJaegerfungi, oh, then I'm clearly off-topic ;(14:19
*** ajmiller has quit IRC14:19
*** ajmiller has joined #openstack-infra14:19
AJaegerzigo, isn't today a national holiday in Canada and thus pabelanger might be out?14:20
*** fkautz has quit IRC14:20
*** hurgleburgler has joined #openstack-infra14:20
zigoAJaeger: That's a separate issue. This one adds the repo to the image, but that's not the info I need. I need to know how to make this: publish to jessie-newton-backports instead of just jessie-newton.14:20
zigoAJaeger: Ah, good to know!14:21
zigoI thought pabelanger was in USA>14:21
*** roxanaghe has quit IRC14:21
fungitoronto, ca14:21
fungii think14:21
*** derekjhyang has joined #openstack-infra14:21
*** petems has joined #openstack-infra14:22
*** matrohon has joined #openstack-infra14:22
*** hongbin has joined #openstack-infra14:22
fungioh, ottawa not toronto (i was close, they're in the same province)14:23
zigoWoooo ! That's crazy. Every city can declare today as off or not.14:23
*** fkautz has joined #openstack-infra14:24
*** tpsilva_ has joined #openstack-infra14:24
fungioh, i don't know about that14:24
zigoAnd also a different day name depending where you are.14:25
fungiso "Colonel By Day" in ottawa14:25
*** thiagop has joined #openstack-infra14:25
*** chlong has joined #openstack-infra14:26
*** eharney has joined #openstack-infra14:26
AJaegerAnd I thought only German holidays were strange...14:27
*** chlong is now known as chlong|gone14:27
*** Hal has joined #openstack-infra14:27
openstackgerritMatt Borland proposed openstack-infra/project-config: WIP: Splitting current/deprecated integ. test
*** Hal is now known as Guest6812114:28
*** mmandel has quit IRC14:28
*** karthikp_ has joined #openstack-infra14:29
*** jaosorior has quit IRC14:30
*** matrohon has quit IRC14:31
*** PollyZ has joined #openstack-infra14:32
*** chlong|gone has quit IRC14:32
*** jaosorior has joined #openstack-infra14:32
*** chlong|gone has joined #openstack-infra14:33
*** chlong|gone is now known as chlong14:34
*** yamamoto has joined #openstack-infra14:36
rcarrillocruzfungi , pleia2 : i responded to the translate devstack thread, let me know what you think14:36
*** mdrabe has quit IRC14:36
fungircarrillocruz: i just replied to your reply with some ideas14:36
Zarahi, infra! wanders lonely as a cloud, with two +2 but no +A. (it's the one that makes storyboard task ids searchable in gerrit)14:37
fungii think autoupdating dns is probably a no-go for this, but there are reasonable alternatives we could implement14:37
openstackgerritMonty Taylor proposed openstack-infra/zuul: Add an internals doc to the documentation
openstackgerritMonty Taylor proposed openstack-infra/zuul: Rename BasePipelineManager to PipelineManager
openstackgerritMonty Taylor proposed openstack-infra/zuul: Move code out of zuul/manager/
*** cloudnull is now known as cloudkiller14:37
mordredShrews, notmorgan: I pushed up some documentation patches based on last thursday ^^14:37
*** tonytan4ever has quit IRC14:38
Shrewsi'd rather review those than the ansible things14:38
* Shrews refocuses14:38
*** hwcomcn has joined #openstack-infra14:38
mordredShrews: the intenrals doc patch probably needs more work - but I think the docstrings one is probably in decent shape14:38
*** hwcomcn has quit IRC14:39
mordredZara: I think this will require a gerrit restart to pick up the change14:39
Zaraaha, right14:39
*** hwcomcn has joined #openstack-infra14:39
mordredand since it's mucking with indexing, I'm not sure if that means it'll be a long downtime or not (if it has to reindex stuff)14:40
mordredzaro: ^^ do you know?14:40
*** PollyZ has quit IRC14:40
ZaraI think there may be others in that series that don't require a restart, but I'm not sure14:40
*** hwcomcn has quit IRC14:41
*** yamamoto has quit IRC14:41
Zarathat is,
*** eric_lopez has quit IRC14:41
mordredyah. I think the one before it in line also will need a restart14:42
mordredfungi: I think is a thing that having PTL vote on is probably good14:42
*** hwcomcn has joined #openstack-infra14:42
*** mdrabe has joined #openstack-infra14:42
*** cloudkiller is now known as cloudnull14:43
Zaramordred: aha, I didn't realise that one would. I guess that means installing it on review-dev is also blocked until then, since it looks like that also relies on the configuration patch.14:43
*** tonytan4ever has joined #openstack-infra14:44
mordredZara: I'm guessing on several things here - but I think zaro will know all of the things once he hath awakened14:45
Zaracool, I'm sorry if people said before and I missed the significance; I remember people discussing restarts but I thought that was just for installing it on review.o.o14:45
*** mmandel has joined #openstack-infra14:47
*** siva_krish has joined #openstack-infra14:47
*** matrohon has joined #openstack-infra14:47
*** thiagolib has quit IRC14:48
fungiyolanda: i think you misread 34899014:48
fungiit's about github descriptions, not gerrit descriptions14:48
*** eric_lopez has joined #openstack-infra14:49
*** clenimar has quit IRC14:49
fungiyolanda: same for 34899114:50
*** signed8bit has joined #openstack-infra14:50
anteayaZara: I also didn't know about restarts for the its-storyboard plugin patches14:50
*** jcoufal has joined #openstack-infra14:50
anteayaso if it was stated previously and was missed, I also missed it14:51
*** jcoufal_ has quit IRC14:51
*** clenimar has joined #openstack-infra14:51
*** tonytan_brb has joined #openstack-infra14:51
*** thiagolib has joined #openstack-infra14:51
anteayaalso today is a holiday in Canada and I'm just here for the third party meeting14:51
*** hwcomcn_ has joined #openstack-infra14:52
openstackgerritMatt Borland proposed openstack-infra/project-config: Splitting current/deprecated integ. test
*** xarses has joined #openstack-infra14:53
*** karthikp_ has quit IRC14:53
fungimordred: have you tried out the script in 348991 at all?14:54
*** tonytan4ever has quit IRC14:54
rcarrillocruzfungi: btw, i'm all over to use Designate in whatever provider14:54
mordredfungi: nope. not even a little bit - I wanted to make sure you thought it was a good idea before I started trying to make sure it works14:55
*** beekneemech is now known as bnemec14:55
fungimordred: i do, though will update the configuration change with some thoughts around timeline and announcing14:55
*** hwcomcn has quit IRC14:55
vsaienkoinfra core team please help to merge, the patch adds a tempest based non-voting job for networking-generic-switch project. We are going to switch to tempest based tests instead of shell script.14:56
*** ajmiller has quit IRC14:56
matt-borlandkrotscheck, is there a reason we run xvfb in the background for npm (and other) testing rather than using xvfb-run?  It seems like there are advantages to running as xvfb-run.  Thoughts?14:56
*** ajmiller has joined #openstack-infra14:57
*** mriedem_meeting is now known as mriedem14:57
*** asselin has joined #openstack-infra14:58
*** esberglu has joined #openstack-infra14:58
*** stew925 has quit IRC14:58
*** eeiden has quit IRC14:59
*** skipp_ has quit IRC14:59
rcarrillocruzfungi: support for recordsets and zones in designate landed in ansible devel
rcarrillocruz i hope we move over from rackdns at some point14:59
*** jistr is now known as jistr|call15:00
fungircarrillocruz: yep, i'd love to be able to use designate too (or even just bind or something) but the trick is finding an interface which is agreeable to the foundation staff with whom we share control of the domain15:00
mtreinishrcarrillocruz, fungi: is the system-config change. All the dependencies have landed, so now we should be ready to create the server and see where things fall apart :)15:00
*** coolsvap has quit IRC15:01
rcarrillocruzoh, ok15:01
fungiif it were just our team, this would be an easy thing to solve15:02
rcarrillocruzfungi: wasn't aware of the sharing thing with the foundation15:02
*** karthikp_ has joined #openstack-infra15:02
rcarrillocruzthat's not entirely a technical problem then15:02
fungiindeed, most of the complication in moving dns would be entirely nontechnical15:02
*** matrohon has quit IRC15:02
*** rwsu has joined #openstack-infra15:05
openstackgerritJesse Pretorius (odyssey4me) proposed openstack-infra/project-config: Implement Xenial/CentOS jobs for OpenStack-Ansible
*** amitgandhinz has quit IRC15:08
*** karthikp_ has quit IRC15:09
*** jaosorior is now known as jaosorior_lunch15:09
*** amitgandhinz has joined #openstack-infra15:09
*** rwsu has quit IRC15:09
rcarrillocruzmtreinish: i kiiind of remember pabelanger said to create the server for firehose?15:10
*** dprince has quit IRC15:10
rcarrillocruzi'm cool doing it, don't want to create a dupe resource or start if he has it in mind15:10
pabelangerrcarrillocruz: go for it15:10
*** kushal has joined #openstack-infra15:10
pabelangeralso, morning!15:10
rcarrillocruzyeah, morning paul!15:10
pabelangerfungi: Looks like I'm going to move PTO to friday now15:11
*** rwsu has joined #openstack-infra15:11
rcarrillocruzmtreinish: OS and flavor?15:11
*** ajmiller_ has joined #openstack-infra15:11
mtreinishrcarrillocruz: it needs to be xenial, the flavor doesn't have to be big15:11
pabelangerrcarrillocruz: I offered up since it was going to be ubuntu-xenial, so there might be puppet issues15:11
mtreinishrcarrillocruz: fungi put a note in the spec for the flavor let me dig that up15:11
*** zhurong has quit IRC15:12
pabelangerrcarrillocruz: 2GB should be the smallest flavor we use15:12
mtreinishrcarrillocruz: 2GB flavor15:12
*** jistr|call is now known as jistr15:12
*** clenimar has quit IRC15:13
fungiapparently i think ahead. i don't even remember writing that now ;_15:13
mtreinishfungi: heh, you've put all the details in there15:14
mtreinishI should have looked at it beforehand. I did everything out of order15:14
*** ajmiller has quit IRC15:15
fungimeh, the order was merely a suggestion in case you needed it15:15
*** eeiden has joined #openstack-infra15:15
*** ericksonsantos has quit IRC15:15
fungimost of that stuff is not actually order-dependent15:15
JayFIf someone has a sec, would appreciate a +2A on to fix a misconfiguration in our gate (which is causing us to build ramdisks everytime)15:15
*** skipp has joined #openstack-infra15:16
*** lucasagomes has quit IRC15:16
AJaegerJayF: alternative for that change would be changing the default - does that one really have the correct value? Perhaps you need both changes...15:16
*** Teresita-Warrior has joined #openstack-infra15:17
JayFAJaeger: We're likely going to change default as well, but we should be explicit either way15:17
openstackgerritMonty Taylor proposed openstack-infra/jeepyb: Add script to reset all descriptions to global default
AJaegerJayF: I agree.15:17
openstackgerritMonty Taylor proposed openstack-infra/system-config: Set a global description override
mordredfungi: tested/fixed:
*** matrohon has joined #openstack-infra15:19
*** karthikp_ has joined #openstack-infra15:19
pabelangermordred: fungi: do you mind helping land adds tripleo-test-cloud-rh(1|2) into ansible-clouds so we can run puppet on them15:20
pabelangerfor AFS15:20
AJaegermordred: did you see fungi's comment on 348992 ? I like his suggested wording a wee bit better.15:21
*** raildo has quit IRC15:23
*** ggnel_t has quit IRC15:24
*** thiagop has quit IRC15:24
rcarrillocruzfungi , pabelanger : I put it on DFW region, yeah?15:25
*** lucasagomes has joined #openstack-infra15:25
fungipabelanger: see mordred's comment on 347158 patchset 2 about using the wrong variable names in one of them. i didn't see any response to that15:25
pabelangerfungi: I commented out of band, let me update the review15:26
*** hongbin has quit IRC15:26
openstackgerritMonty Taylor proposed openstack-infra/jeepyb: Add script to reset all descriptions to global default
*** hongbin has joined #openstack-infra15:27
mordredfungi, pabelanger: looking at comments now - I just added one more nicety to that ^^ (turns out having homepage set when we have urls in the description anyway looks weird)15:27
*** bhavik1 has quit IRC15:27
*** karthikp_ has quit IRC15:28
*** karthikp_ has joined #openstack-infra15:28
anteayaI'm going to enjoy the rest of my holiday Monday15:29
anteayasee you tomorrow15:29
*** yottatsa has quit IRC15:29
openstackgerritMonty Taylor proposed openstack-infra/system-config: Set a global description override
*** EricGonczer_ has joined #openstack-infra15:29
AJaegeranteaya: enjoy!15:29
*** ericksonsantos has joined #openstack-infra15:30
*** clenimar has joined #openstack-infra15:31
AJaegerfungi, germqtt release tarball looks wrong in log file with zuul-cloner: ;(15:32
openstackgerritMerged openstack-infra/system-config: Add openstackci-tripleo-test-clouds to ansible-clouds.yaml
*** thiagop has joined #openstack-infra15:32
*** mdrabe has quit IRC15:32
*** raildo has joined #openstack-infra15:32
*** mdrabe has joined #openstack-infra15:33
fungiAJaeger: yep, i have a feeling zuul-cloner doesn't handle tag refs based on that error15:33
fungithat's the sort of complication i was worried we'd run into15:33
*** psachin has quit IRC15:34
*** matrohon has quit IRC15:34
*** sdake has quit IRC15:34
AJaegerI had expected that our recent fixes for fetching should have fixed this. corvus ^15:35
* AJaeger reverts now...15:35
fungiAJaeger: that error _may_ be benign15:35
fungiAJaeger: see where it says "Prepared 'openstack-infra/germqtt' repo at revision '37b472a13419cfc1cdc6179b11f136fbfdb85064'15:36
*** sambetts is now known as sambetts|afk15:36
fungi`git show-ref 0.1.0` says 37b472a13419cfc1cdc6179b11f136fbfdb8506415:36
AJaegerAh! I was looking for 37xxx but couldn't find it.15:36
fungiAJaeger: i'm checking to make sure the artifacts are sane now15:37
openstackgerritMerged openstack-infra/project-config: Ensure IRONIC_BUILD_DEPLOY_RAMDISK is always explict
AJaegerso, tar-ball might be sane this time - but we cannot gurantee that for the next time?15:37
*** yamamoto has joined #openstack-infra15:37
*** vincentll has quit IRC15:38
fungiAJaeger: well, the fallback ref zuul-cloner retrieved was the tag, not the ref the tag points to15:38
fungiwhich means this _should_ work fine, albeit with confusing debug messages15:39
AJaegervery confusing debug messages ;(15:39
*** kushal has quit IRC15:39
fungiversioning and content looks correct when i pip install germqtt into a virtualenv too15:41
fungiit checked out the tag by its sha rather than by its name15:41
mtreinishfungi: yeah, I checked the code in the tarball on pypi it looks correct15:41
fungibut it definitely checked out the sha of the tag, not the sha of the ref that got tagged15:42
fungiwhich is the behavior we want15:42
fungithe messages do worry me that we're ending up using an unintended fallback path for this which just happens to do the right thing15:42
AJaegerfungi, my worry as well - I fear it just worked by accident this time and might fail for a stable release15:43
*** harlowja_at_home has joined #openstack-infra15:43
fungiprobably the step that checks for a branch matching $ZUUL_REFNAME should also look for a tag matching that15:43
*** yamamoto has quit IRC15:43
*** Cormite has quit IRC15:43
*** stew925 has joined #openstack-infra15:44
corvusfungi, AJaeger: good morning15:44
fungioh, and yes the requirements fallback behavior for this will be very weird for point release tags. zuul-cloner will need to somehow know what stable branch of requirements to checkout for a given tag15:44
fungizuul doesn't, i think, actually have sufficient information to be able to figure that out15:45
*** amitgandhinz has quit IRC15:45
rcarrillocruzmtreinish: i'm spinning the node fyi, didn't tell you i already started15:46
rcarrillocruzgoing thru puppet bootstrap now15:46
AJaegercorvus, fungi:
*** amitgandhinz has joined #openstack-infra15:46
AJaegerAnd the checked out tree IS master at HEAD ;(15:46
AJaeger(note that cachedir was empty)15:47
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Revert "Use constraints for venv"
AJaegerfungi, revert ^15:47
mtreinishrcarrillocruz: heh, ok cool, thanks15:47
corvuswhy is it being run with '--branch' ?15:47
fungii suspect that's in error. we should only specify --branch if zuul isn't providing sufficient branch information15:48
*** M-docaedo_vector has quit IRC15:48
AJaegercorvus: see jenkins/jobs/macros.yaml for zuul-git-prep-upper-constraints builder15:48
*** rbuzatu has joined #openstack-infra15:49
fungihowever, separate from that, i don't think zuul has sufficient context to know what branch should integrate with a given tag15:49
AJaegerThat one always adds --branch15:49
fungigit tags are branchless, yes?15:49
fungithe gerrit push for a tag certainly doesn't include any branch context15:50
fungiso if the gerrit event for it does mention a branch, i wonder how it even gets that right at all15:50
corvuslet me be upfront here -- i do not understand the 'upper constraints' system, so i am unlikely to be able to suggest a way to implement or fix it15:50
corvusi will say that, from what i can see here, zuul-cloner appears to be doing exactly what was asked of it15:50
*** matrohon has joined #openstack-infra15:51
fungicorvus: there's a file in a (branched) repo which jobs for other repos want to retrieve a copy of (from the "correct" branch)15:51
*** mikelk has quit IRC15:51
zaroZara, mordred : change does require a restart and reindex.  i'm thinking we wouldn't do a reindex just for that but just add it in for the next time we need to reindex, like project name change or upgrade.15:51
AJaegercorvus, ignore upper constraints. What we need is that as part of the release we need to check out a project at a given tag - and requirements HEAD for the same branch15:51
fungiAJaeger: which is where this falls apart15:51
fungithere is no corresponding branch for a git tag at all15:51
AJaegerfungi ;(15:51
*** matrohon has quit IRC15:51
*** abregman is now known as abregman|afk15:52
corvusAJaeger: thanks, that's a problem statement i can work with :)  and yes, fungi is saying things that match my understanding :(15:52
fungiwhat we have will work for projects which tag on master, but we'd need lots of extra (fragile) logic to identify a branch corresponding to a given tag15:52
*** Sukhdev has joined #openstack-infra15:52
AJaegerSo, let's revert my change (proposed already) - and then go back to drawing board?15:53
*** rbuzatu has quit IRC15:53
AJaegerUnfortunately I need to step out for an hour or so now ;(15:53
fungiAJaeger: yep, we need to regroup on this topic15:53
corvusi mean, one could inspect a git repo to see which branches a tag is present on.  that number can legitimately be 0..n.  with the new release process, there might be other factors that further restrict that though, so perhaps the assumption could be made.15:53
*** M-docaedo_vector has joined #openstack-infra15:53
fungiyeah, i think it's tractable, but nontrivial15:54
*** tsg has joined #openstack-infra15:54
fungiand probably can't be generalized to all use cases for git, but potentially can for our use case15:54
*** tsg has quit IRC15:55
fungiwe did some similar logic for the tag merge proposal job15:55
*** M-docaedo_vector has quit IRC15:55
pabelangerAJaeger: re: 348890. I think ianw objection is that nodepool gate jobs are running on ubuntu-xenial for devstack.  We should move that back to ubuntu-trusty since nodepool.o.o is still running ubuntu-trusty15:55
pabelangerI can look at that now15:55
pleia2rcarrillocruz: thanks fo replying, it's good to have the ideas floating around15:55
*** yottatsa has joined #openstack-infra15:55
AJaegerpabelanger: Ah... thanks for looking into it.15:56
*** thorst is now known as thorst_afk15:56
pleia2also, it's late here and I'm quite jetlagged, but I just wanted to say I love err2d2, nice one SpamapS :D15:57
asselinrcarrillocruz, did you ever get a chance to look at this?
mordredpleia2, fungi:
rcarrillocruzasselin: i'll get to it this week, i'm cleaning other ansible openstack issues15:58
rcarrillocruzbut in short15:58
rcarrillocruzi suspect that's going to be an ER15:58
rcarrillocruzas passing explicitly verify no on the task param itself, should work15:58
rcarrillocruzbut don't quote me15:58
rcarrillocruzi need to verify15:59
pleia2mordred: thumbs up15:59
fungiAJaeger: corvus: i think the tricky part will be if a ref appears in more than one branch at the time it gets tagged (which can very easily happen in situations like the coordinated release where we branch stable from master and then might need no backports during the rc period) then we need a priority ordering to decide which branch that goes with (maybe earliest non-master in an alphanumeric sort15:59
fungisince we use progressive branch naming)15:59
*** rossella_s has quit IRC16:00
*** coolsvap has joined #openstack-infra16:00
*** bhavik has joined #openstack-infra16:00
*** rossella_s has joined #openstack-infra16:00
*** armax has joined #openstack-infra16:00
*** jistr is now known as jistr|biab16:01
asselinrcarrillocruz, ok, let me know. right now I'm using patched version of cloud-launcher16:01
Zarazaro: thanks for explaining, and yeah, makes sense. should we include any other patches from that series next time a restart is needed?16:01
pleia2mordred: btw, apache does the same, at the meeting where we talked about this a few weeks back I had some links in the agenda16:02
* pleia2 wikihistory diggint16:02
*** dtantsur is now known as dtantsur|bbl16:03
*** achanda has joined #openstack-infra16:03
openstackgerritKien Ha proposed openstack-infra/jenkins-job-builder: gitlab trigger: Support new "trigger-open-merge-request-push" options
pleia2mordred: but the wording on the git one is nice, so it's good to point to16:04
*** Guest15832 is now known as redrobot16:04
Zara(btw, all, it looks like last time lp data was migrated to storyboard, the encoding for some pasted things wasn't preserved, and we ended up with some weird artifacts hanging around in story descriptions. so that's something to watch out when the time comes.)16:04
openstackgerritPaul Belanger proposed openstack-infra/project-config: Switch to gate-{name}-linters-{node}
mordredpleia2: it sure would be nice if we could get the nice little "mirrored from" like apache has on their16:05
Zara(I've just gone through tidying up so that scripts will work nicely again, but weirdness should be preserved in plenty of comments when it's the right time to look.)16:05
*** e0ne has quit IRC16:05
pleia2mordred: alas, they don't grant them anymore as far as I've read16:05
fungiup to 187 out of 798 votes cast for an infra team mascot, with ~8 hours remaining16:05
mordredpleia2: yah16:05
*** ayoung has quit IRC16:06
mordredpleia2: and jono left already, so there's nobody on the inside to bug16:06
*** eharney has quit IRC16:07
corvusbut *still* the repos allow folks to open pull requests16:07
fungii wonder if that's a different kind of mirroring implementation in github16:07
pleia2mordred: yeah, I know a few folks thanks to codeconf last month but not quite the right people (and I think they'd rather see us either not mirror or just give up and use them)16:07
mordredpleia2: well, I'd also rather stop mirroring myself ... but that seems to annoy people16:08
fungii'm with them on the just not mirroring to github thing. not so much on the giving up and using github16:08
corvuspleia2: would they be willing to prevent anyone from mirroring our repos?16:08
*** yottatsa has quit IRC16:08
*** matrohon has joined #openstack-infra16:08
pleia2corvus: I highly doubt it16:08
corvusthen i don't think that's really a viable option16:08
*** achanda has quit IRC16:08
openstackgerritMerged openstack-infra/project-config: Make pika jobs voting
openstackgerritMerged openstack-infra/project-config: grafana: Follow xenial renames
mordredyah - before we mirrored to github, other "unofficial" mirrors that were run poorly became defacto mirrors that people talked about and stuff16:09
corvusso they're not really giving us a choice -- "give up and use github" is the only one they are offering16:09
openstackgerritMerged openstack-infra/project-config: Remove broken coverage jobs
*** markvoelker has quit IRC16:09
mordredk. I'm going to send the email to the list16:09
corvus(fortunately, it's not the only one we have)16:09
pleia2mordred: yeah, people like the UI and there are features we can't keep up with pretty-wise16:09
*** Sukhdev has quit IRC16:10
corvuswell, i would really rather not get into yet another discussion about github, but github is pretty easy to beat on features16:10
*** matrohon has quit IRC16:10
*** Sukhdev has joined #openstack-infra16:10
*** jaosorior_lunch has quit IRC16:11
Zarathe only time I use it is when I want to search a repo, so I don't think I'd have any reason to keep using it if I could get comparable search elsewhere.16:12
mtreinishpleia2: syntax highlighting/coloring and multiline links are the big ones. Have that on cgit (or whatever) and I don't think most people would care if we dropped github mirroring (or at least for me)16:12
mordredmtreinish: ++16:12
*** jaosorior has joined #openstack-infra16:12
pleia2mtreinish: yeah, those are the two people cite all the time16:12
fungiZara: we do have too16:12
mordredcorvus: in other news - I realized in writing that we do not have a docs check job for zuul16:12
openstackgerritMatt Borland proposed openstack-infra/project-config: Splitting current/deprecated integ. test
*** yottatsa has joined #openstack-infra16:13
mordredcorvus: nevermind16:13
*** nwkarste_ has quit IRC16:13
mordredcorvus: I realized just now that it's in a position in the list that fooled me16:13
pleia2we should link codesearch to cgit somehow :\16:13
pleia2maybe edit a footer or something16:13
*** nwkarsten has joined #openstack-infra16:13
*** sputnik13 has joined #openstack-infra16:14
*** notnownikki has quit IRC16:14
fungipleia2: you mean from? we already link it to cgit16:14
pleia2I wonder if I could put something near the useless search bar16:14
*** tesseract- has quit IRC16:14
pleia2fungi: er, yeah16:14
fungiokay, then yes i totally agree16:14
fungimaybe the search field could just call out to codesearch?16:14
mordredthat would be super neat16:15
pleia2that would be good16:15
mtreinishrcarrillocruz: how's the firehose startup going? anything I can help with?16:15
fungiit uses get parameters after all16:15
fungiso would probably be trivial to template the action url16:15
mordredyah - should be a 'simple' matter of html templating16:15
pleia2mordred: easier than writing C, which is how you change anything else in cgit ;)16:15
*** pt_15 has quit IRC16:16
*** nwkarste_ has joined #openstack-infra16:16
pleia2anyway, time to stop computering and get some rest, night16:18
mtreinishpleia2: cgit's cgit has colors: just a newer version?16:18
*** tpsilva_ has quit IRC16:18
*** nwkarsten has quit IRC16:18
Zara(atm I would type 'st' into my browser to get the repo to come up in search history, browse to repo, then use the 'search this repo' box to search. so if the search box on cgit automatically goes to a search filtered by the relevant project, that should be just as efficient.)16:19
rcarrillocruzmtreinish: it's up,  i'm checking with pabelanger an issue with xenial, it appears it's more strict in terms of root logins compared to trusty16:19
pleia2mtreinish: I actually thought we had some syntax higlighting at some point, but now blanking16:19
Zara(a reason I don't currently use hound by default is that I have to filter it each time.)16:19
*** notnownikki has joined #openstack-infra16:20
pleia2maybe we broke it at some point /o\16:20
rcarrillocruzpabelanger: give me 5min, gotta bring a pacifier to my wife, she's walking around Mario16:21
pabelangerrcarrillocruz: we need to add UseDNS yes to /etc/ssh/sshd_config now, not sure why but ubuntu-xenial we need to define that16:21
*** karthikp_ has quit IRC16:21
*** hwcomcn_ has quit IRC16:21
pleia2could be selinux again, I'd have to look16:21
mordredpleia2: looks like there is support for using python pygments for highlighting too16:21
pabelangerrcarrillocruz: just tested and root can login again16:21
*** ayoung has joined #openstack-infra16:21
rcarrillocruz++ , we can puppet that then I guess16:21
rcarrillocruzmtreinish: ^16:21
openstackLaunchpad bug 1588457 in Release Notes for Ubuntu "UseDNS default changed to no, locking out authorized_keys from="hostname" users when upgrading to Xenial" [Undecided,Fix released]16:21
pabelangerrcarrillocruz: that is our bug^16:21
rcarrillocruzneed to create a CNAME for the server16:21
rcarrillocruzgive me 5min, i'll be back and do it16:22
SpamapSpleia2: :-D16:23
pleia2mordred: yeah, that's the cgitrc setting I linked to in the puppet module16:23
pabelangerfungi: thoughts on ^? ubuntu-xenial has defaulted UseDNS no for /etc/ssh/sshd_config (because of security?) Should we follow sure and stop using DNS hostnames in /etc/ssh/sshd_config?16:23
zaroZara: yes, all of them :)16:23
*** karthik__ has joined #openstack-infra16:23
mordredpleia2: ++16:23
pleia2mordred: so we do set it up, it's just broken16:24
zaroZara: basically anything that adds/updates a configuration requires a restat16:24
pleia2and it's 10PM here, so maybe I will look tomorrow :)16:24
pabelangerfungi: not security related, just makes configuration easier^16:24
fungipabelanger: which hostname do we use? just the one for puppetmaster? i'm in favor of switching that to its ip address16:24
pabelangerfungi: ya, we have PermitRootLogin without-password for puppetmaster.o.o now. which is breaking on xenial16:25
pabelangerand localhost16:25
Zarazaro: haha, okay, cool, I'll remember that.16:25
fungipabelanger: i'm in full agreement that basing openssh decisions on dns lookups (without the safety of dnssec to validate them) is unsafe16:26
pabelangerfungi: ack, let me get a patch16:26
*** matrohon has joined #openstack-infra16:26
openstackgerritKien Ha proposed openstack-infra/jenkins-job-builder: gitlab trigger: Support new "trigger-open-merge-request-push" options
*** hurgleburgler has quit IRC16:28
Zara(whoa, if you hover over an author in cgit master, their avatar pops out!)16:28
*** harlowja_at_home has quit IRC16:30
*** PollyZ has joined #openstack-infra16:30
openstackgerritJulien Danjou proposed openstack-infra/project-config: ceilometer: gate on Python 3.5
*** yottatsa has quit IRC16:31
openstackgerritPaul Belanger proposed openstack-infra/system-config: Use IPv6 address for puppetmaster.o.o
*** asettle has quit IRC16:32
openstackgerritKien Ha proposed openstack-infra/jenkins-job-builder: gitlab trigger: Support new "trigger-open-merge-request-push" options
pabelangerfungi: do you know any clouds not using IPv6? ^16:32
*** jlanoux has quit IRC16:33
*** yamahata has quit IRC16:33
*** dprince has joined #openstack-infra16:33
pabelangerfungi: yup, I'll need to add IPv4 too16:33
*** julim has quit IRC16:34
fungipabelanger: most of them16:34
*** claudiub has quit IRC16:35
fungipabelanger: rackspace and vexxhost are the only ones where we have working ipv6 for our servers at the moment afaik16:35
*** yottatsa has joined #openstack-infra16:35
*** javeriak has joined #openstack-infra16:36
rcarrillocruzfungi: unsure how to create a cname with rackdns CLI, i infer correct here ?16:36
*** julim has joined #openstack-infra16:36
pabelangerfungi: okay, this is going to be more work. Since puppet-ssh is configured to only support hosts16:36
mordredanybody know where is kept?16:36
fungipabelanger: can the acl specify a list of ip addresses on one line/entry?16:36
pabelangerfungi: not today:
fungircarrillocruz: i think that's correct16:37
pabelangerI need to see if Match address, is valid16:37
pabelangerthat is the easier route16:38
fungipabelanger: right, that's what i was asking16:38
funginot what does our puppet template support, but what does openssh's configuration language support16:38
pabelangerfungi: I think it works, test16:38
mordredor, anybody know of any puppet examples in our corpus that perform "install this module from pip for python3" ?16:39
*** yamamoto has joined #openstack-infra16:39
rcarrillocruzfungi: yup, at least it didn't error out16:40
openstackgerritChristian Fetzer proposed openstack-infra/jenkins-job-builder: gitlab trigger: Update Gitlab to >=1.3.0
*** matrohon has quit IRC16:40
rcarrillocruzweee, it resolves fine16:40
rcarrillocruzmtreinish: node launched16:41
openstackgerritMerged openstack-infra/project-config: Revert "Use constraints for venv"
rcarrillocruzdns created16:41
pabelangerfungi: okay, a list works. Fixing16:41
*** tphummel has joined #openstack-infra16:44
*** yamamoto has quit IRC16:44
openstackgerritJim Rollenhagen proposed openstack-infra/project-config: Add ironic multitenant job to nova experimental queue
jrollmriedem: ^16:45
*** tesseract- has joined #openstack-infra16:46
*** mriedem has left #openstack-infra16:47
*** mriedem has joined #openstack-infra16:47
mriedemjroll: +116:47
jrollnp, ty16:47
openstackgerritPaul Belanger proposed openstack-infra/puppet-ssh: Add @trusted_ssh_type for user to override
openstackgerritPaul Belanger proposed openstack-infra/system-config: Use IPv6 address for puppetmaster.o.o
pabelangerfungi: rcarrillocruz: 2 patches to fix ubuntu-xenial16:48
*** daemontool__ has joined #openstack-infra16:48
*** jistr|biab is now known as jistr16:49
mordredpabelanger: sooooooooo16:50
*** rbuzatu has joined #openstack-infra16:50
mordredpabelanger: nevermind16:50
openstackgerritMerged openstack-infra/zuul: Add initial support for jobs authentication config
*** dtantsur|bbl is now known as dtantsur16:51
pabelangermordred: pew :)16:51
mordredpabelanger: I was going to ask what we do if we lose/cycle puppetmaster host16:51
mordredpabelanger: but in those cases I think we can do a playbook as one of us with a sudo argument to update it16:51
openstackgerritMerged openstack-infra/zuul: Add some v3 notes
pabelangermordred: Ya, that's the issue with dropping DNS.  And how I plan on fixing our 2 xenial servers right now16:52
*** daemontool_ has quit IRC16:52
*** osnexus has joined #openstack-infra16:53
*** flepied1 has joined #openstack-infra16:53
*** tesseract- has quit IRC16:54
*** nmagnezi has joined #openstack-infra16:54
*** rbuzatu has quit IRC16:56
*** flepied has quit IRC16:56
*** matrohon has joined #openstack-infra16:57
*** tesseract- has joined #openstack-infra16:58
*** tesseract- has quit IRC16:58
*** Apoorva has joined #openstack-infra16:58
*** karthik__ has quit IRC16:59
*** gyee_ has joined #openstack-infra16:59
*** gyee_ has quit IRC16:59
*** matrohon has quit IRC16:59
*** gyee has joined #openstack-infra17:00
AJaegerpabelanger: did you figure out ianw's comment regarding nodepool and devstack - and xenial? I don't see any devstack jobs run for nodepool...17:00
*** gildub has quit IRC17:01
pabelangerAJaeger: yup, we are good.  They are in jenkins/jobs/nodepool.yaml and still running ubuntu-trusty17:01
*** degorenko is now known as _degorenko|afk17:02
AJaegerpabelanger: ah, thanks!17:02
*** sdake has joined #openstack-infra17:02
openstackgerritAdam Spiers proposed openstack-infra/irc-meetings: move HA meeting an hour later, to 0900 UTC
*** clenimar has quit IRC17:03
mtreinishrcarrillocruz: cool, yep I can ping it now17:04
*** thiagop has quit IRC17:04
*** thiagop has joined #openstack-infra17:04
*** cbader has quit IRC17:04
*** markvoelker has joined #openstack-infra17:05
mtreinishrcarrillocruz, pabelanger: we still need the +A on I'm not sure the order for that when launching a new node17:05
*** achanda has joined #openstack-infra17:06
*** sshnaidm is now known as sshnaidm|afk17:07
openstackgerritAdam Spiers proposed openstack-infra/irc-meetings: move HA meeting an hour later, to 0900 UTC
rcarrillocruzi approved17:08
rcarrillocruzi should've probably created the hiera associated k/v17:08
*** abregman|afk is now known as abregman17:09
mtreinishrcarrillocruz: the gerrit ones require a user being created17:09
*** HeOS has quit IRC17:09
pabelangerrcarrillocruz: if you review 349590 and 349596 that should fix our root SSH issue in xenial17:10
mtreinishI guess we could reuse an existing one17:10
*** dmsimard is now known as dmsimard|afk17:10
*** achanda has quit IRC17:10
mtreinishoh, actually I'd need to change the puppet, because the username is being set in openstack_project::firehose and not overriden in site.pp17:10
*** ayoung has quit IRC17:11
*** achanda has joined #openstack-infra17:11
*** achanda has quit IRC17:11
*** rbuzatu has joined #openstack-infra17:12
*** achanda has joined #openstack-infra17:12
*** yamahata has joined #openstack-infra17:12
rcarrillocruzmtreinish: i assume germqtt_gerrit_ssh_private_key is the key for 'germqtt' user ?17:12
rcarrillocruzso it can read gerrit events, etc17:12
mtreinishrcarrillocruz: yeah17:13
openstackgerritMerged openstack-infra/project-config: networking-midonet: Consume gate_hook if exists
rcarrillocruzwas that user created at all?17:13
mtreinishnope, not yet17:13
rcarrillocruzhaven't checked17:13
mtreinishthat's what I was just saying :)17:13
*** rbuzatu has quit IRC17:13
rcarrillocruzyeah, wasn't sure where you meant17:13
fungipabelanger: are we going to run into issues if 349596 gets approved and applied before 349590?17:13
rcarrillocruzis it expected to be 'germqtt' or...17:13
*** ericksonsantos has quit IRC17:14
mtreinishrcarrillocruz: it's germqtt: L1817:14
AJaegercorvus, fungi: For tagging/constraings: I can run "git show-ref --dereference 8.1.1" (or similar) and then use the sha like "git branch --contains 2c07b060bb9fdf8a244fa700e91861f8ec32b494" to get stable/mitaka.17:14
mtreinishwe can change it if needed though17:14
AJaegercorvus, fungi: But that means I need an uptodate stable/mitaka branch, otherwise it won't work.17:14
rcarrillocruzcool, let me refresh gerrit cli foo to create the user17:14
fungiAJaeger: also, see my followup comments: we can't guarantee that `git branch --contains ...` will return only one branch name17:16
openstackgerritMerged openstack-infra/system-config: Add firehose.o.o config to system-config
fungiAJaeger: so we need logic for selecting which containing branch is the correct one to use17:16
AJaegerfungi, indeed, we can't, we could try to find the oldest...17:16
*** Sukhdev has quit IRC17:16
rcarrillocruzmtreinish, pabelanger : would "Non-Interactive" group be sufficient for 'germqtt' ?17:16
mtreinishrcarrillocruz: I don't know, but I would assume so. Since all it needs is the event stream. But I don't know what that means so I can't be sure17:17
*** sdake has quit IRC17:17
rcarrillocruzi think that group grants that17:17
rcarrillocruzand you won't need any email notifications, do you?17:17
rcarrillocruzi.e. when i create a user, i'm supposed to put an email, full name, groups, and ssh key17:18
mtreinishnope, literally all it needs is to use gerritlib to get an event stream17:18
AJaegerfungi, will we have a uptodate branches?17:18
*** thorst_afk is now known as thorst17:19
*** hongbin has quit IRC17:19
*** thiagop has quit IRC17:19
*** raildo has quit IRC17:20
mtreinishrcarrillocruz: the germqtt code is super simple17:20
*** clenimar has joined #openstack-infra17:21
rcarrillocruzpuppet runs are failing on the node17:22
*** raildo has joined #openstack-infra17:22
rcarrillocruzso user is not even created17:22
*** ericksonsantos has joined #openstack-infra17:22
*** ayoung has joined #openstack-infra17:24
*** thiagop has joined #openstack-infra17:24
*** julim has quit IRC17:24
*** Goneri has quit IRC17:24
*** rbuzatu has joined #openstack-infra17:24
mtreinishrcarrillocruz: any errors I can look at? Or is it something before my poorly written puppet comes into play?17:25
openstackgerritJuan Antonio Osorio Robles proposed openstack/diskimage-builder: TESTING UNDERCLOUD WITH CERTMONGER - DO NOT MERGE
fungiAJaeger: zuul-cloner does a remote update, so knows the current remote branch heads so use -a and you can filter on those17:25
rcarrillocruzmtreinish: errors are not in manifest logic:
*** signed8bit is now known as signed8bit_Zzz17:26
rcarrillocruzand i'm not sure if that's benign, looks to me it fails as it doesn't have master setting on puppet.conf17:26
rcarrillocruzeven though it has [agent] section17:26
rcarrillocruzor well17:27
rcarrillocruzthe master is on the agent section17:27
fungiAJaeger: specifically, `git branch -a --contains $tag|grep '^  remotes/origin/'|cut -d/ -f3-` or something17:27
rcarrillocruzbut we are on a puppetmasterless17:27
rcarrillocruzi think that's just there for getting certs for the puppetdb17:27
rcarrillocruzpabelanger: ^17:27
openstackgerritMikhail S Medvedev proposed openstack-infra/zuul: (wip) Treat builds without a registered job as cancelled
AJaegerfungi, ah that's easier than what I did. Thanks17:29
*** signed8bit_Zzz is now known as signed8bit17:31
*** Ravikiran_K has quit IRC17:31
*** dtantsur is now known as dtantsur|afk17:31
*** acoles is now known as acoles_17:32
*** nwkarste_ has quit IRC17:32
fungiAJaeger: the remaining topic is that we need some ranking mechanism. like prefer the first remotes/origin branch name from an alpahnumeric besides master or HEAD, and then fall back on master if there isn't one17:32
*** Na3iL has quit IRC17:33
fungier, from an alphanumeric sort17:33
fungigiven that we shouldn't have nontrafitional stable/<series> branches in openstack/requirements and this logic is specific to that repo (for now) we likely don't need to get overly complex supporting things like version sort orders, case sensitivity, planning for feature branches, et cetera17:34
fungiwe could maybe even just hardcode to `...|grep ^  remotes/origin/stable/`17:35
AJaegerfungi, yeah - we only need to look for branches that are known to requirements repo, so stable/X17:35
fungiAJaeger: note however that this gets tricky for projects like gnocchi that don't use normal stable branch names17:36
fungiper that thread on the -dev ml last week17:36
AJaegerI don't think we can do anything for gnocchi here - we would need to have a mapping between stable/<gnocchi-tree> vs stable/<requirements>17:37
fungibut maybe they just don't use upper-constraints.txt and then it doesn't matter17:37
fungiyeah, they explicitly map their branch names to coordinated stable release branches in their integration test jobs17:37
AJaegerI know and don't like it ;(17:37
fungiand it looks like they've only added to the mapping on an ad hoc basis as they discover they can't land a stable branch patch for something17:38
AJaegerlet's first get the "normal" case done;)17:38
*** amotoki has quit IRC17:38
*** amotoki has joined #openstack-infra17:38
fungiagreed, best to assume repos follow our expected conventions and let the special snowflakes deal with their specialness on their own17:38
*** amotoki has quit IRC17:38
*** yamamoto has joined #openstack-infra17:41
AJaegerThis gives me the oldest stable branch a tag is in : git branch -a --contains $TAG|grep '^  remotes/origin/'|cut -d/ -f3-|sort|grep ^stable|head -n117:42
*** nwkarsten has joined #openstack-infra17:42
openstackgerritMerged openstack-infra/zuul: Remove TestSchedulerConfigParsing
openstackgerritMerged openstack-infra/zuul: Individually skip all TestScheduler tests
openstackgerritMerged openstack-infra/zuul: Create AnsibleZuulTestCase
pabelangerrcarrillocruz: cannot remember off the top of my head what we use it for these days17:43
*** nwkarsten has quit IRC17:44
*** nwkarsten has joined #openstack-infra17:44
*** karthik__ has joined #openstack-infra17:45
*** eharney has joined #openstack-infra17:45
*** nwkarste_ has joined #openstack-infra17:46
fungiAJaeger: yep, you could factor out the additional grep by changing the first one to look for '^  remotes/origin/stable/' instead17:46
*** nwkarste_ has quit IRC17:46
*** yamamoto has quit IRC17:46
*** nwkarste_ has joined #openstack-infra17:47
*** nwkarste_ has quit IRC17:47
*** nwkarste_ has joined #openstack-infra17:48
*** e0ne has joined #openstack-infra17:48
*** jaosorior has quit IRC17:49
fungimordred: would your announcement about github description flattening be good to get mentioned in thingee's weekly -dev ml digest too?17:49
*** nwkarsten has quit IRC17:49
fungijust thinking about how to get the word out to some of those teams who don't pay close attention to that ml and might otherwise be caught off-guard17:49
fungion the other hand, i'm not too keen on making public mention of github, so maybe the fewer times we invoke its name the better17:50
*** yeeg has joined #openstack-infra17:50
openstackgerritPaul Belanger proposed openstack-infra/system-config: Remove openstackci-tripleo-test-cloud-rh2 from ansible-clouds
*** gyee has quit IRC17:50
*** yeeg has quit IRC17:50
fungior it may make itself manifest in our presence and rend us all asunder17:50
*** nwkarsten has joined #openstack-infra17:51
*** yeeg has joined #openstack-infra17:51
*** yeeg has quit IRC17:51
pabelangerfungi: mordred: We should land that quickly, to avoid puppetmaster ansible-playbook from using openstackci-tripleo-test-cloud-rh2. I forgot they only have a single tenant.  openstackci-tripleo-test-cloud-rh1, which is coming back online will have 2 tenants17:51
*** nwkarsten has quit IRC17:52
*** gyee has joined #openstack-infra17:52
*** nwkarsten has joined #openstack-infra17:52
*** jcoufal has quit IRC17:52
AJaegerfungi, is what I'm testing, looks fine so far.17:53
*** markvoelker_ has joined #openstack-infra17:53
*** nwkarsten has quit IRC17:53
*** bnemec has quit IRC17:53
*** markvoelker has quit IRC17:53
*** nwkarst__ has joined #openstack-infra17:54
AJaegerfungi, so, I basically need to run two zuul-cloner invocations. First check out the project, then find branch for tag, then checkout requirements branch.17:54
*** pt_15 has joined #openstack-infra17:54
*** nwkarste_ has quit IRC17:54
*** nwkarst__ has quit IRC17:54
*** nwkarste_ has joined #openstack-infra17:55
*** PollyZ has quit IRC17:55
*** nwkarste_ has quit IRC17:55
*** bnemec has joined #openstack-infra17:55
fungiAJaeger: oh, good point, you need a local git clone to work from. but ultimately the overhead of calling z-c twice for different repos rather than once for two repos is likely insignificant17:56
fungiso that seems fine17:56
*** nwkarste_ has joined #openstack-infra17:56
*** sambetts|afk has quit IRC17:57
prometheanfireAJaeger: talking about the cross-project testing?17:57
AJaegerprometheanfire: constraints in post queue17:57
fungiAJaeger: release pipeline17:57
*** nwkarste_ has quit IRC17:58
prometheanfirewe just added that17:58
AJaegerfungi is right ;)17:58
prometheanfirehave been useful so far17:58
*** rbuzatu has quit IRC17:58
*** nwkarsten has joined #openstack-infra17:58
*** nwkarsten has quit IRC17:58
*** sambetts has joined #openstack-infra17:59
fungiprometheanfire: specifically working out how to reverse-map tags to branches so that we can apply the upper-constraints.txt from openstack-requirements when invoking tox to build release artifacts17:59
fungier, from openstack/requirements17:59
*** nwkarsten has joined #openstack-infra17:59
*** rbuzatu has joined #openstack-infra18:00
prometheanfirereverse map tags? as in figure out which branch it applies to, so you know which branch of requirements to pull UC from to build release artifacts?18:00
*** nwkarste_ has joined #openstack-infra18:01
*** tosky has quit IRC18:01
*** nwkarst__ has joined #openstack-infra18:02
AJaegerprometheanfire: yes. So, for neutron 9.0.0b2 we want master, but for 8.1.2 we want stable/mitaka HEAD of requirements repo18:02
*** nwkars___ has joined #openstack-infra18:02
rcarrillocruzis /etc/ansible/hostfile managed in git repo somewhere?18:03
prometheanfireso you need to crawl git history, and since branches can disapear as well, it gets more complex18:03
rcarrillocruzi can't seem to find where we populate it on puppet18:03
rcarrillocruzfungi: ^ ?18:03
*** nwkars___ has quit IRC18:03
*** nwkarsten has quit IRC18:04
openstackgerritBen Kero proposed openstack-infra/project-config: Move {name}-docs to ubuntu-xenial
prometheanfirefungi: AJaeger: git branch --contains tag18:04
prometheanfireor is it not that simple18:04
*** nwkars___ has joined #openstack-infra18:04
*** nwkarste_ has quit IRC18:05
*** asingh has joined #openstack-infra18:06
*** rbrndt has quit IRC18:06
*** nwkarst__ has quit IRC18:06
*** nwkars___ has quit IRC18:06
*** Goneri has joined #openstack-infra18:06
*** karthik__ has quit IRC18:06
*** njohnston has joined #openstack-infra18:06
*** karthik__ has joined #openstack-infra18:07
*** nwkarste_ has joined #openstack-infra18:07
*** nwkarste_ has quit IRC18:07
*** nwkarste_ has joined #openstack-infra18:08
*** cody-somerville has joined #openstack-infra18:09
rcarrillocruzseems a leftover, since ansible.cfg default points to /etc/ansible/hosts anyways18:10
*** julim has joined #openstack-infra18:10
*** nwkarste_ has quit IRC18:10
*** jesusaur has quit IRC18:10
*** EricGonczer_ has quit IRC18:11
*** nwkarsten has joined #openstack-infra18:11
*** csomerville has quit IRC18:11
fungiprometheanfire: it's almost that simple, but needs to get worked into the logic used by the job18:12
fungiso the devil is in the implementation18:12
prometheanfireshould we focus on talking here or in -releases? :P18:12
prometheanfirebut ya18:12
*** nwkarste_ has joined #openstack-infra18:13
fungithe discussion in -release was intended to keep the release managers up to date on the situation with release jobs, less to hash out implementation details18:13
karthik__sdague: Hi18:13
*** nwkarst__ has joined #openstack-infra18:14
fungiprometheanfire: for that, we have gerrit (once AJaeger pushes up his work in progress)18:14
prometheanfireah, cool18:14
*** nmagnezi has quit IRC18:15
*** nwkarsten has quit IRC18:16
mordredrcarrillocruz: yah - we don't use hostfile anyumore18:16
*** njohnston has left #openstack-infra18:16
AJaegerfungi, we have jobs that run in both post and release pipeline - like docs job. Do we need to duplicate the job?18:17
fungiprometheanfire: the other main complication is that we reuse the same logic currently for building branch artifacts in the post pipeline and release artifacts in the release pipeline18:17
*** nwkarste_ has quit IRC18:17
openstackgerritMerged openstack-infra/germqtt: Add docs explaining mqtt topic layout
*** e0ne has quit IRC18:19
fungiAJaeger: ideally we would reuse the same builder for both. i'm just trying to think if it gets that complicated to use $ZUUL_REFNAME that way18:19
AJaegerfungi, let me push what I have and then we can discuss that - nearly finished, just writing commit message18:20
*** gouthamr has quit IRC18:20
*** nwkarsten has joined #openstack-infra18:20
*** PollyZ has joined #openstack-infra18:21
fungiAJaeger: also thinking through whether this complicates zuul-git-prep-upper-constraints where we're relying on --branch to work around the lack of $ZUUL_BRANCH in some pipelines18:21
fungiAJaeger: i wonder if that predates Nakato's zuul-cloner improvements and can be dispensed with now?18:22
fungiAJaeger: we still need to use --branch in zuul-git-branch-prep-upper-constraints for obvious reasons, but its use can probably be restricted to periodic pipeline jobs only18:23
*** piet has joined #openstack-infra18:23
*** e0ne has joined #openstack-infra18:23
*** permalac has joined #openstack-infra18:23
*** nwkarst__ has quit IRC18:24
corvusi plan on altering the periodic pipelines to include a branch parameter sometime in zuulv318:24
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Setup constraints for release
corvusso eventually, we won't even need it there18:25
fungithat would seriously help make our jobs more consistent18:25
AJaegerfungi, corvus, there's a first patch up. I'm a bit tired, but this woke me up ;) Please review - and feel free to take over and rewrite if you have ideas on improvements...18:26
*** nwkarste_ has joined #openstack-infra18:26
* AJaeger is sure that some $ZUUL_REF/URL variables need to be passed in somehow.18:26
fungiAJaeger: thanks!18:26
corvusAJaeger, fungi: can we avoid using a jenkins script for this and use it as a macro directly?18:27
corvusi'd really hate to spend 3 days fixing a bug in this18:27
AJaegercorvus, sure, I can make this a macro...18:27
AJaegerAh, you don't want it baked into our images?18:27
AJaegerPlease review - and comment. Next iteration will be macro...18:28
*** nwkarst__ has joined #openstack-infra18:28
*** hurgleburgler has joined #openstack-infra18:28
fungistarting out as a macro for now would be good, yes. can move it to a separate script later once it's stabilized18:28
corvusfungi: is there any reason to do that?18:28
AJaegerback in a few minutes...18:28
fungicorvus: perhaps not. i had assumed so given that we have many other things that could be builder macros as scripts baked into our images instead18:29
*** nwkarsten has quit IRC18:29
corvusfungi: afaik, most of that dates to pre-jjb.  :)18:29
fungifor example, and could be builders18:29
corvus(or follows a tradition so dated)18:29
*** ayoung has quit IRC18:30
fungibut if the preference is to just keep things in macros unless they need to run on persistent job nodes, works for me18:30
openstackgerritMerged openstack-infra/tripleo-ci: Use delorean-current for puppet modules
corvusit's probably not worth moving any existing scripts until post-v318:30
corvusbut i don't think we need to make new ones without a good reason18:30
corvusjjb has an include feature too, so we can even still stick 'em in their own files i think18:30
mtreinishrcarrillocruz: I went to get lunch, did you get past that issue you were hitting before I stepped away?18:30
*** yottatsa has quit IRC18:31
*** nwkarste_ has quit IRC18:31
*** nwkarsten has joined #openstack-infra18:32
odyssey4mepabelanger any chance on getting merged some time soon?18:33
*** jcoufal has joined #openstack-infra18:33
*** nwkarste_ has joined #openstack-infra18:33
*** nwkarste_ has quit IRC18:33
odyssey4me(or any other project-config cores :))18:33
*** nwkarste_ has joined #openstack-infra18:34
*** gouthamr has joined #openstack-infra18:34
*** nwkarste_ has quit IRC18:34
rcarrillocruzsorry, not really around now, i'll recheck when pabelanger changed land, re: xenial blocking ssh18:34
rcarrillocruzi tried manually to run remote_puppet_else playbook from puppetmaster but it fails, since it can ssh to it18:34
rcarrillocruzi'll let you know18:34
*** nwkarste_ has joined #openstack-infra18:35
*** Sukhdev has joined #openstack-infra18:35
*** nwkarste_ has quit IRC18:35
*** nwkarst__ has quit IRC18:36
*** nwkarste_ has joined #openstack-infra18:36
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Setup constraints for release
mtreinishrcarrillocruz: ok, thanks for helping18:37
rcarrillocruzhate phone typing18:37
*** nwkarsten has quit IRC18:37
mtreinishheh, no worries. That's why I don't even attempt irc from the phone :)18:37
AJaegerfungi, corvus: converted to builder macro. I still wonder whether ZUUL_REF/ZUUL_URL etc. are setup correctly. Do I need to set or unset them for either of the zuul-cloner invocations?18:37
*** markvoelker_ has quit IRC18:38
openstackgerritMerged openstack-infra/tripleo-ci: Implement a new job for undercloud testing only
*** markvoelker has joined #openstack-infra18:39
*** markvoelker has quit IRC18:39
fungiAJaeger: ZUUL_URL should always be correct. also ZUUL_REFNAME should be the tag name (it's the fallback to ZUUL_REF we were seeing in the earlier job causing it to checkout the tag by its sha rather than by its name)18:40
AJaegerfungi, will it cause a problem if those are set for the requirements checkout invocation?18:41
*** whoops has joined #openstack-infra18:41
fungiAJaeger: zuul-cloner already expects they may not exist and has a series of fallbacks it looks for18:42
*** notnownikki has quit IRC18:42
*** karthik__ has quit IRC18:42
*** yamamoto has joined #openstack-infra18:43
AJaegerfungi, I mean: Those are set for the tagged-repo. Should we unset them for the second zuul-cloner call?18:43
*** karthik__ has joined #openstack-infra18:43
fungiAJaeger: the second zuul-cloner invocation will likely need to specify --branch since we need a means of communicating to it what requirements branch we've mapped to, but those variables won't cause a problem because zuul will see that they don't exist18:43
AJaegerfungi, yes, it has --branch.18:44
fungii need to look at 349627 first though and stop reviewing in my head what i think you've written18:44
AJaegerfungi, no hurry ;)18:44
pabelangerodyssey4me: working bring up tripleo-test-cloud-rh1, I should be able to move back to MairaDB last this week18:45
*** mdrabe has quit IRC18:46
*** yamamoto has quit IRC18:48
*** nwkarsten has joined #openstack-infra18:48
fungiAJaeger: yeah, that's fine. i think the place we can _drop_ --branch is in the zuul-git-prep-upper-constraints macro since zuul-cloner should now be smart enough to look at $ZUUL_REFNAME and $ZUUL_BRANCH18:49
AJaegerfungi, feel free to propose a change for that ;)18:51
fungiAJaeger: and i'm wondering if later on we should refactor zuul-git-prep-upper-constraints to handle the additional logic from 34962718:51
*** nwkarst__ has joined #openstack-infra18:51
fungionce we're sure it's working correctly18:51
*** nwkarste_ has quit IRC18:51
fungiso that we rely on mostly the same codepaths everywhere and can get rid of some of this duplication18:52
*** eric_lopez has quit IRC18:52
*** nwkarste_ has joined #openstack-infra18:53
AJaegerfungi, I can't think straight right now how to use this for post queue, that's for another day.18:53
*** nwkarste_ has quit IRC18:53
*** nwkarste_ has joined #openstack-infra18:54
*** nwkarsten has quit IRC18:55
*** doug-fish has joined #openstack-infra18:55
*** nwkarst__ has quit IRC18:56
*** nwkarst__ has joined #openstack-infra18:58
*** nwkarst__ has quit IRC18:58
*** nwkarste_ has quit IRC18:58
*** e0ne has quit IRC19:00
*** nwkarste_ has joined #openstack-infra19:00
*** nwkarste_ has quit IRC19:00
*** rbuzatu has quit IRC19:00
pabelangerfungi: I think it works, test ? is online and under the control of puppetmaster.o.o now19:00
pabelangerscroll wheel fail19:00
*** nwkarsten has joined #openstack-infra19:00
*** nwkarsten has quit IRC19:00
pabelangerfungi: AJaeger rcarrillocruz: do you mind reviewing 349659? tripleo-test-cloud-rh1 AFS mirror is onlin and under control of puppetmaster19:01
*** eric_lopez has joined #openstack-infra19:01
*** nwkarsten has joined #openstack-infra19:02
*** rbrndt has joined #openstack-infra19:02
*** jsavak has quit IRC19:02
*** nwkarsten has quit IRC19:03
*** jsavak has joined #openstack-infra19:03
*** e0ne has joined #openstack-infra19:04
*** rbuzatu has joined #openstack-infra19:04
*** nwkarsten has joined #openstack-infra19:04
*** nwkarsten has quit IRC19:05
*** nwkarsten has joined #openstack-infra19:05
openstackgerritMatt Borland proposed openstack-infra/project-config: Splitting current/deprecated integ. test
*** nwkarste_ has joined #openstack-infra19:06
*** nwkarste_ has quit IRC19:06
*** fifieldt has quit IRC19:07
*** nwkarste_ has joined #openstack-infra19:07
*** nwkarste_ has quit IRC19:07
*** pvaneck has joined #openstack-infra19:08
*** nwkarste_ has joined #openstack-infra19:08
*** nwkarste_ has quit IRC19:09
*** mdrabe has joined #openstack-infra19:09
*** nwkarste_ has joined #openstack-infra19:09
*** Sukhdev has quit IRC19:09
*** nwkarste_ has quit IRC19:10
*** nwkarsten has quit IRC19:10
*** Sukhdev has joined #openstack-infra19:10
*** nwkarsten has joined #openstack-infra19:10
*** nwkarsten has quit IRC19:11
*** mmandel has quit IRC19:11
*** mmandel has joined #openstack-infra19:11
*** barnaby has joined #openstack-infra19:12
*** nwkarsten has joined #openstack-infra19:13
*** nwkarsten has quit IRC19:13
*** nwkarsten has joined #openstack-infra19:13
AJaegerinfra-root, there's an email on our mailing list mentioning that is not mirrored to - I just confirmed this. Could one of you look at it?19:14
*** nwkarsten has quit IRC19:14
*** harlowja has quit IRC19:14
AJaegerThis is one of the reasons why I prefer no mirroring - it's not reliable to setup ;(19:14
*** EricGonczer_ has joined #openstack-infra19:14
*** nwkarsten has joined #openstack-infra19:14
*** nwkarsten has quit IRC19:14
*** nwkarsten has joined #openstack-infra19:15
*** nwkarsten has quit IRC19:15
*** harlowja has joined #openstack-infra19:15
fungiAJaeger: fixed. as usual manage-projects was unable to successfully grant gerrit's account permission to push into it for mirroring19:16
*** piet has quit IRC19:16
AJaegerfungi, thanks. Will you answer the email or shall I?19:16
*** nwkarsten has joined #openstack-infra19:17
fungiAJaeger: i can. just wrapping up another19:17
*** sdake has joined #openstack-infra19:18
* AJaeger waves good night...19:18
*** nwkarste_ has joined #openstack-infra19:18
*** nwkarste_ has quit IRC19:18
*** fifieldt has joined #openstack-infra19:18
*** achanda has quit IRC19:19
rcarrillocruzhave a good one AJaeger19:20
*** nwkarste_ has joined #openstack-infra19:21
*** mriedem1 has joined #openstack-infra19:21
openstackgerritMerged openstack-infra/project-config: Revert "Disable tripleo-test-cloud-rh1 until we have AFS mirrors"
*** nwkarst__ has joined #openstack-infra19:21
*** nwkarsten has quit IRC19:22
*** nwkarst__ has quit IRC19:23
*** mriedem has quit IRC19:23
*** nwkarsten has joined #openstack-infra19:24
*** nwkarsten has quit IRC19:24
asinghHi All. i have , it has 2 +2 and +1 workflow , the dependent patch has merged few days back, not sure why is not getting merged . can someone take a look at it?19:24
mtreinishrcarrillocruz: heh, at the speed the gates moving it'll be a while before any of those patches land19:24
rcarrillocruzand it randomly failed the last test19:24
*** nwkarsten has joined #openstack-infra19:25
rcarrillocruzi just rechecked19:25
* rcarrillocruz waves fist19:25
*** nwkarsten has quit IRC19:25
*** nwkarste_ has quit IRC19:25
*** nwkarsten has joined #openstack-infra19:25
*** nwkarsten has quit IRC19:26
rcarrillocruzdependent change got +1 tho19:26
*** ihrachys has joined #openstack-infra19:26
fungiasingh: 339177 was approved while its dependency was still under review. now that the dependency has merged you need a new approval vote on it19:26
odyssey4meinfra-roots an extra review of would be appreciated19:27
odyssey4mefungi has already made his mark19:27
fungimtreinish: someone has a 33-change series in nova they keep rebasing, so it's been quite the hog on check pipeline resources19:27
*** nwkarste_ has joined #openstack-infra19:27
*** nwkarste_ has quit IRC19:27
*** nwkarste_ has joined #openstack-infra19:28
*** nwkarste_ has quit IRC19:28
mtreinishfungi: ugh, we should get mriedem1 to lock that down19:29
*** nwkarste_ has joined #openstack-infra19:29
mriedem1that would be mdbooth19:29
*** mriedem1 is now known as mriedem19:29
*** nwkarste_ has quit IRC19:29
*** nwkarste_ has joined #openstack-infra19:30
*** nwkarste_ has quit IRC19:30
*** nwkarsten has joined #openstack-infra19:32
asinghthanks fungi: is it okay to have same core reviewers that have reviewed the patch?19:32
*** nwkarsten has quit IRC19:32
mtreinishmriedem: sure, but I was asking you to wield your ptl influence and tell mdbooth to try not to flood the gate with constant rebases19:32
fungiasingh: sure, any core reviewer can just add a new workflow +1 to that change, or the same core reviewer who did it before can set it to workflow +0 and then back to workflow +1 again19:32
*** nwkarsten has joined #openstack-infra19:33
*** nwkarsten has quit IRC19:33
mriedemmtreinish: he'll probably just say nova core should review it faster :)19:33
mriedemdansmith: ^19:33
fungimriedem: or approve it faster!19:34
*** nwkarsten has joined #openstack-infra19:34
dansmithmriedem: yeah19:34
*** nwkarsten has quit IRC19:34
fungi(merged changes need no rebasing!)19:34
dansmithmtreinish: mriedem: last week I asked diana to split the bottom chunk off, which she did, but he probably stitched it back together19:35
*** nwkarsten has joined #openstack-infra19:35
*** nwkarsten has quit IRC19:35
pabelangerianw: regarding ntp-wait, do you also need to restart ntpd on ubuntu to make it work?19:36
*** karthik__ has quit IRC19:36
*** nwkarsten has joined #openstack-infra19:36
*** karthik__ has joined #openstack-infra19:37
fungiin theory ntp-wait itself doesn't rely on ntpd being restarted. it's just a utility to watch ntpd's state and determine when the time has properly updated so that other tasks dependent on having correct time can run19:37
openstackgerritMerged openstack-infra/tripleo-ci: Enable Ceph in OVB/HA2
*** nwkarsten has quit IRC19:37
fungiso that observation likely indicates there's something breaking ntpd itself early in boot19:38
*** nwkarsten has joined #openstack-infra19:38
fungiianw's conjecture about lack of network setup early enough in boot seems like a possibility, i agree19:38
openstackgerritMonty Taylor proposed openstack-infra/zuul: Add an internals doc to the documentation
pabelangerfungi: right, so that is what I am seeing:
openstackgerritMonty Taylor proposed openstack-infra/zuul: Rename BasePipelineManager to PipelineManager
openstackgerritMonty Taylor proposed openstack-infra/zuul: Move code out of zuul/manager/
mordredcorvus, Shrews: I believe I have taken care of your reviews19:40
*** nwkarste_ has joined #openstack-infra19:40
*** nwkarste_ has quit IRC19:40
fungipabelanger: ahh, yep, dns resolution failing at ntpd start, and it doesn't reevaluate its configuration to properly resolve later19:41
*** nwkarste_ has joined #openstack-infra19:41
fungipabelanger: so ideally ntpd wouldn't start until dns resolution is working properly19:41
*** nwkarste_ has quit IRC19:41
pabelangerfungi: yup, so is systemd configure for centos-719:42
*** nwkarste_ has joined #openstack-infra19:43
*** nwkarsten has quit IRC19:43
*** jsavak has quit IRC19:43
fungipabelanger: do any of those units listed in Unit.After depend on networking?19:43
*** jsavak has joined #openstack-infra19:44
pabelangerfungi: ntpdate is the only one install, and it is stopped19:44
pabelangerso that looks to be our issue19:44
*** nwkarsten has joined #openstack-infra19:44
fungiwondering if network hotplug is adding interfaces after ntpd starts, or if there's something else external to the instance which is causing dns lookups to fail soon after boot19:44
pabelangeractually, rsyslog is installed but it is also missing network target19:45
*** yamamoto has joined #openstack-infra19:45
*** nwkarsten has quit IRC19:45
mordredfungi: remember back when all of your hardware existed at boot time and also when boot was serial?19:46
*** nwkarst__ has joined #openstack-infra19:46
*** nwkarst__ has quit IRC19:46
fungimordred: i still have dreams about it, yes19:46
mordredfungi: I'm so glad that "progress" has brought us all of the new "tools" to make our "lives" "better"19:47
fungimordred: we no longer need lives. systemd has a replacement for that19:47
*** nwkarste_ has quit IRC19:47
mordredfungi: oh right. I forgot19:48
*** nwkarste_ has joined #openstack-infra19:48
mordredfungi: it probably doens't log reliably either does it?19:48
*** eharney has quit IRC19:48
fungimine never logged well to start with, so no great loss19:48
*** nwkarst__ has joined #openstack-infra19:48
*** eharney has joined #openstack-infra19:49
*** nwkarst__ has quit IRC19:49
fungijust reading an ml thread on debian-devel wherein md provides high praise for systemd-networkd as "nowadays usable"19:49
openstackgerritBen Nemec proposed openstack-infra/tripleo-ci: Delete ping test environment in periodic jobs
fungi(the ifupdown2 bashing thread)19:49
*** nwkarsten has joined #openstack-infra19:50
*** nwkarst__ has joined #openstack-infra19:50
*** yamamoto has quit IRC19:50
*** nwkarst__ has quit IRC19:51
mordredfungi: I will remind us all that slackware still uses sysvinit19:51
*** coolsvap has quit IRC19:51
*** nwkarst__ has joined #openstack-infra19:51
mtreinishfungi: I still don't have bonded networks starting up properly at home...19:52
*** nwkarste_ has quit IRC19:52
*** nwkarste_ has joined #openstack-infra19:53
fungimtreinish: yeah, that is actually pretty easy to do in ifupdown19:53
fungidefine your interfaces, define your bond interface, done19:53
*** nwkars___ has joined #openstack-infra19:53
phschwartzhas anyone seen an issue where you have a min-ready set for a label in nodepool  and a much greater max-servers set for the providers, but you never get more nodes then the min-ready?19:54
fungii used to have an ifupdown config for my xen servers that created 802.1q vlan subinterfaces on bonds across physical interfaces, bound them into a bridge and added virtual interfaces to the bridge19:54
mordredfungi: but that's not modern and event-driven!19:54
*** nwkarsten has quit IRC19:54
mtreinishfungi: heh, yep I have a bash script that does that. But I've never gotten it to work as part of the startup process with systemd.19:55
*** Apoorva has quit IRC19:55
fungiphschwartz: that generally means that nodepool is unable to track the demand in gearman19:55
*** nwkarst__ has quit IRC19:56
*** shardy has quit IRC19:56
*** dkranz has joined #openstack-infra19:56
*** nwkars___ has quit IRC19:57
*** nwkarste_ has quit IRC19:57
*** jesusaur has joined #openstack-infra19:57
*** nwkarsten has joined #openstack-infra19:57
phschwartzfungi: is there a way to check that and to see the current demand it sees?19:58
*** nwkarste_ has joined #openstack-infra19:58
rcarrillocruznothing personal against upstart, but i sobbed when i looked at that diagram when fixing a glean init issue...19:59
mordredI mean19:59
fungiphschwartz: /var/log/nodepool/debug.log19:59
*** javeriak has quit IRC19:59
mordredrcarrillocruz: you don't happen to know how to install a python3 module with pip in our puppet do you?20:00
rcarrillocruzhmm, i'm afraid i haven't battled with that no20:01
rcarrillocruzwhat is it about20:01
phschwartzfungi: also, have you ever seen an issue were 2 nodes built from the same image have an issue were 1 will allow sudo and 1 will not (revoke-sudo builder is not used and the failure/pass are both on the same job)20:01
*** jsavak has quit IRC20:01
mordredwell, I started hacking on a change to switch our cgit to using pygments for syntax highlighting20:01
*** e0ne has quit IRC20:01
mordredrcarrillocruz: but the pygments helper script that comes with cgit depends on python3 for some reason20:02
*** nwkarste_ has quit IRC20:02
*** nwkarsten has quit IRC20:02
*** jsavak has joined #openstack-infra20:02
mtreinishmordred: ? :)20:02
rcarrillocruzif you have a change up or some paste, link me20:03
rcarrillocruzi can look with better eyes in my morning20:03
mordredrcarrillocruz: OH! well, it lies. it doesn't need python320:03
*** nwkarsten has joined #openstack-infra20:03
mordredit just has python3 in its first line20:03
mordredmtreinish: yah - but there's no pip3 on the box either20:03
*** nwkarsten has quit IRC20:04
rcarrillocruzmordred: btw, how's your ansible image builder role? did you have time to look at it? if not, i'm happy to pull it from your queue of stuff20:04
rcarrillocruzi.e. forking and pushing a PR on your github or something20:04
mordredrcarrillocruz: that's a great question :)20:04
*** nwkarsten has joined #openstack-infra20:04
rcarrillocruzprometheanfire was asking something related to it the other day20:04
rcarrillocruzabout 'why o why use puppet on our images'20:04
*** jcoufal_ has joined #openstack-infra20:05
*** nwkarsten has quit IRC20:05
pabelangerfungi: so, I think our fix is to also enable ntpdate service, at least in centos-7. I'm just testing that in a few regions before suggesting a patch20:05
pabelangerianw: ^20:05
mordredrcarrillocruz: I haven't hacked on it in a while -so feel free to jump on it20:05
rcarrillocruzk, i'll sniff around your github repos and find it20:05
corvusrcarrillocruz: i don't expect us to use ansible instead of puppet to build images20:06
corvusrcarrillocruz: we're moving to using dib elements20:06
*** e0ne has joined #openstack-infra20:06
*** nwkarste_ has joined #openstack-infra20:06
rcarrillocruzcorvus: so the plan is to patch stuff that we really need puppet today with elements20:07
*** jcoufal has quit IRC20:07
corvusrcarrillocruz: we need very little puppet for a nodepool worker image20:07
*** azweb76 has joined #openstack-infra20:08
rcarrillocruzwell, that's great cos then we just use ansible to do the glance upload20:08
rcarrillocruzhow are you folks uploading these days images20:08
rcarrillocruzplain glance commands20:08
corvusrcarrillocruz: nodepool20:08
fungiphschwartz: does not sound familiar, no. though i have seen strangeness in years past where we would ask nova to boot an instance from a snapshot and it would seemingly boot from another (often broken) snapshot instead which seemed like one we had deleted days or weeks before. like nova's image cache table getting corrupted somehow20:08
rcarrillocruzhmm, ok, need to look, haven't looked at this stuff, we still had snapshot based images in Gozer20:09
phschwartzfungi: hmm20:09
phschwartzfungi: that is interesting20:09
*** HeOS has joined #openstack-infra20:09
*** yottatsa has joined #openstack-infra20:11
*** nwkarsten has joined #openstack-infra20:11
rcarrillocruzo-k, those are a few verbs i wasn't aware of...20:12
*** Apoorva has joined #openstack-infra20:12
* rcarrillocruz reads nodepool docs avidly20:12
*** nwkarst__ has joined #openstack-infra20:12
*** e0ne has quit IRC20:12
openstackgerritThiago Paiva Brito proposed openstack-infra/puppet-openstackci: Pinning JJB and Zuul for 3rd party CI
*** nwkars___ has joined #openstack-infra20:14
*** nwkarste_ has quit IRC20:15
*** mriedem1 has joined #openstack-infra20:16
*** nwkarsten has quit IRC20:16
openstackgerritMonty Taylor proposed openstack-infra/system-config: Use pygments for syntax highlighting for git.o.o
*** nwkarsten has joined #openstack-infra20:16
mordredpleia2: ^^ I _think_ that should be good when you awaken20:17
mordredpleia2: or, rather... I think that should be good. you may enjoy looking when you awaken20:17
mordredits quality has no relationship to your state of sleeping20:17
*** nwkarst__ has quit IRC20:18
*** nwkarste_ has joined #openstack-infra20:18
*** nwkars___ has quit IRC20:19
*** mriedem has quit IRC20:19
*** nwkarsten has quit IRC20:21
pabelangerfungi: okay, some testing seems to should that we should also enable the ntpdate service on centos-7 to setup the proper dependencies for ntpd20:21
*** nwkarsten has joined #openstack-infra20:24
fungipabelanger: that sounds like something that needs fixing upstream in fedora/rhel too20:25
fungisince people besides us may want to stop relying on the (officially deprecated) ntpdate utility in favor of ntp-wait20:25
pabelangerfungi: Agreed. I plan on submitting a fix into fedora for sure20:26
corvusmordred: does that need or already have a stylesheet?20:26
*** ayoung has joined #openstack-infra20:26
fungialternatively, is there a simple way to declare ntpdate as an alias for networking from systemd's dependency handling perspective?20:26
pabelangerI am not sure right now20:26
*** nwkarste_ has quit IRC20:28
fungior we could simply install a custom ntpd unit that has the corrected dependency we expect20:28
fungipabelanger: basically a backport of whatever fix you're going to propose upstream20:29
*** amitgandhinz has quit IRC20:29
*** amitgandhinz has joined #openstack-infra20:30
pabelangerfungi: Right, so we can do 2 steps, enable ntpdate service for centos-7, fix upstream fedora. Once patch lands in fedora, revert ntpdate service and run our own custom ntpd20:30
*** nwkarste_ has joined #openstack-infra20:31
pabelangeralso open a bug into centos and see if we can backport service file there too20:31
*** nwkarst__ has joined #openstack-infra20:32
*** jesusaur has quit IRC20:32
*** Apoorva_ has joined #openstack-infra20:33
*** nwkarsten has quit IRC20:34
*** Jeffrey4l has quit IRC20:35
*** nwkarsten has joined #openstack-infra20:35
*** nwkarste_ has quit IRC20:35
*** Jeffrey4l has joined #openstack-infra20:35
fungiall that sounds reasonable20:35
*** dkranz has quit IRC20:35
*** Apoorva has quit IRC20:36
*** rbuzatu has quit IRC20:36
corvusmordred: also, we're supposed to have syntax highlighting now, but i don't see it.  do you know why?20:37
*** bhavik has quit IRC20:37
*** nwkarste_ has joined #openstack-infra20:38
mordredcorvus: it already has a stylesheet, although I believe we can provide one to override20:38
*** honza has quit IRC20:38
*** nwkarst__ has quit IRC20:38
mordredcorvus: and no, we do not know why it's broken currently - that is a thing pleia2 was going to look in to in the morning20:38
openstackgerritPaul Belanger proposed openstack-infra/system-config: Enable ntpdate service for centos-7
pabelangerfungi: ianw: ^20:39
corvusmordred: yeah, i wonder if the output from pygments will match the existing stylesheet or if we'll need a new one?20:40
mordredcorvus: fwiw, is an example of running the pygments highlighter on the script that is the highlighter20:40
*** nwkarst__ has joined #openstack-infra20:40
rcarrillocruzpabelanger: not sure if you're gonna get linter issue due to rockets misalignment?20:40
*** nwkarsten has quit IRC20:41
corvusmordred: the current thing, if it worked, would output things like "<span class="hl kwa">print</span>"20:41
rcarrillocruznot sure if we have that setting on the linter job at the gate20:41
pabelangerrcarrillocruz: shouldn't because it is system-config20:41
pabelangerbut I can quickly fix20:41
rcarrillocruzah right20:41
rcarrillocruzi hate rocket things20:41
rcarrillocruzplease no :D20:41
*** honza has joined #openstack-infra20:41
fungiyeah, we don't do style enforcement on system-config because we don't intend to put it on puppetforge20:41
corvusmordred: looks like pygments outputs '<span class="k">' for the same20:42
rcarrillocruzand i'm glad we do then20:42
fungipabelanger: rcarrillocruz: as long as it's readable and maintainable, i don't have any style concerns20:42
*** honza is now known as Guest5282320:42
rcarrillocruzi find it one of the most irritating things to get -1 about20:42
corvusmordred: so i expect we would need a stylesheet update20:42
*** vvalderrv1 has joined #openstack-infra20:42
mordredcorvus: ya. so looks like we'd need to rework stylesheet ... yeah20:42
cjvvalderrv1: howdy!20:42
cjzxiiro: I've recommended that vvalderrv1 read up on jjb.  Is there a canonical primer?20:43
rcarrillocruzok, so , the dib images in nodepool. is there a cron that upload them periodically or we do that manually?20:43
*** nwkarsten has joined #openstack-infra20:43
fungipabelanger: 349613 failed gate-infra-puppet-apply-centos-7 (i haven't looked into why)20:44
corvusrcarrillocruz: nodepool uploads images when it finishes building them20:44
corvusrcarrillocruz: and it re-builds them every day20:44
*** nwkars___ has joined #openstack-infra20:44
*** nwkars___ has quit IRC20:44
*** nwkarste_ has quit IRC20:44
*** nwkarste_ has joined #openstack-infra20:46
fungi188 out of 798 votes cast so far for infra mascot with just over 3 hours to go20:46
*** yamamoto has joined #openstack-infra20:46
* cj checks mail20:46
*** mriedem1 is now known as mriedem20:47
rcarrillocruzi guess it consumes the same nodepoo.conf setting as snapshot based images updates then20:47
fungicj: you would have gotten a civs ballot from me if you're an infra contributor over the past 12 months20:47
*** nwkarst__ has quit IRC20:47
rcarrillocruzmordred, so did you start your ansible build/upload role before this was in nodepool?20:47
rcarrillocruzor you have another use case in mind?20:47
cjhurm... I wonder if any of my changes have been accepted yet...20:47
fungicj: From: Jeremy Stanley <>; Subject: Poll: OpenStack Infra Team Mascot20:47
mordredrcarrillocruz: ah! yes - the ansible role was for us to make base images for infra20:47
mordredso that we would not be using cloud-provider images in our long-lived machines, but would have an automated way to manage it20:48
fungicj: sent tuesday of last week20:48
cjNope.  looks like I'm not in the crowd20:48
*** thiagolib has quit IRC20:48
*** nwkarsten has quit IRC20:48
pabelangerfungi: possible, it was untested20:48
rcarrillocruzubuntu-infra-minimal-foo etc20:48
fungicj: get a change in soon and you can vote me out of the ptl position! ;)20:48
mordredrcarrillocruz: yah20:48
rcarrillocruzmakes sense now20:48
mordredrcarrillocruz: I think once we get into that the two will become related20:49
cjfungi: this one's nearly ready...20:49
*** nwkarst__ has joined #openstack-infra20:49
*** nwkarst__ has quit IRC20:49
mordredbecause one of the reasons we puppet our nodepool images right now is to get our admin accounts on them and stuff20:49
*** Goneri has quit IRC20:49
mordredso figuring out a good plan for dealing with duplication of effort without making anything too difficult20:49
cjI need a test environment to exercise the IRC comms, but once I get tests passing, I think it'll be good to go.20:49
mordredis part of the hard part here20:49
pabelangerfungi: oh, 349613, looks like zuul-launcher aborted the job. let me check that20:49
fungicj: ahh, yep. i started to try 287056 out a while back and realized we'd need to convert our config first20:50
*** dprince has quit IRC20:50
*** rfolco_ has quit IRC20:50
fungicj: looks like pleia2 has since added a change for that part20:50
*** shashank_hegde has joined #openstack-infra20:50
*** nwkarste_ has quit IRC20:50
cjoh, great!  I had thought about a conversion script or a deprecated old-config loader, but haven't had time to implement.20:50
fungithough it still needs a little work apparently20:50
fungibut yes, close20:51
pabelangerfungi: looks like slow git clones from ovh-bhs120:51
*** nwkarste_ has joined #openstack-infra20:51
*** nwkarste_ has quit IRC20:51
fungipabelanger: strange. i wouldn't think they'd stop packets at the border, but those mounties are getting more vigilant20:51
zxiirocj: just the main documentation20:51
cjzxiiro: okay.  url?20:51
cjvvalderrv1: ^^20:51
*** nwkarste_ has joined #openstack-infra20:51
*** hrybacki is now known as hrybacki|appt20:51
corvusjhesketh: i was thinking more about your comment on the test changes, and now i'm thinking about subclassing the ansible launcher to do everything *except* run ansible... is that what you were suggesting?20:52
*** nwkarste_ has quit IRC20:52
cj prolly, eh?20:52
zxiirocj vvalderrv1
*** yamamoto has quit IRC20:52
cjgreat, the docs are published well then :-)20:52
zxiirocj vvalderrv1 it's a holiday today so my responses will ve slow20:52
*** siva_krish has quit IRC20:52
*** nwkarste_ has joined #openstack-infra20:53
cjaw, thanks for checking in, then!  Sorry to keep you from your vacating!  :-)20:53
*** nwkarste_ has quit IRC20:54
*** nwkarste_ has joined #openstack-infra20:55
*** nwkarste_ has quit IRC20:55
openstackgerritBen Nemec proposed openstack-infra/tripleo-ci: Enable net-iso for ovb jobs
zxiirono problem. the jjb documentation is pretty good I think. plus our intern's been beefing it up all summer :)20:56
mgagnemordred: looking at shade, is there a way to actually list images and filters? I only found get_image() in doc, I found list_images() in source but there is no filters argument :-/20:56
*** nwkarste_ has joined #openstack-infra20:56
*** nwkarste_ has quit IRC20:56
mgagnemordred: ok, found "search_images" :O20:56
*** akshai has quit IRC20:57
*** siva_krish has joined #openstack-infra20:57
openstackgerritPaul Belanger proposed openstack-infra/nodepool: Default config-drive to true
mgagnesorry for the disturbance :D20:57
*** nwkarste_ has joined #openstack-infra20:57
rcarrillocruzyeah, in general20:57
rcarrillocruzall objects have list and search20:57
zxiiroI'm gonna try to convince him to stay on as a reviewer after his internship is over20:57
rcarrillocruzif not , we should add them20:57
*** nwkarste_ has quit IRC20:57
openstackgerritJoshua Harlow proposed openstack-infra/err2d2: Add basic README.rst and LICENSE
fungipabelanger: i've approved 349692, but the inline code comment could use some followup clarification pointing out that ntpd has an implicit dependency on networking that's getting fulfilled indirectly through ntpdate's systemd unit, but that the dependency should be made explicit upstream and backported. maybe can update that comment once you have an upstream bug url tracking it?20:58
openstackgerritJoshua Harlow proposed openstack-infra/err2d2: Add basic README.rst and LICENSE
mgagnercarrillocruz: well, I found so inconsistencies where some calls have "filters" and others don't20:58
pabelangerfungi: Agreed. I'll follow up with a new commit once I have upstream bugzilla in place20:58
mgagneso I'm not sure if it's on purpose or not20:58
rcarrillocruzyeah, cos some have sense to have filters20:59
*** nwkarsten has joined #openstack-infra20:59
rcarrillocruzwhereas others not20:59
rcarrillocruzand well... it's shade20:59
*** nwkarsten has quit IRC20:59
*** Goneri has joined #openstack-infra20:59
*** karthik__ has quit IRC20:59
*** thiagop has quit IRC20:59
*** nwkarsten has joined #openstack-infra21:00
*** nwkarsten has quit IRC21:00
*** jsavak has quit IRC21:00
*** nwkarsten has joined #openstack-infra21:01
*** jesusaur has joined #openstack-infra21:01
*** amitgandhinz has quit IRC21:01
*** amitgandhinz has joined #openstack-infra21:01
*** yottatsa has quit IRC21:01
*** nwkarste_ has joined #openstack-infra21:02
*** nwkarste_ has quit IRC21:02
*** nwkarste_ has joined #openstack-infra21:04
*** nwkarsten has quit IRC21:05
*** nwkarste_ has quit IRC21:05
*** nwkarsten has joined #openstack-infra21:06
*** raildo has quit IRC21:06
*** nwkarsten has quit IRC21:06
*** nwkarste_ has joined #openstack-infra21:08
*** thorst has quit IRC21:08
*** nwkarste_ has quit IRC21:08
*** thorst has joined #openstack-infra21:09
mgagneso it looks like create_image doesn't like when you have 2 images with the same name :-/21:09
mordredmgagne: yah. it does not like that21:10
mgagnewhy is that?21:10
mordredmgagne: because nobody has yet given a compelling use case as to why that should be allowed and what the workflow should look like vs. just not letting it happen21:11
*** nwkarste_ has joined #openstack-infra21:11
pabelangerfungi: mordred: have a moment to review: adds debian-openstack source.list files into configure_mirror.sh21:11
mordredmgagne: the simple case for images is "I want to boot the image named foo"21:11
mgagnemordred: well, that's my workflow and I can't migrate to shade because of that21:11
mordredmgagne: can you explain how it works for you?21:12
*** jkilpatr has quit IRC21:12
mgagnemordred: it's a similar system to the one found at RAX where old images are preserved and only marked "hidden". Image always has the same name.21:13
*** thorst has quit IRC21:13
mgagneand only one is visible to the end user.21:14
fungipabelanger: one concern with 341856, see inline21:14
mgagneso this means I can't "stage" the new image because it has the same name as an existing one.21:14
mordredmgagne: gotcha. so - for images currently the de-duplication is in service of avoiding uploading multiple times ... I think if we added a "force_upload" parameter or something that woudl skip the duplicate checks, the rest of the things should work21:15
mgagnewould force_upload override the existing one?21:15
*** ldnunes has quit IRC21:15
fungipabelanger: also i have a feeling that until you start signing the debian mirror, it'll be unusable without the same bit from the ubuntu conditional that turns off signature checking21:15
mordredmgagne: no - glance is perfectly happy to let you upload multiples21:16
*** r0ckS0LiD has joined #openstack-infra21:16
mgagnemordred: or should it be named "allow_duplicate_names" instead?21:16
mordredmgagne: yah - maybe so21:16
mgagnemordred: well, my goal is not to override the existing one21:16
mordredI'm with you21:16
fungipabelanger: are any jobs even using the plain (not openstack-newton) debian mirror? if we're building it with reprepro i'm unsure how it's working for any jobs21:16
mgagneall of this because somehow, glanceclient broke itself and I can't upload with the script I've been using for years or maybe someone in our team changed something :-/21:17
*** sdake has quit IRC21:20
*** kzaitsev_mb has joined #openstack-infra21:20
openstackgerritMonty Taylor proposed openstack-infra/shade: Add ability to upload duplicate images
mordredmgagne: something like that ^^21:21
*** skipp has quit IRC21:22
mgagnemordred: I think allow_duplicates would be more appropriate21:22
openstackgerritMonty Taylor proposed openstack-infra/shade: Add ability to upload duplicate images
mordredmgagne: with the s instead of without?21:22
mgagnethat's the most common name I found on google21:22
openstackgerritMonty Taylor proposed openstack-infra/shade: Add ability to upload duplicate images
*** stew925 has quit IRC21:23
mordredmgagne: k. I'm pretty sure that's the only place where we're doing the enforcement21:23
*** flepied1 has quit IRC21:23
*** eeiden has quit IRC21:24
mordredmgagne: ACTUALLY - reading through it again, it looks like it should only complain if the md5 and sha256 checksums of the file you have locally match the ones of the image in the cloud ... did you get the error "image {name} exists and is up to date" ?21:25
mgagnehmm let me see21:25
mgagne"shade.exc.OpenStackCloudException: Multiple matches found for Ubuntu 16.04 LTS (Xenial Xerus)"21:25
mordredmgagne: can you paste me your script?21:26
mordredmgagne: and then the larger traceback?21:26
mgagnegonna take some time21:26
mordred(just want to make sure I'm solving your actual problem)21:26
mordredah - ok. well, let's skip that then ...21:26
mordredis it easier to paste the traceback?21:26
mordredaha. ok. I grok21:27
mordredmgagne: yes - the patch I just made should fix the problem for you21:27
mgagnein fact, I will checkout and test21:27
mordredbut I'm going to think about it for just a few to see if there is a 'better' way to solve the same problem (I think not, but let me check anyway)21:28
corvusmgagne: do you have a cgit stylesheet for pygments?21:28
mgagnecorvus:  I don't think so?21:28
mgagneif I had one, it's long gone21:28
corvusmgagne: just thought i'd ask since you were talking about it just yesterday.  or maybe it was almost 3 years ago, but seems like just yesterday.  :)
mgagneand I think it was for gitweb21:29
mgagnehehe, I remember =)21:30
mgagnewould have to dig and try to find if I have it somewhere21:30
corvusmgagne: well, you also say that pygments can generate the stylesheet too... so that may be easiest21:30
corvusmordred: ^21:30
mordredcorvus: I mean, it generates a stylesheet when it runs21:31
mordredcorvus: if we want to colorize things differently we need a stylesheet21:31
mgagnethat should be it I guess21:31
corvusmordred: oh, it embeds a stylesheet in the output?21:31
mgagneI think it's full of <span> by default21:31
mordredcorvus: that's what the output of the script is21:32
mordredcorvus: sorry - I misunderstood what you were talking about earlier21:32
corvusmordred: ah i see21:32
*** stevebaker has joined #openstack-infra21:32
mgagne"I alright hacked gitweb to use pygments, shouldn't be hard to do the same with cgit" is what I said21:33
corvusmordred: so that will *work* but be inefficient since it includes 4k of stylesheet in every page, yeah?21:33
mgagnebash history doesn't go that far =(21:33
mordredcorvus: ish? it's only 61 lines of stylesheet - so it's really not _that_ inefficient21:33
*** ajmiller_ has quit IRC21:33
corvusmordred: i did not make up the number 4k, fwiw.  :)21:34
mgagnecorvus: does it help?
mordredcorvus: oh, that's actually 4k? :)21:34
mgagnepygmentize -S <style-name> -f html21:34
corvus4310 bytes21:34
*** Apoorva_ has quit IRC21:34
*** thorst has joined #openstack-infra21:34
mgagneit was a simple command, nothing fancy21:34
*** Apoorva has joined #openstack-infra21:35
mordredcorvus: we can split out that with a different wrapper script and only emit a link at the top21:35
mordredcorvus: that would be fairly trivial21:35
*** nwkarsten has joined #openstack-infra21:35
mordredcorvus: I'll amend my patch in just a bit and try that21:36
corvusmordred: we should already have a stylesheet, so if we can just add those 61 lines to it, can we ask pygments not to output the stylesheet?21:36
mordredcorvus: yah21:36
*** eeiden has joined #openstack-infra21:36
*** rbuzatu has joined #openstack-infra21:37
*** nwkarste_ has quit IRC21:37
ianwpabelanger: cool21:38
*** thorst has quit IRC21:38
*** nwkarsten has quit IRC21:38
*** nwkarsten has joined #openstack-infra21:39
mgagnemordred: image is uploading atm, no duplicate errors so far21:40
*** skipp has joined #openstack-infra21:41
*** psilvad has quit IRC21:41
*** rbuzatu has quit IRC21:41
mordredmgagne: woot!21:42
*** mriedem has quit IRC21:42
*** matt-borland has quit IRC21:43
*** nwkarste_ has joined #openstack-infra21:43
*** rosmaita has joined #openstack-infra21:43
*** eharney has quit IRC21:44
mgagnereally? :-/ Image creation failed: Unable to set 'virtual_size' to '10737418240'. Reason: '10737418240' is not of type u'null', u'integer'21:44
*** nwkarsten has quit IRC21:44
*** weshay is now known as weshay_bbiab21:44
*** ihrachys has quit IRC21:44
*** _ari_ has quit IRC21:44
ianwfungi / pabelanger: added a comment to ; there's nothing really to fix as such, ntpd is working as expected.  basically if the clock isn't already in sync when ntpd starts, you're going to have to wait.  chronyd & systemd's ntp trade-off clock stability / usability in a way that suits us better, i think21:46
mgagnemordred: this is caused by this piece of code:
*** burgerk has quit IRC21:46
mgagnemordred: which cast properties as str() unless the prop is found in a list of "known" integer parameters21:46
mgagnemordred: this means an operator cannot introduce custom integer properties without patching shade ^^'21:47
*** ajmiller has joined #openstack-infra21:47
*** armax has quit IRC21:48
*** yamamoto has joined #openstack-infra21:48
*** jamielennox|away is now known as jamielennox21:48
*** dimtruck is now known as zz_dimtruck21:50
*** ajmiller has quit IRC21:52
openstackgerritMonty Taylor proposed openstack-infra/system-config: Use pygments for syntax highlighting for git.o.o
mordredcorvus: ^^ that21:53
*** sdake has joined #openstack-infra21:53
*** sdake has quit IRC21:53
*** sdake has joined #openstack-infra21:53
*** yamamoto has quit IRC21:54
*** tonytan4ever has joined #openstack-infra21:54
pabelangerianw: oh, interesteing. We should see how ntpdate server is installed then.  Because there is still the issue that ntpd depends on ntpdate21:55
*** PollyZ has quit IRC21:55
pabelanger host name not found: would seem to be a DNS issue21:56
mordredmgagne: sigh. have I mentioned these APIs drive me CRAZY ?21:56
*** ajmiller has joined #openstack-infra21:57
*** tonytan_brb has quit IRC21:57
mgagneand now glance-api made virtual_size read-only while the Cinder SolidFire driver image caching relies on this same property in kilo.21:57
mordredI cannot believe that we have a strongly-typed key-value metadata store21:57
mordredwell, this is the reason that having operator created arbitrary metadata was a bad design choice - BUT - that's not going to help you deal with this21:58
*** sdake has quit IRC21:58
jgriffithmgagne: sigh21:58
mgagnejgriffith: you sir have the solidfire keyword on notification21:58
jgriffithmgagne: that's not the only thing they broke, I updated those in M I believe, but that likely doesn't help you21:58
jgriffithmgagne: yes, sorry... I'll pipe down unless asked :)21:58
mgagnemordred: since I consider glance a glorified artifact store for images, I should be free to set whatever metadata I want =)21:59
mordredmgagne: well ... your users should in the images they upload for sure. but arbitrary operator set metadata that carries system meaning is _hell_ for end users21:59
*** ajmiller_ has joined #openstack-infra22:00
*** Teresita-Warrior has quit IRC22:00
mordredmgagne: I'm trying to come up with a patch that would give you a way to do what you need without breaking exisitng behavior or api, fwiw22:00
mgagnemordred: true. that's also sad that there is no way to communicate to the end user the actual version of the image or the distro or whatever without concatenating the whole thing in the name.22:00
mordred(just so you don't think I'm just complaining)22:01
ianwpabelanger: i think if we just let packagers sort it out ... really the thing is the ntp puppet should make sure ntpdate service is enabled so the clock is in sync before ntpd starts22:01
mordredmgagne: +1000000022:01
mgagnemordred: I will find a way22:01
mordredmgagne: I mean, I have a TERRIBLE idea22:01
mordredwhich will work22:01
mgagnemordred: which might just be reverting to glance v1 so I can sleep tonight22:01
pabelangerianw: yup, agree22:01
*** stew925 has joined #openstack-infra22:01
jgriffithmgagne: if you're interested in not reverting but something else... I just moved that check and use the response form qemu_img_info now22:02
*** HeOS has quit IRC22:02
jgriffithmgagne: I'm trying to remove any and all interactions with Glance that I can at this point22:03
mgagnejgriffith: yea, I read about it a couple of months ago but didn't bother repatching yet, wasn't broken (yet) to bother fixing22:03
*** ajmiller has quit IRC22:03
jgriffithmgagne: fair enough, that was a very angry evening for me :)  Between this and f'ing "visibility"22:03
*** esberglu has quit IRC22:04
jgriffithanyway... sorry for interrupting22:04
mgagnejgriffith: np, creating an item in our backlog to take a look at the latest version of the image caching code22:04
*** siva_krish has quit IRC22:05
*** abregman has quit IRC22:06
openstackgerritMonty Taylor proposed openstack-infra/shade: Add ability to provide direct glance metadata
mordredmgagne: something like that ^^22:07
mordredthe previous patch also doesn't properly do what we expect on task-upload based systems, so it'll need another revision22:08
mordredI also think this is hella ugly22:08
mgagnemordred: how about straight_pipe=True and only one argument is used: raw_request_body :D22:08
mgagnemordred: I will work on updating our system, I feel like this whole house of cards is falling apart22:08
mordredmgagne: I may want to move some things aroud and put a create_image method on OperatorCloud that has this option22:08
*** mdrabe has quit IRC22:08
*** amitgandhinz has quit IRC22:09
*** PollyZ has joined #openstack-infra22:09
*** mdrabe has joined #openstack-infra22:09
mgagneI just wanted to upload an image and now I'm here, refactoring the whole thing :D22:09
mordredmgagne: welcome to my life22:09
mordredmgagne: I just wanted to upload an image and I had to write shade22:09
mordredmgagne: anyway - thanks for pointing out the two issues - I understand them and will keep poking at them to get a solution that works better, even if that won't completely help you before you go to sleep22:10
mgagnemordred: =) thanks for the help ;)22:10
*** sdake has joined #openstack-infra22:13
*** vvalderrv1 has left #openstack-infra22:13
*** HeOS has joined #openstack-infra22:14
*** jcoufal has joined #openstack-infra22:14
*** jcoufal_ has quit IRC22:14
*** mdrabe has quit IRC22:17
*** Goneri has quit IRC22:18
*** kushal has joined #openstack-infra22:19
*** xyang1 has quit IRC22:21
*** kgiusti has left #openstack-infra22:22
ianwpabelanger: urgh, i guess for it to go in puppetlabs-ntp we'd have to do it properly and template out all the step-files (for people behind firewalls) and add a bunch of parameters, and figure it out across centos/debian/freebsd22:22
*** jcoufal has quit IRC22:24
*** esberglu has joined #openstack-infra22:25
*** jcoufal has joined #openstack-infra22:25
*** mmandel has quit IRC22:26
*** cschwede has quit IRC22:26
*** kushal has quit IRC22:28
notmynamehas there ever been requests to have different filesystems (ie not ext4) on the boxes that run tests?22:30
*** camunoz has joined #openstack-infra22:30
notmyname(you might guess what my next question will be)22:31
*** armax has joined #openstack-infra22:31
fungiianw: the difference is that "modern" ntpd performs a rapid convergence at start, so the wait isn't long22:32
*** kzaitsev_mb has quit IRC22:32
fungiianw: what needs to be fixed is that the systemd unit file for ntpd doesn't declare a dependency on networking22:32
fungiinstead getting it indirectly through ntpdate's dependencies, which are ignored if ntpdate is not used22:33
fungiianw: relying on ntpdate is a poor long-term strategy as it's heavily deprecated and no longer recommended since it's not receiving upstream maintenance for some time now22:35
*** jcoufal has quit IRC22:36
*** jcoufal has joined #openstack-infra22:36
*** flepied has joined #openstack-infra22:39
ianwfungi: i dunno, see what the maintainer says in the bug...22:39
*** signed8bit is now known as signed8bit_Zzz22:40
*** rfolco_ has joined #openstack-infra22:41
*** yamahata has quit IRC22:42
ianwit's not so much the network, but what ntp does when it's more than 0.128s out (goes into it's synchronisation loop for 10 minutes or so)22:42
ianwthat's why if you "service ntpd restart" it just comes good straight away22:42
*** john-davidge has quit IRC22:42
fungiianw: the failure noted is that ntpd failed to resolve the names of its ntp servers. that's the network yes?22:42
fungiwhen restarting it, dns resolution happened fine22:43
fungithat tells me it was probably initially started before networking was up22:43
ianwfungi: no, it picks up the servers just fine when the network comes up, it knows how to do that22:43
ianwsee bug 1361382 in ntp "ntp-wait hangs after boot for a long time, unless ntpd is restarted" [Unspecified,Closed: notabug] - Assigned to mlichvar22:44
fungiianw: does it still go into rapid time convergence mode if it reresolves them after start?22:44
ianw"When the initial offset is larger than 0.128s, ntpd will step the clock and then it will wait for at least 900 seconds (in default configuration) before it reports it's in the synchronized state.22:44
openstackgerritChristopher Aedo proposed openstack-infra/infra-specs: Add spec for hosted IRC client
ianwso if the clock is out by more than 0.128s when you start, well, you just gotta wait until ntpd thinks it's synchronised22:44
*** esberglu has quit IRC22:45
ianwi think this is the "stepout" value, that maybe you can tinker22:45
*** r-daneel has joined #openstack-infra22:47
fungiianw: or are we not using iburst?22:47
*** esberglu has joined #openstack-infra22:49
*** yamamoto has joined #openstack-infra22:50
*** thorst has joined #openstack-infra22:51
openstackgerritMerged openstack-infra/puppet-ssh: Add @trusted_ssh_type for user to override
openstackgerritMerged openstack-infra/system-config: Use IPv6 address for puppetmaster.o.o
*** yamamoto has quit IRC22:54
*** yamahata has joined #openstack-infra22:55
*** tphummel has quit IRC22:56
*** rbrndt has quit IRC22:57
fungiianw: reading the bug, i agree chronyd on rh derivative platforms seems fine if they're moving away from as their default time sync implementation22:57
*** jcoufal has quit IRC22:57
*** jcoufal has joined #openstack-infra22:58
*** rbuzatu has joined #openstack-infra22:58
*** tphummel has joined #openstack-infra22:59
*** thorst_ has joined #openstack-infra23:00
*** nwkarste_ has quit IRC23:00
*** sdake has quit IRC23:01
*** thorst has quit IRC23:02
*** fbo has quit IRC23:03
*** rbuzatu has quit IRC23:03
*** asettle has joined #openstack-infra23:03
*** sdake has joined #openstack-infra23:03
fungimordred: (came through just now) is a prime example for your github description thread if you need one23:04
*** nwkarste_ has joined #openstack-infra23:04
mtreinishpabelanger, rcarrillocruz: ^^^ the 2 ssh fixes landed23:04
pabelangermtreinish: ya, was just about to check puppetmater23:05
*** tphummel has quit IRC23:05
pabelangerfungi: ianw: no iburst:
pabelangerthat is what centos-7 has23:06
*** tonytan4ever has quit IRC23:07
*** jcoufal_ has joined #openstack-infra23:07
fungipabelanger: i'm coming around to ianw's suggestion that long term we can switch away from using if that's not the distro's preference anyway23:07
*** jcoufal has quit IRC23:08
*** nwkarste_ has quit IRC23:08
*** asettle has quit IRC23:09
fungipabelanger: as long as we have some supported mechanism on each distro for making sure time is in sync with a trustworthy global time service prior to starting jobs, that's all that really matters to me23:09
*** tphummel has joined #openstack-infra23:09
fungiwe have plenty of distro-specific special casing already, so one more won't matter23:09
*** rfolco_ has quit IRC23:09
*** amitgandhinz has joined #openstack-infra23:10
*** r0ckS0LiD has quit IRC23:10
*** mriedem has joined #openstack-infra23:13
*** thorst_ has quit IRC23:13
*** thorst has joined #openstack-infra23:14
fungiuh oh...
*** amitgandhinz has quit IRC23:14
*** sdague has quit IRC23:14
fungisetuptools 25.1.2 (out today) failing to install sphinx with "UnicodeDecodeError: 'ascii' codec can't decode byte 0xcc in position 8: ordinal not in range(128)23:14
*** gouthamr has quit IRC23:15
*** jcoufal_ has quit IRC23:15
fungifrom the traceback quoted in the issue, looks like it might choke on non-ascii filenames in packages23:15
*** achanda has joined #openstack-infra23:16
*** nwkarsten has joined #openstack-infra23:16
*** esberglu has quit IRC23:17
*** jcoufal has joined #openstack-infra23:18
*** nwkarsten has quit IRC23:21
*** thorst has quit IRC23:22
*** PollyZ has quit IRC23:22
*** ianychoi has quit IRC23:23
*** jesusaur has quit IRC23:24
*** esberglu has joined #openstack-infra23:25
craigeping clarkb23:26
*** jcoufal has quit IRC23:28
notmynamefungi: clarkb: mordred: is there any support in -infra for running jobs with a different filesystem available? right now, the worker nodes have ext4, but I'd like to use xfs23:28
greghaynesnotmyname: in theory an image could be made which uses that fs23:30
funginotmyname: it would require completely separate images and node types if you wanted the rootfs reformatted with a different filesystem, so that sets the bar pretty high from a resource perspective (adding another image type for any reason other than supporting a new distro has been avoided so far)23:30
*** jcoufal has joined #openstack-infra23:30
*** weshay_bbiab is now known as weshay23:30
funginotmyname: on the other hand, you can format a filesystem on a loop device however you want and use that23:30
notmynameyeah, that's what I was thinking23:30
pabelangerfungi: mtreinish: blarg, apparently we still need UseDNS yes in sshd_config on ubuntu-xenial, even after we've changed to match address23:31
pabelangerfungi: mtreinish: I am not sure why that is23:31
fungipabelanger: is it forcing reverse dns lookups maybe?23:31
*** sdake has quit IRC23:31
fungipabelanger: could be there's another option we need to disable along with it?23:32
notmynamefungi: is that something we could set up and configure in code in our repo or does it need to be done in the infra repo?23:32
notmynamefungi: oh, this might be a devstack thing. (I'm not sure)23:32
pabelangerfungi: just found:
pabelangertesting that right now23:32
funginotmyname: it depends entirely on how the job in question is being run, yes23:32
*** achanda has quit IRC23:32
pabelangerwe do add something into authorized_keys23:32
*** nwkarsten has joined #openstack-infra23:33
notmynamefungi: basically, I need the mount points that are in the swift ring to refer to xfs partitions23:33
funginotmyname: if you want your python unit tests run on a special filesystem, that would need configuration in a custom job to do all that. on the other hand for a devstack-based job you might be able to do it with an in-repo devstack plug-in or devstack-gate hook23:33
funginotmyname: in which case you could just specify those in the job definition and then keep the implementation logic in the swift repo23:35
*** ianychoi has joined #openstack-infra23:35
notmynamefungi: where should I look first? I want my py* unit tests and in-repo functional tests to be run with xfs23:35
pabelangerfungi: looks like the from="" in authorized_keys is the issue. Updating it to use the IP address23:36
funginotmyname: different places for each of those. i'm winding down for the evening and not in a position to go into much detail but perhaps our team members in apac timezones (ianw, jhesketh) can chime in with suggestions23:37
notmynamefungi: oh I understand. I'm about to take off too23:37
* jhesketh waves23:37
ianwnotmyname: loopback?23:38
notmynameianw: yeah, likely. do you recall what's used in devstack today for swift? is it loopback devices that are created and then added to the ring?23:39
notmynameor, where would I go look to find the answer to that?23:39
fungiianw: yeah, the question is how to format a file xfs and loop-mount it, then tell (e.g. a custom swift unit test job or devstack) to use that for the swift ring23:39
ianwnotmyname: yeah, i'm pretty sure swift is loopback23:39
ianwnotmyname: special infra stuff though probably isn't the way to go though, because you want people to run it locally too23:40
notmynameif it's already a loopback device(s) for data drives in the swift test, then it should be simple to change it from ext4 to xfs23:40
*** edmondsw has quit IRC23:40
*** sdake has joined #openstack-infra23:40
ianwnotmyname: maybe it is already?
notmynameianw: right. and I want it in the swift repo anyway (preferably), basically because it's easier to get changes landed there that affect swift rather than bugging another team (ie you)23:41
notmynameianw: oh, very interesting!23:41
ianwISTR some changes around this, a while ago... memory hazy :)23:41
fungii'm not immediately familiar with how/what swift's ring is configured to use locally in unit tests. fs interactions are something i'd expect to find mocked in normal unit testing, but here in openstack we have a very loose definition of "unit"23:41
*** flepied has quit IRC23:42
notmynamefungi: the cause here is that ext4 has a pretty bad xattr story compared to xfs, and we'd like to start testing some functionality related to xattrs. currently we're mocking it out, but that leaves a gap in the testing (but also is why we don't see problems today)23:42
openstackgerritPaul Belanger proposed openstack-infra/system-config: Also use IP address in from fields for authorized_keys
pabelangerianw: fungi: mtreinish: ^ should also have been updated to our IPs for sshd23:43
notmynameianw: do the standard python jobs run in a devstack environment or is it configured some other way?23:44
funginotmyname: would definitely be a custom job definition to do that in unit tests since you would need root access to set up a loop mounted fs, but implementation in the job definition would depend somewhat on how you expected to leverage it in your unit tests23:44
notmynamefungi: ianw: for context, this is the specific patch we want and that is currently failing (seemingly from ext4's failing wrt xattrs)
notmynameah, it's /tmp that is ext423:46
ianwnotmyname: you mean like what you get when you type "tox -epy27"?23:46
funginotmyname: our "normal" python unit tests are just run on a node booted from a very minimal image, some common packages installed, in some cases a very minimal database configuration applied for mysql/postgres, and then root permission is dropped and tox runs the tests within a virtualenv under an unprivileged user23:46
notmynamefungi: got it23:46
*** xarses has quit IRC23:47
notmynameso yeah, it's not the devstack setup. this is a unit test (py27 job) that is writing to some temp files in /tmp and failing because that's ext423:47
*** gildub has joined #openstack-infra23:47
ianwnotmyname: yeah, i think there's no reason that you couldn't setup a loopback and format etc from your unit tests.  tricky to make sure you clean it up properly.  sounds like a good job to give to the intern :)23:47
notmynameianw: lol23:48
funginotmyname: we could probably add a custom job macro to mount a loop device at /tmp (or specify $TEMPDIR as a different place where we've done that) and then proceed as normal with the job. so the deviation from "normalcy" would be pretty minimal23:48
*** zz_dimtruck is now known as dimtruck23:48
notmynamefungi: oh, that's an interesting idea23:48
fungiianw: loop-mounting a filesystem without root access is tricky (impossible? at least without nonstandard kernel capabilities applied)23:48
fungimy information may be outdated though23:49
notmynamefungi: our test isn't run as root, right? so that would have to be created as part of the image?23:49
notmynamefungi: and then the swift job definition would set the envvar TEMPDIR and we'd be good?23:49
notmyname"it's just typing"23:50
ianwnotmyname: yeah, good point that we drop sudo by default before unit tests23:50
funginotmyname: we have stages ("builders") in the job that can run before the stage where root permission is dropped23:50
greghaynesyou could also add a little setuid script that does the loopback device creation23:50
*** yamamoto has joined #openstack-infra23:51
fungiso it would be a custom job, but would probably just need one fairly trivial builder added beyond what the usual unit test job template does23:51
notmynamefungi: so we'd write the builder and have it in the swift repo?23:51
fungigreghaynes: yep, though that setuid script either needs to be baked into the image or installed by a builder before the revoke-sudo builder runs anyway23:51
notmyname(sorry, builders are new to me)23:51
greghaynesfungi: Yep23:52
funginotmyname: builder macros are what you see in
greghaynesah, builder before dropping sudo sounds like win :)23:52
ianwgreghaynes: i guess the problem is that what if you mount an image that has an setuid version of bash on it?23:52
ianwi guess we also don't care23:53
notmynamefungi: oh ok, so a new builder in that file and then add it to the job definition? jobs are composed of the builders, right?23:53
notmynamefungi: is there a problem with the fact that this is a "standard" py27 job?23:53
funginotmyname: if you look at the gate-{name}-python27 job-template in you'll see it has a list of builders it runs in sequence23:53
greghaynesianw: Yea, its wierd here what we care about, but I think the builder is a better idea anyhow23:53
*** thorst has joined #openstack-infra23:54
greghaynessince I think lack of sudo is more of a 'dont shoot yourself in the foot' rather than a 'we are protecting ourselves from privilege escalation'23:54
funginotmyname: you'd need a separate job definition rather than using the jgate-{name}-python27 job-template, but i think it would only differ by the addition of that one extra builder23:54
notmynamefungi: ah ok. then would there be a problem with swift not running the "normal" py27 job? it would fail23:55
fungigreghaynes: ianw: yes, the revoke-sudo builder is there to ensure that you can locally run, e.g., `tox -e py27` without needing root privs23:55
funginotmyname: i think if swift's unit tests can no longer run _without_ a specific local filesystem available, that may run afoul of the tc's "consistent testing interface"23:56
*** yamamoto has quit IRC23:56
fungibut to have a second job that runs it with xfs seems fine23:56
notmynamefungi: I suppose that depends on what the definition of "consistent testing interface" is.23:57
fungiand you can always disable tests in the normal job run if they need a specific filesystem23:57
*** jcoufal has quit IRC23:57
*** dingyichen has joined #openstack-infra23:57
*** jcoufal has joined #openstack-infra23:58
notmynameok, to summarize (1) make a new builder that creates an xfs loopback (2) make a new job-template copied from the py27 one that addes the new builder (3) add the new job in the swift gate23:58
funginotmyname: the definition of "consistent testing interface" is
funginotmyname: you got it23:58
notmynamefungi: heh "the tests that don't test everything work just fine!" ;-)23:58
funginotmyname: well, if you had ext4-specific tests and xfs-specific tests you'd have to disable some one way or the other unless you mocked them23:59
mtreinishpabelanger: is localhost not an issue? should you change that to jsut to be safe?23:59
funginot all tests are guaranteed to be relevant to every possible testing platform23:59

Generated by 2.14.0 by Marius Gedminas - find it at!