Wednesday, 2018-05-09

*** rlandy has quit IRC00:07
*** Goneri has joined #oooq00:29
hubbotFAILING CHECK JOBS on stable/ocata: tripleo-ci-centos-7-undercloud-upgrades @ https://review.openstack.org/56429100:48
*** Goneri has quit IRC01:09
*** username_ has joined #oooq01:20
*** username_ is now known as username__01:23
*** saneax is now known as saneax-_-|AFK02:02
*** username__ has quit IRC02:35
*** myoung|ruck|bbl is now known as myoung|ruck02:38
hubbotFAILING CHECK JOBS on stable/ocata: tripleo-ci-centos-7-undercloud-upgrades, tripleo-ci-centos-7-undercloud-oooq @ https://review.openstack.org/56429102:48
*** udesale has joined #oooq03:36
*** links has joined #oooq03:43
*** udesale has quit IRC04:00
*** udesale has joined #oooq04:01
*** ykarel|away has joined #oooq04:18
*** udesale_ has joined #oooq04:36
*** udesale has quit IRC04:39
*** ratailor has joined #oooq04:42
*** marios has joined #oooq04:46
hubbotFAILING CHECK JOBS on stable/ocata: tripleo-ci-centos-7-undercloud-upgrades, tripleo-ci-centos-7-undercloud-oooq @ https://review.openstack.org/56429104:48
*** ykarel|away is now known as ykarel05:07
ykarelfailed: [localhost] (item=[u'current-tripleo-rdo', u'swift-proxy-server']) => {"changed": false, "item": ["current-tripleo-rdo", "swift-proxy-server"], "msg": "Error pushing image docker.io/tripleomaster/centos-binary-swift-proxy-server: Error: Status 403 trying to push repository tripleomaster/centos-binary-swift-proxy-server: \"Push failed due to a network error. Please try again. If the problem persists, it may be due to a slow connection.\""}05:07
*** udesale_ has quit IRC05:22
bandinia nice and simple one, folks. great way to start the morning ;) https://review.openstack.org/#/c/490590/ (ignore the old august jenkins failures)05:35
*** jfrancoa has joined #oooq05:36
*** pgadiya has joined #oooq05:37
*** pgadiya has quit IRC05:37
*** jtomasek has joined #oooq05:41
*** jtomasek has quit IRC05:41
*** ratailor has quit IRC05:41
*** jtomasek has joined #oooq05:42
ykarelbandini, strange05:42
ykarelbandini, 3 months back only commit message merged for this fix: https://review.openstack.org/#/c/509084/05:43
ykarel4th PS had the change command --> shell05:43
ykarelnot sure how it happened05:44
bandiniykarel: oh wow. that is so weird05:44
bandinioh well05:45
*** quiquell|off has quit IRC05:45
ykarelwhats going on i can see your patch also is merged 4 months ago: https://review.openstack.org/#/c/534232/105:48
ykarelohh this patch is too old05:48
ykarelcan be ignored https://review.openstack.org/#/c/490590/05:48
*** quiquell has joined #oooq05:49
*** udesale_ has joined #oooq05:51
ykarelbandini, commented05:53
*** udesale_ has quit IRC05:53
*** udesale has joined #oooq05:53
*** kopecmartin has joined #oooq05:54
bandinioh I see, ack05:55
*** ratailor has joined #oooq05:55
*** ratailor has quit IRC06:05
*** ratailor_ has joined #oooq06:05
*** saneax-_-|AFK is now known as saneax06:10
*** holser__ has joined #oooq06:14
*** quiquell has quit IRC06:26
*** quiquell has joined #oooq06:27
*** saneax is now known as saneax-_-|AFK06:30
*** saneax-_-|AFK is now known as saneax06:33
*** jaosorior has joined #oooq06:34
*** ccamacho has quit IRC06:38
*** brault has joined #oooq06:38
*** ccamacho has joined #oooq06:39
*** holser___ has joined #oooq06:43
*** holser__ has quit IRC06:44
hubbotFAILING CHECK JOBS on stable/ocata: tripleo-ci-centos-7-undercloud-upgrades, tripleo-ci-centos-7-undercloud-oooq @ https://review.openstack.org/56429106:48
*** quiquell is now known as quiquell|ack06:58
*** quiquell|ack is now known as quiquell|afk06:58
*** ratailor__ has joined #oooq07:08
*** ratailor_ has quit IRC07:10
*** ratailor_ has joined #oooq07:15
*** florianf has joined #oooq07:16
*** ratailor__ has quit IRC07:17
*** tesseract has joined #oooq07:22
*** amoralej|off is now known as amoralej07:36
sshnaidm|roverrasca, hi07:37
*** tosky has joined #oooq07:37
sshnaidm|roverrasca, is it still relevant? https://review.openstack.org/#/c/553465/07:37
*** bogdando has joined #oooq07:46
*** dtantsur|afk is now known as dtantsur07:53
*** ykarel is now known as ykarel|lunch07:58
sshnaidm|roverarxcruz, what's about https://review.openstack.org/#/c/565716/ ? why is it DNM?08:02
arxcruzsshnaidm|rover: checking08:03
arxcruzsshnaidm|rover: i was testing first if the ironic tests would pass08:04
arxcruzthen i need to figure out which tests to run08:04
arxcruzthat's why is still a dnm08:04
sshnaidm|roverarxcruz, ok, better to have DNM in first line, so that people will see it without opening a patch08:05
arxcruzsshnaidm|rover: gotcha08:05
*** apetrich has quit IRC08:07
quiquell|afksshnaidm|rover: Reproducers + libvirt is usable ?08:07
*** quiquell|afk is now known as quiquell08:08
sshnaidm|roverquiquell|afk, for some yes, not for me though08:08
sshnaidm|roverquiquell, if you will have dns problem like me, you can run with https://review.openstack.org/#/c/565839/08:10
quiquellsshnaidm|rover: Using it reduce deploy time ?08:16
sshnaidm|roverquiquell, using what?08:16
quiquellreproducer + Ãlibvirt08:16
sshnaidm|roverquiquell, well, depends, rdo cloud I think would be faster, but it's sometimes not available08:18
quiquellAhhh, so reproducer + libvirt can be used as a fallback08:18
quiquellokok08:18
quiquellThanks08:18
sshnaidm|roverquiquell, although when I use docker and http proxy with libvirt, I think it could be faster than rdo cloud08:18
quiquellWhat http proxy ?08:19
sshnaidm|roverquiquell, just squid, it's not supported yet, but I use it with my old POC: https://review.openstack.org/#/c/543429/08:20
sshnaidm|roverquiquell, I hope we could use it with current libvirt reproducer too08:20
sshnaidm|roversometime08:22
quiquellsshnaidm|rover: Would be great08:22
*** skramaja has joined #oooq08:26
*** udesale_ has joined #oooq08:28
*** udesale has quit IRC08:28
*** lucas-bbl is now known as lucasagomes08:31
rascasshnaidm|rover, yes, I will take care of this very soon08:32
rascasshnaidm|rover, this is something we really need08:32
rascafrom a baremetal perspective08:32
*** jaosorior has quit IRC08:36
*** jaosorior has joined #oooq08:41
hubbotFAILING CHECK JOBS on stable/ocata: tripleo-ci-centos-7-undercloud-upgrades, tripleo-ci-centos-7-undercloud-oooq @ https://review.openstack.org/56429108:48
*** udesale_ has quit IRC08:59
*** udesale_ has joined #oooq08:59
*** udesale_ has quit IRC09:00
*** udesale has joined #oooq09:00
*** panda|off is now known as panda09:01
*** gkadam__ has joined #oooq09:06
quiquellsshnaidm|rover: To check toci_type in reproducers at runtime09:11
quiquellhttps://review.openstack.org/#/c/56716209:11
quiquellWith this you can change from multinode to ovb without changing the reproducer09:11
*** ykarel|lunch is now known as ykarel09:28
*** zoli is now known as zoli|lunch09:31
*** apetrich has joined #oooq09:37
*** jaganathan has quit IRC09:39
*** ratailor__ has joined #oooq09:44
*** ratailor_ has quit IRC09:47
quiquellpanda: Do we have a card to poke at undercloud upgrade ?10:11
sshnaidm|roverquiquell, you need to configure your editor to strip spaces automatically from end on line10:17
quiquellsshnaidm|rover: Did I do it again ? damn...10:18
quiquellI have to...10:18
pandaquiquell: there are currently no cards10:31
quiquellpanda: I am testing it changing the overcloud-upgrade role, maybe we need a card to state the status of it10:33
pandaquiquell: what are you testing exactly ?10:34
quiquellUsing trowns patch I pass the correct release file to undercloud-upgrade10:34
quiquellAnd changing the role so it doesn't read from UPGRADE_RELEASE10:34
quiquellinstead use {{ release }}Ã10:34
pandaquiquell: can we chat a sec ?10:36
quiquellhttps://bluejeans.com/789106523210:37
*** saneax is now known as saneax-_-|AFK10:38
*** zoli|lunch is now known as zoli10:46
hubbotFAILING CHECK JOBS on stable/ocata: tripleo-ci-centos-7-undercloud-upgrades, tripleo-ci-centos-7-undercloud-oooq @ https://review.openstack.org/56429110:48
*** ykarel_ has joined #oooq10:59
*** gkadam_ has joined #oooq10:59
*** gkadam__ has quit IRC10:59
*** pliu has quit IRC10:59
*** ykarel has quit IRC11:00
*** tcw has quit IRC11:01
*** pliu has joined #oooq11:02
*** lucasagomes is now known as lucas-hungry11:06
*** tcw has joined #oooq11:09
*** atoth has joined #oooq11:25
*** ratailor__ has quit IRC11:47
*** apetrich has quit IRC11:53
*** quiquell is now known as quiquell|lunch11:54
*** apetrich has joined #oooq11:55
weshaymyoung|ruck, updating the status a bit11:59
weshaythis meeting is about osp-13 status, osp-14 planning12:00
weshaymyoung|ruck, keep it tight on queens and on promotions :)12:01
weshaypanda, howdy12:07
*** lucas-hungry is now known as lucasagomes12:07
weshaysshnaidm|rover, what change did you have to make to get the patch expiration to work?12:08
sshnaidm|roverweshay, what do you mean?12:11
weshaysshnaidm|rover, so I see patches are being abandoned was that done via a policy, script, how?12:11
sshnaidm|roverweshay, using my hands and eyes :)12:12
weshayoh no12:12
weshayk12:12
sshnaidm|roverweshay, better to look at every case separately12:13
sshnaidm|roverweshay, please see a mail12:13
weshayah thanks12:13
weshaysshnaidm|rover, thanks for cleaning it up!12:14
weshaysshnaidm++12:14
weshaywould like to see it done automatically12:15
weshaythat was quite a bit of work12:15
*** tcw has quit IRC12:15
*** tcw has joined #oooq12:17
*** tcw has quit IRC12:18
sshnaidm|roverweshay, yeah, but it's for 2 years12:18
*** tcw has joined #oooq12:18
sshnaidm|roverweshay, next time it will be much easier12:18
weshaypanda, you around?12:20
weshaysshnaidm|rover, panda need a review for the ui guys https://review.openstack.org/#/c/518043/12:21
sshnaidm|roverweshay, it's not tested in ci, right?12:23
weshaysshnaidm|rover, https://ci.centos.org/artifacts/rdo/jenkins-tripleo-quickstart-extras-gate-newton-delorean-full-minimal-5726/console.txt.gz12:23
weshay2018-05-08 22:17:37.531 |  TASK [validate-ui : Execute the simple tripleo-ui validation] ******************12:24
weshayis the only test we have for htat12:24
weshaythat12:24
sshnaidm|roverweshay, and it's for newton12:24
weshayya :(12:24
weshaysshnaidm|rover, I wonder if it's retrying a more recent release12:25
weshayin ci.centos12:25
weshaypanda, where you at12:26
sshnaidm|roverweshay, I have problem with expressions like "release not in ['master', ....]"12:26
sshnaidm|roverweshay, every time when master moves it's broken12:26
sshnaidm|roverweshay, better to use from older releases, like "release in [newton, ocata]   release not in [newton, ocata]"12:27
weshaysshnaidm|rover, ya.. I'd be more comfortable doing that if there was more support for leaving if statements on branches that are not executed upstream12:28
pandaweshay: hey, sorry, lunch12:28
weshaypanda, ok.. I need 2 min.. and then we should sync12:28
sshnaidm|roverweshay, we have support from newton afaik12:29
sshnaidm|roverso it could be as oldest branch12:29
sshnaidm|roverweshay, maybe we need an ansible lint rule for that..12:29
*** rlandy has joined #oooq12:32
weshaysshnaidm|rover, ya.. but we have a history of removing the older branches from if statements12:33
weshaypanda, k.. ready12:33
weshaypanda, https://bluejeans.com/u/whayutin/12:33
*** trown|outtypewww is now known as trown12:34
weshaysshnaidm|rover, so I can update the patch to flip it.. we just need to consider our policy on removing old branches from if statements12:34
sshnaidm|roverweshay, I don't understand - what is the issue with removing old branches?12:35
sshnaidm|roverweshay, if we remove "newton" everywhere it still works12:35
weshaysshnaidm|rover, except when I run on osp 1012:35
sshnaidm|roverweshay, not sure I understand12:36
weshayok.. so if I say.. when: release in [ kilo, liberty, newton, ocata ] vs.. when release not in [ pike, queens, master ] the logic of the when statement is likely to change if we rip out old releases from if statements12:37
weshayso I've been leaning on not doing the former12:38
weshayI can change the when's so there is less maint for us as you suggested.. just looking for that to be considered when we from time to time remove the older branches from when statements12:39
weshaythat is all :)12:39
sshnaidm|roverweshay, I don't understand why logic of when statement should change? If no old releases anymore, you just remove "when"12:40
sshnaidm|roverweshay, removing old branches support and using releases in "when" are different things, btw12:41
sshnaidm|roverweshay, if ocata and newton are deprecated for oooq - we just remove this "when"12:42
weshaysshnaidm|rover, so do we only consider what is deprecated upstream, or do we take into account osp?12:42
sshnaidm|roverweshay, I don't understand how it's related in this case12:43
weshaywell.. I'm speaking to how I generally craft the when statements12:43
weshayto be defensive with regards to older releases12:44
sshnaidm|roverweshay, what's problem with osp when you use older branches in "when" statements?12:44
weshaysshnaidm|rover, we tend to remove the older branches from when statements from time to time12:45
sshnaidm|roverweshay, yeah12:46
hubbotFAILING CHECK JOBS on stable/ocata: tripleo-ci-centos-7-undercloud-upgrades, tripleo-ci-centos-7-undercloud-oooq @ https://review.openstack.org/56429112:48
*** skramaja_ has joined #oooq12:49
*** skramaja has quit IRC12:50
weshaysshnaidm|rover, did you see https://bugs.launchpad.net/tripleo/+bug/177002112:50
openstackLaunchpad bug 1770021 in tripleo "hubbot: TQE check jobs have not been kicked by hubbot since 4/25" [Critical,Triaged]12:50
sshnaidm|roverweshay, yeah, didn't find any problem, waiting for myoung|ruck to explain12:51
weshayk12:51
quiquell|lunchsshnaidm|rover, myoung|ruck, weshay: I have put together a dashboard for ruck/rover12:51
quiquell|lunchhttp://38.145.34.131:7272/12:51
quiquell|lunchFrom stuff I needed in my previous ruck/rovering12:52
*** quiquell|lunch is now known as quiquell12:52
weshaysshnaidm|rover, so look at https://review.openstack.org/#/c/56045212:52
sshnaidm|roverquiquell++12:52
hubbotsshnaidm|rover: quiquell's karma is now 112:52
trownnice12:52
sshnaidm|roverquiquell, you read my thoughts, just started to do it with grafana :D12:53
quiquellsshnaidm|rover: Let me know if it's of any use or we can change something12:53
*** skramaja has joined #oooq12:53
quiquellIs still some WIP12:53
*** skramaja_ has quit IRC12:54
sshnaidm|roverquiquell, what do you use for dashboard?12:54
quiquelltipboard12:54
quiquellDidn't customize it too much12:54
quiquellBut you can12:54
quiquellJust want to show the info12:54
sshnaidm|roverquiquell, wanna switch to telegraph+grafana?12:54
quiquellLet me check, didn't want graphs I want green/red boxes :-)12:55
quiquellAnd links pointing to logs12:55
quiquellThat's pending tough12:55
sshnaidm|roverquiquell, grafana has boxes too12:56
quiquellWill check12:56
sshnaidm|roverquiquell, with alerts, etc12:56
quiquellThen we decide12:56
quiquellI choose the simplest to just show it to you12:56
sshnaidm|roverquiquell, looks great, I think we can do it more official12:57
quiquellLet's play a little with it first12:57
quiquellI have to move it to config-ci and do a prper ansible role to setup things12:57
sshnaidm|roverquiquell, does it take data from your scripts?12:57
quiquellI push the data from the scripts to tipbaord12:58
quiquellevery 5 minutes12:58
sshnaidm|roverquiquell, great12:58
quiquellI can update each box in parallel12:58
quiquellNot done yet12:58
*** amoralej is now known as amoralej|lunch12:58
sshnaidm|roverquiquell, so take a look at grafana, because we use it in Ci already, need to create a dashboard, it would be nice to reuse it instead of new tool12:59
sshnaidm|roverquiquell, all your scripts will be integrated without a problem12:59
quiquellsshnaidm|rover: They just render data into pandas, it's just a matter of using it13:00
myoung|ruckweshay: I saw your changes, makes sense.  I had put things to Yellow overall due to the CIX issues, but from a focused osp13 status, agree with your edits to team status.  It certainly made tripleo-ci go quickly on the call :)13:01
sshnaidm|roverquiquell, I was thinking about Telegraf which sends to influxdb adn then rendered by grafana: https://docs.influxdata.com/telegraf/v1.6/concepts/data_formats_input/13:01
*** Goneri has joined #oooq13:02
weshaymyoung|ruck, the osp board determines the status13:03
weshaycolor13:03
quiquellmyoung|ruck: Would be possible to do team meetings at 1 PM UTC instead of 2 PM UTC in general ?13:04
myoung|ruckquiquell: the only problem with 1pm UTC is that for folks in mountain time (wes) that's a 7am meeting13:07
*** rasca has quit IRC13:08
*** rasca has joined #oooq13:08
myoung|ruckweshay: re: boards and color, that's why I had indicated yellow, we have 2 rdo2 tiles that are yellow...including master.  when I wrote the status it was broadly, and I was going to caveat and say "yellow due to rdo2 and CIX, green for beta"13:09
quiquellOk, I just mean one hour earlier13:09
weshaymyoung|ruck, which queens card on the board is yellow?13:09
myoung|ruck...hence..."green for beta, yellow for reasons above"  When I wrote the status earlier I hadn't realized the meeting was narrowly focused on queens/13 beta13:10
myoung|ruckI had assumed it was "beta/13 status, 14 planning" - so master mattered.  Not trying to be pedantic here...I noticed the edits from my phone in the car.  Thanks for streamlining the status :)13:11
*** ykarel_ is now known as ykarel13:13
myoung|ruckquiquell: re:  UTC 1400 meetings not working, please bring it up in retrospective, or we can discuss at next standup (if it doesn't take too much time).  AFAIK 1300 UTC is 7am for some team members, and the 1300-1400 block has been problematic for other team members as well13:13
quiquellmyoung|ruck: 7am is a bitch, no need for discussions, I totally understand13:14
*** yolanda has joined #oooq13:16
quiquelltrown: Do you have brain cycle for me over bj ?13:16
myoung|rucksshnaidm|rover: have a moment to sync?13:17
sshnaidm|rovermyoung|ruck, yep13:17
myoung|ruckcool...my place or yours?13:17
sshnaidm|rovermyoung|ruck, paste a link13:17
*** yolanda_ has quit IRC13:18
myoung|rucksshnaidm|rover: https://bluejeans.com/705085945513:18
trownquiquell: sure13:19
quiquellcool, https://bluejeans.com/789106523213:19
*** saneax-_-|AFK is now known as saneax13:21
rlandytrown: weshay: we could use more design around the dry-run option. how we will trigger that. what it will spew out exactly13:22
rlandyhttps://review.openstack.org/#/c/567060/13:23
rlandystarted here13:23
*** amoralej|lunch is now known as amoralej13:31
sshnaidm|rovermyoung|ruck, supybot.plugins.GateStatus.changeIDs13:33
*** myoung|ruck is now known as myoung13:36
sshnaidm|rovermyoung, you don't hear me13:38
*** jaosorior has quit IRC13:43
myoungsshnaidm|rover: https://review.rdoproject.org/etherpad/p/ruckrover-sprint1313:47
myoung%gatestatus13:50
hubbotFAILING CHECK JOBS on stable/ocata: tripleo-ci-centos-7-undercloud-upgrades, tripleo-ci-centos-7-undercloud-oooq @ https://review.openstack.org/56429113:50
*** jaosorior has joined #oooq13:50
weshayquiquell, myoung trown rlandy panda fyi..  panda and I have reworded and reordered the cards13:50
weshayprobably needs some discussion after your review13:51
*** ykarel is now known as ykarel|away13:53
myoung%config plugins.GateStatus.changeIDs13:53
hubbotmyoung: I9af1028314859b63cab9f55d2c906b1344ca2300 I214272a6f25feb75496e44eb0a16269c6ee4cfe2 I4c5bdf00ce8cf7eabf669b248b99cb8443e82fab If12c8fe9bd0bea98a4842f279399285344f2224613:53
myoung%config plugins.GateStatus.changeIDs13:55
hubbotmyoung: I9af1028314859b63cab9f55d2c906b1344ca2300 I214272a6f25feb75496e44eb0a16269c6ee4cfe2 I4c5bdf00ce8cf7eabf669b248b99cb8443e82fab If12c8fe9bd0bea98a4842f279399285344f2224613:55
rlandyupdating review13:56
*** hubbot has quit IRC13:57
*** hubbot has joined #oooq13:57
myoung%config plugins.GateStatus.changeIDs13:58
hubbotmyoung: I0cbf9ffb8552411e4dd891c38702ff8d1f6db5b1 I214272a6f25feb75496e44eb0a16269c6ee4cfe2 I4c5bdf00ce8cf7eabf669b248b99cb8443e82fab If12c8fe9bd0bea98a4842f279399285344f2224613:58
myoungsshnaidm|rover:  \o/ ^^13:58
*** myoung is now known as myoung|ruck13:58
myoung|ruck%gatestatus13:58
hubbotFAILING CHECK JOBS on stable/ocata: tripleo-ci-centos-7-undercloud-upgrades, tripleo-ci-centos-7-undercloud-oooq @ https://review.openstack.org/56429113:58
weshaymyoung|ruck, ya.. fyi.. the program call is release specific so we need to target our status and summary to the release13:59
sshnaidm|roverhrybacki, ping14:13
hrybackisshnaidm|rover: pong14:13
sshnaidm|roverhrybacki, did you create a ovb stack on rdo cloud recently?14:13
hrybackisshnaidm|rover: not this week, no14:14
sshnaidm|roverhrybacki, can you try it? because I can't..14:14
hrybackisshnaidm|rover: sure, gimmige a few14:14
sshnaidm|roverhrybacki, ops told me to use smaller flavors, but I'm not sure how it works with ovb case14:15
hrybackisshnaidm|rover: yeah capacity is an issue but this again raises my point of -- how are we supposed to test out true CI?14:15
quiquell00..14:19
quiquellups14:20
weshaymyoung|ruck, sshnaidm|rover so is this fixed? https://bugs.launchpad.net/tripleo/+bug/177002114:22
openstackLaunchpad bug 1770021 in tripleo "hubbot: TQE check jobs have not been kicked by hubbot since 4/25" [Critical,Triaged]14:22
*** skramaja has quit IRC14:22
weshayasking because I see \0/14:23
sshnaidm|roverweshay, yep14:23
weshaycool, thanks14:23
myoung|ruckweshay: aye14:25
hrybackisshnaidm|rover: nope, I can't re-deploy either now14:25
hrybackiare medium flavors large enough?14:25
sshnaidm|roverhrybacki, maybe, need to check..14:26
sshnaidm|roverhrybacki, actually we need only stack created, no need to install there undercloud14:26
*** quiquell is now known as quiquell|off14:27
myoung|rucksshnaidm|rover, weshay, I just updated the LP with the details of what we did, and what the bot is monitoring now.  https://bugs.launchpad.net/tripleo/+bug/1770021/comments/114:29
openstackLaunchpad bug 1770021 in tripleo "hubbot: TQE check jobs have not been kicked by hubbot since 4/25" [Critical,Fix released] - Assigned to Sagi (Sergey) Shnaidman (sshnaidm)14:29
myoung|rucksshnaidm|rover: please update ^^ if I missed anything14:29
sshnaidm|rovermyoung|ruck, no, that's fine14:30
hrybackisshnaidm|rover: that's true (at this point anyways)14:34
hrybackimedium sized flavors would at least get you far enough to figure out the key issue?14:34
sshnaidm|roverhrybacki, I think so14:36
rlandyweshay: panda: myoung|ruck: trown: could you pls review the description, DoD and see if this captires what we intend?14:37
rlandyhttps://trello.com/c/Qnd1IJY1/753-create-a-new-upstream-zuul-job-that-executes-toci-dry-run-with-linter-and-unit-tests14:37
* myoung|ruck is CIX call but looks anyway14:38
myoung|ruckpanda: do you care if I drop the "Maturity: " from the labels?  All of your new labels start with "Needs", and the ":" character in label borks trello's filtering links on the URI (trello bug)14:39
trownrlandy: card looks good, I dont see any reason to not run the job on openstack infra14:40
trownrlandy: should be <10min job14:41
*** jaosorior has quit IRC14:41
trownrlandy: the patch is not exactly what I was thinking, but I will comment there14:42
*** apetrich has quit IRC14:43
*** chandankumar is now known as chkumar|off14:45
*** udesale has quit IRC14:46
rlandytrown: thanks - will review your comments14:49
*** apetrich has joined #oooq14:50
*** moguimar has quit IRC14:50
*** kopecmartin has quit IRC14:52
myoung|ruckrlandy: the card looks great to me.  IMHO a minor addition to DoD would be "a README.md exists (somewhere) that describes how to run the linter/dry run tests locally" - but that's just 'imho'15:06
rlandymyoung|ruck: sure - can add that15:06
pandarlandy: you should skip collectlogs for the dry run15:07
rlandytrown: thanks for your comments - wrt what the option kicking the dry run is named - let's bike shed a little here. I am fine changing the name but I am not sure about unit15:08
pandamyoung|ruck: Is it ok if we just change the : to - ?15:08
rlandyskipping the playbook run does not essentially mean we run unit tests15:08
rlandypanda: I was going to move the collected vars to the $LOGS fir to collect logs would then collect something15:09
myoung|ruckpanda: sure (caveat: for me personally visually parsing "(Maturity * 5)" for each and every card does not add value)15:09
rlandyI'll updated the patch and pls comment again there if you want it still removed15:09
trownrlandy: right but the toci_jobtype is what we are parsing there... it is not a cli option15:09
myoung|ruckpanda: "Needs $something" is more clear...but I'm not stuck on it.  The broken links are the actual issue.15:09
trownrlandy: so if we will run dryrun in a unit test job makes more sense to have the jobtype be named something not "dryrun"15:10
*** dtrainor has quit IRC15:10
trownrlandy: "dryrun" in a jobtype would make me think the job does nothing15:10
*** moguimar has joined #oooq15:10
rlandytrown: I don't know a better way to invoke this option w/o adding something to jobtype15:10
myoung|ruckpanda: if you update the labels to whatever-you-want-not-including-: in place they all autoupdate15:11
rlandyalso I am not sure how we plan to kick the unit tests15:11
pandacan we add a parameter to toci_quickstart.sh --dry-run ?15:11
trownrlandy: ya I think going off jobtype makes sense, but just pointing out what jobtype will want that option is a unittest job15:12
pandaalso the unit test don't necesarl=ily need to be invoked by any existing scripts15:12
rlandy^^ exactly15:12
rlandyI am not sure linking dry run and unit tests is correct15:13
rlandyquestion above15:13
rlandyhow do we plan on kicking the unit tests?15:13
rlandyI can think of a ansible role to do it15:13
rlandythen we wouls have a fs15:13
pandawith a call to tox ? maybe look at the other openstack-tox jobs ?15:14
rlandyand I would understand the unit test naming15:14
rlandyif we kick it indep, the two are unrelated15:14
trownpanda: we cant skip collectlogs if we want logs collected :P15:18
pandawhat do we collect from dry runs ?15:19
rlandyyep15:20
trownpanda: the log of the dry run15:20
rlandyI don't have the dry run data all in yet15:20
trownpanda: if we dont collect anything, what is the point of running the job15:20
trownpanda: oh nevermind... if we put stuff in $LOGS_DIR we are good without running collect-logs15:22
pandatrown: how much data is the current collect will be used in a dry run15:22
pandayep ok15:25
*** links has quit IRC15:25
*** saneax is now known as saneax-_-|AFK15:26
*** ykarel|away has quit IRC15:26
rlandyso we are still not agreed on the option name?15:26
*** saneax-_-|AFK is now known as saneax15:26
* rlandy fixes other comments15:26
trownrlandy: I am fine with the option name, just not the jobtype (aka name) that uses that option15:27
*** saneax is now known as saneax-_-|AFK15:28
rlandyok15:29
pandatrown: did you find a card to work on ?15:32
trownpanda: im working on python script with quiquell15:32
myoung|rucksshnaidm|rover: just saw this gate failure in the queue, and it seems odd to me...any idea what this might be?  https://review.rdoproject.org/etherpad/p/ruckrover-sprint13 @ L5615:32
trownpanda: do you have a card to work on?15:33
pandatrown: did you find a card to work on ??15:33
pandaoops15:33
pandatrown: you're working on this one https://trello.com/c/oSW5OQer/732-implement-the-script-that-creates-releases-file  ?15:34
trownyes15:34
pandatrown: ok, I can work on the unit tests card15:34
pandatrown: do we have any agreement on the output ?15:35
rlandywhy is https://trello.com/c/nkEay4We/752-create-a-linter-script-for-the-tool  in progress?15:35
trownpanda: ugh... there is now a seperate unit tests card? I thought we were doing tdd so the unit tests are what I am working on... as is quiquell15:35
pandarlandy: bevcause we started from the in progress column, and added cards there, I'll move to backlog again15:36
pandatrown: ok15:36
rlandypanda: ack - thanks15:36
rlandywe need to see what is really in flight here15:36
pandarlandy: the cards are also sorted by dependency, so how are you working on a job taht calls the linter if you don't know how to call it ?15:37
pandamaybe we created the cards after that one15:37
pandawe need to regroup15:37
pandatrown: rlandy 15 minutes meeting ?15:38
pandano more, promise15:38
rlandyok15:38
rlandyyour bj?15:39
pandayep15:39
trownim a hard stop at 15min15:40
trownmeetings are getting way out of hand15:40
*** tosky has quit IRC15:48
*** tosky has joined #oooq15:48
*** ykarel|away has joined #oooq15:55
hubbotFAILING CHECK JOBS on stable/ocata: tripleo-ci-centos-7-undercloud-upgrades, tripleo-ci-centos-7-undercloud-oooq @ https://review.openstack.org/56429115:57
myoung|rucksshnaidm|rover: fyi - adding rhos-13 jobs https://code.engineering.redhat.com/gerrit/#/c/13826916:00
*** zoli is now known as zoli|gone16:02
*** zoli|gone is now known as zoli16:02
*** bogdando has quit IRC16:06
*** dtrainor has joined #oooq16:13
*** ykarel|away has quit IRC16:16
*** panda is now known as panda|off16:18
*** dtrainor has quit IRC16:21
*** trown is now known as trown|lunch16:21
*** dtrainor has joined #oooq16:24
*** marios has quit IRC16:25
*** dtrainor has quit IRC16:26
*** hamzy has quit IRC16:28
*** jfrancoa has quit IRC16:31
*** udesale has joined #oooq16:32
myoung|rucksshnaidm|rover: FYI first rhos-13 gate job is in flight: https://rhos-dev-jenkins.rhev-ci-vms.eng.rdu2.redhat.com/job/tq-gate-rhos-13-ci-rhos-ovb-minimal-pacemaker-public-bond/1/console16:33
*** florianf has quit IRC16:34
*** sshnaidm|rover is now known as sshnaidm|off16:36
*** lucasagomes is now known as lucas-afk16:39
*** jfrancoa has joined #oooq16:45
weshayrlandy, ok.. had time to review https://wiki.openstack.org/wiki/Tripleo-upgrades-fs-variables16:49
weshaythat does indeed look good16:49
weshaypanda|off, rlandy given the scope in the card, I agree that can be closed16:50
weshayquiquell++16:50
panda|offwe are missing updates16:52
weshaytrue,16:52
*** atoth_ has joined #oooq16:55
*** atoth has quit IRC16:57
*** tosky has quit IRC16:57
weshayk.. panda|off thanks, I also updated where the job runs16:59
*** tosky has joined #oooq16:59
rlandycool17:08
*** tesseract has quit IRC17:15
weshaymyoung|ruck, https://thirdparty.logs.rdoproject.org/ appears to be down can please raise the issue w/ rdo-infra17:27
*** hamzy has joined #oooq17:27
*** hubbot has quit IRC17:28
myoung|ruckweshay: was about to hit enter in #rdo, then saw your ping17:29
*** trown|lunch is now known as trown17:29
*** dmellado has quit IRC17:29
*** atoth has joined #oooq17:34
*** atoth_ has quit IRC17:37
*** gkadam_ has quit IRC17:43
*** jtomasek has quit IRC17:44
*** udesale has quit IRC17:50
*** holser___ has quit IRC18:12
*** jtomasek has joined #oooq18:22
*** jfrancoa has quit IRC18:24
weshayrlandy, trown myoung|ruck fyi.. rdo services are down atm.. it will impact local reproducers etc18:32
weshaymyoung|ruck, sshnaidm|off expect it to impact promotion jobs as well18:33
weshayconversation in #rhos-ops18:33
rlandythanks - looking18:37
*** d0ugal has quit IRC18:37
*** d0ugal_ has joined #oooq18:37
weshayrlandy, re: the dry run.. here's a snip from a run from your review http://paste.openstack.org/show/720698/18:43
weshayis that roughly what you are getting as well?18:43
rlandylooking18:46
rlandyweshay: pretty much - you should see files in /tmp/logs18:47
rlandyI have more playbook output18:48
rlandygetting18:48
weshaycool.. nice start :)18:49
*** dtantsur is now known as dtantsur|afk18:49
rlandyweshay: https://paste.fedoraproject.org/paste/av9xRieSq4H2DrQh1AF4hA18:49
rlandyweshay: essentially people should not read the console output18:50
rlandythey get the relevant info in the log files18:50
rlandythat are put dir into the logs dir18:50
rlandyso no collect logs anymore18:50
rlandyif there are no playbook executed18:50
weshayrlandy, what do you think about putting some thought into breaking up " $QUICKSTART_INSTALL_CMD" so that it's very readable18:50
weshaylike every param on each own line as a possibility18:51
weshaynot sure if you do, but I find it very difficult to parse our playbook commands18:51
weshayansible-playbook18:51
weshay--e foo18:51
rlandyweshay: any playbook command18:51
weshay-e bar18:51
weshayetc18:51
weshayrlandy, ya18:51
* rlandy tests18:51
rlandylet's see what we can do18:51
weshayawesome thanks18:52
rlandyweshay: I also started a doc section18:52
rlandyvery prelim18:52
rlandybut somewhere to start collecting info18:52
*** dtrainor has joined #oooq18:54
*** dtrainor has quit IRC18:56
*** dtrainor has joined #oooq18:56
rlandytrown: wrt https://review.openstack.org/#/c/567060/4/toci_quickstart.sh - if I indent that text going into the collect_logs.sh file will be indented. we ok with that?18:58
*** dmellado has joined #oooq18:58
*** dtrainor has quit IRC18:59
*** dtrainor has joined #oooq18:59
trownrlandy: hmm ... maybe it doesnt matter... I had trouble reading it at first because I thought you were missing "fi" ... but bash doesnt care about indentation18:59
rlandyfine - ok18:59
trownmaybe we should move all that huge block of collect-logs stuff to a collect-logs function ...19:01
trownprobably not important for your patch though...19:01
*** hamzy has quit IRC19:01
rlandyI added the indents for now19:01
rlandylet's see how we go19:02
rlandyactually function is much cleaner - implementing19:04
*** hubbot has joined #oooq19:16
*** hamzy has joined #oooq19:18
*** agopi is now known as agopi|brb19:20
*** links has joined #oooq19:23
*** links has quit IRC19:45
*** atoth has quit IRC19:46
weshaymyoung|ruck, if you can reschedule the scrum tomorrow back 1/2 hour I can attend19:52
weshayresponded w/ no for now19:52
hubbotFAILING CHECK JOBS on stable/ocata: tripleo-ci-centos-7-undercloud-upgrades @ https://review.openstack.org/564291, master: tripleo-ci-centos-7-3nodes-multinode, tripleo-quickstart-extras-gate-newton-delorean-full-minimal @ https://review.openstack.org/56044519:58
myoung|ruckweshay: ack, moved back 30.  a few folks have requested it as well, will raise at retro as well.  I'll send out a doodle to the squads to see where folks would like to normally meet as well.20:07
*** matbu has quit IRC20:28
rlandyweshay: hello ...20:34
rlandyweshay: question about where to add the unittest job20:35
weshayrlandy, k20:35
weshayrlandy, so you guys are free to do what you think is best20:35
weshaybut imho20:35
rlandyweshay: so looking at https://github.com/openstack-infra/tripleo-ci/tree/master/zuul.d20:35
rlandywe could run a multinode job20:35
weshaythe dry-run is for humans20:35
rlandynamed with unittest20:36
rlandymaybe that is overkill20:36
weshaythe unit and lint run the tests20:36
weshayseems like they could be one job20:36
myoung|ruckweshay, rlandy, sshnaidm|off, so results from very first TQ gate job using rhos-13 are....*drumroll*20:36
myoung|ruck^^ Looks like we're hitting a bug weshay opened last year: https://bugs.launchpad.net/tripleo/+bug/1737356, a patch was posted (https://review.openstack.org/#/c/529548) which was -1 by slagel, and subsequently abandoned.20:37
openstackLaunchpad bug 1737356 in tripleo "Hostname validation for undercloud is too strict" [Medium,In progress] - Assigned to Martin André (mandre)20:37
weshayundercloud install fail :)20:37
rlandyhmmm20:37
rlandyhaven't looked at that in a while20:37
* myoung|ruck is attempting to understand why a non-FQDN is such a huge deal...not to mention why "subnode-0" works if that's the case20:38
weshayrlandy, so we generate some fake subunit data20:39
weshayand we also generate some subunit data for the ping test and a few other things20:40
rlandylooking at what we have for other units tests20:40
rlandyadding a unittest-jobs.yaml20:40
weshaythe linting could probably just be tacked onto openstack-tox-linters20:40
weshayrlandy, ya.. so if you think we should run the unit tests separately that is fine by me20:41
weshayrlandy, one could also just run https://review.openstack.org/#/c/567320/1/scripts/emit-releases-file.py w/ pytest20:42
weshayon a dryrun job that always runs too.. based on which files are changed etc.20:42
weshaywe could key off playbooks, toci..20:42
weshaypytest does exit 1, when it fails20:43
rlandyI would kick the dry run job is anything toci* changed20:43
weshayand if we generated subunit results they could also be added to elastic recheck..  sshnaidm|off could help there20:44
rlandyyeah and maybe fs?20:44
weshayfs000 ?20:44
weshaylolz20:44
rlandyfs-1?20:44
weshayya.. maybe we use a featureset for a dryrun20:44
weshaywe have a featureset00120:44
rlandyfs {-1}20:44
weshaylolz20:45
rlandybut that isn't actually what I meant20:45
rlandywhatever, I'll put in something just to get comments20:45
rlandythen I'll change it20:45
weshayrlandy, would be interesting to depends on https://review.openstack.org/#/c/567320/1/scripts/emit-releases-file.py20:45
weshayand execute w/ pytest20:45
weshayor go w/ it's own job.. get input from others20:46
weshaytbh I'm not sure what is "right"20:46
weshayI'll be back on later tonight20:47
rlandyok - no rush20:47
rlandyjust very little design on it so I am thinking out aloud on irc20:48
*** holser__ has joined #oooq20:48
weshaymyoung|ruck, this may also be related https://review.openstack.org/#/c/566911/20:49
weshaymyoung|ruck, may want to ff after that lands20:49
*** Goneri has quit IRC20:50
myoung|ruckack, i've made a note, 12 branches are 6 weeks old anyway, and 13 branches are 13 days old.  will FF both after this lands20:55
*** trown is now known as trown|outtypewww21:00
*** hamzy has quit IRC21:01
*** holser__ has quit IRC21:11
*** hamzy has joined #oooq21:11
*** hubbot has quit IRC21:13
*** dmellado has quit IRC21:15
*** hamzy has quit IRC21:15
myoung|ruckfyi - we still appear to have ongoing intermittent networking issues with dlrn, ops is on it21:16
*** dmellado has joined #oooq21:23
*** dmellado has quit IRC21:27
myoung|ruck%gatestatus21:29
*** hamzy_ has joined #oooq21:29
*** hamzy_ is now known as hamzy21:30
*** matbu has joined #oooq21:38
*** hubbot has joined #oooq21:49
*** myoung|ruck is now known as myoung|ruck|bbl21:59
*** hubbot has quit IRC22:01
*** hubbot has joined #oooq22:06
*** jtomasek has quit IRC22:07
*** rlandy has quit IRC22:09
sshnaidm|offmyoung|ruck|bbl, weshay it's related to undercloud install: https://review.openstack.org/#/c/566911/22:17
*** hamzy has quit IRC22:39
*** myoung|ruck|bbl is now known as myoung|ruck|off22:55
*** myoung|ruck|off is now known as myoung|ruck22:57
*** tosky has quit IRC23:38

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