Perl 6 - the future is here, just unevenly distributed

IRC log for #puppet-openstack, 2016-08-17

| Channels | #puppet-openstack index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
01:35 ducttape_ joined #puppet-openstack
01:47 ilbot3 joined #puppet-openstack
01:47 Topic for #puppet-openstack is now CI status: green - Puppet OpenStack modules http://docs.openstack.org/deve​loper/puppet-openstack-guide/
01:57 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: scenario003: re-enable trove  https://review.openstack.org/351953
02:01 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: Revert "run_tests: stop mistral testing"  https://review.openstack.org/339813
02:08 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: gnocchi: export db-sync output in logs  https://review.openstack.org/356178
02:47 openstackgerrit Michael Liu proposed openstack/puppet-magnum: Add itesm for keystone_authtoken using class keystone::resource::authtoken  https://review.openstack.org/351980
03:07 dhollinger joined #puppet-openstack
03:13 ducttape_ joined #puppet-openstack
03:39 openstackgerrit Keith Schincke proposed openstack/puppet-ceph: Update Jewel Ceph RGW to support Keystone V3  https://review.openstack.org/347956
04:36 ducttape_ joined #puppet-openstack
04:52 openstackgerrit Merged openstack/puppet-openstacklib: Add missing namevar method to openstack_config provider  https://review.openstack.org/355235
05:13 jaosorior joined #puppet-openstack
05:38 ducttape_ joined #puppet-openstack
05:46 rcernin joined #puppet-openstack
05:47 openstackgerrit Karthik S proposed openstack/puppet-neutron: Set firewall driver for SR-IOV ports  https://review.openstack.org/343505
06:06 openstackgerrit Merged openstack/puppet-tripleo: Defaults empty hash to kernel_modules and sysctl_settings  https://review.openstack.org/355615
06:06 openstackgerrit Merged openstack/puppet-tripleo: Align hiera keys with service names  https://review.openstack.org/353582
06:11 openstackgerrit OpenStack Proposal Bot proposed openstack/puppet-openstack-integration: Promote RDO repository to latest consistent URL  https://review.openstack.org/356245
06:31 sofie joined #puppet-openstack
06:35 openstackgerrit Charlie Llewellyn proposed openstack/puppet-cinder: Add variable to set availabilty zone fallback  https://review.openstack.org/356255
06:35 openstackgerrit Michele Baldessari proposed openstack/puppet-tripleo: Enable aodh systemd services on non-bootstrap nodes  https://review.openstack.org/356110
06:39 ducttape_ joined #puppet-openstack
07:00 tesseract- joined #puppet-openstack
07:00 gfidente joined #puppet-openstack
07:16 johnma joined #puppet-openstack
07:20 openstackgerrit Merged openstack/puppet-tripleo: Add cinder-backup profiles  https://review.openstack.org/311218
07:21 johnma left #puppet-openstack
08:00 markvoelker joined #puppet-openstack
08:07 ducttape_ joined #puppet-openstack
08:09 chem joined #puppet-openstack
08:21 openstackgerrit Carlos Camacho proposed openstack/puppet-tripleo: Fix deployment steps for AODH profiles  https://review.openstack.org/352372
08:24 priteau joined #puppet-openstack
08:31 paramite joined #puppet-openstack
08:33 dgurtner joined #puppet-openstack
08:54 openstackgerrit Carlos Camacho proposed openstack/puppet-nova: Removing WARNING: line has more than 140 characters in puppet-nova profiles  https://review.openstack.org/355409
08:55 xarses_ joined #puppet-openstack
08:57 derekh joined #puppet-openstack
08:59 openstackgerrit Lukas Bezdicka proposed openstack/puppet-ironic: Introduce ironic::ipxe class  https://review.openstack.org/354125
09:00 openstackgerrit Erno Kuvaja proposed openstack/puppet-manila: Add CephFSNative driver logic  https://review.openstack.org/353928
09:01 markvoelker joined #puppet-openstack
09:08 ducttape_ joined #puppet-openstack
09:25 gfidente joined #puppet-openstack
09:29 maestropandy joined #puppet-openstack
09:30 maestropandy left #puppet-openstack
09:32 openstackgerrit Merged openstack/puppet-tacker: Add authtoken class  https://review.openstack.org/351779
10:14 openstackgerrit Charlie Llewellyn proposed openstack/puppet-cinder: Fixed alginment and added test  https://review.openstack.org/356359
10:17 openstack joined #puppet-openstack
10:20 openstackgerrit Ivan Berezovskiy proposed openstack/puppet-oslo: Add support for oslo privsep management  https://review.openstack.org/356007
10:21 openstackgerrit Ivan Berezovskiy proposed openstack/puppet-nova: [WIP] Manage privsep helper for nova  https://review.openstack.org/356005
10:45 openstackgerrit Erno Kuvaja proposed openstack/puppet-ceph: Expose osd crush update on start option  https://review.openstack.org/356376
10:46 hjensas joined #puppet-openstack
10:46 hjensas joined #puppet-openstack
10:50 openstackgerrit Carlos Camacho proposed openstack/puppet-nova: Removing WARNING: line has more than 140 characters in puppet-nova profiles  https://review.openstack.org/355409
10:53 paramite joined #puppet-openstack
10:59 openstackgerrit Erno Kuvaja proposed openstack/puppet-ceph: Expose osd crush update on start option  https://review.openstack.org/356376
11:03 openstackgerrit Saravanan KR proposed openstack/puppet-neutron: SR-IOV numvfs config is made persistent across reboots  https://review.openstack.org/352494
11:07 openstackgerrit Erno Kuvaja proposed openstack/puppet-ceph: Expose osd crush update on start option  https://review.openstack.org/356376
11:10 ducttape_ joined #puppet-openstack
11:11 openstackgerrit Michele Baldessari proposed openstack/puppet-tripleo: Configure galera-monitor on all controller nodes  https://review.openstack.org/356390
11:32 openstackgerrit Alexey Deryugin proposed openstack/puppet-trove: Configure keystone authtoken options  https://review.openstack.org/346685
11:36 jaosorior joined #puppet-openstack
11:49 ansmith joined #puppet-openstack
11:53 EmilienM hello
11:59 morazi joined #puppet-openstack
12:09 markvoelker joined #puppet-openstack
12:10 ducttape_ joined #puppet-openstack
12:10 openstackgerrit Lukas Bezdicka proposed openstack/puppet-ironic: Introduce ironic::ipxe class  https://review.openstack.org/354125
12:12 EmilienM iberezovskiy, degorenko: hey! any progress on https://review.openstack.org/#/c/355632/ ? :)
12:16 cdelatte joined #puppet-openstack
12:16 dprince joined #puppet-openstack
12:17 EmilienM iberezovskiy, degorenko: tripleo is waiting for this patch, it's kind of urgent
12:18 chem joined #puppet-openstack
12:19 openstackgerrit Pradeep Kilambi proposed openstack/puppet-tripleo: Move ceilometer api to run under apache wsgi  https://review.openstack.org/352476
12:22 johnmilton joined #puppet-openstack
12:23 Jokke_ iurygregory: thanks for all your help on the reviews. https://review.openstack.org/#/c/353928/ is passing the tests again!
12:26 openstackgerrit Carlos Camacho proposed openstack/puppet-nova: Removing WARNING: line has more than 140 characters in puppet-nova profiles  https://review.openstack.org/355409
12:29 openstackgerrit Merged openstack/puppet-tripleo: Fix deployment steps for AODH profiles  https://review.openstack.org/352372
12:29 iurygregory Jokke_, you're welcome o/
12:30 iurygregory morning people
12:30 guimaluf joined #puppet-openstack
12:40 johnmilton joined #puppet-openstack
12:42 openstackgerrit Juan Antonio Osorio Robles proposed openstack/puppet-tripleo: Add manifests to inject and trust CA certificates  https://review.openstack.org/353021
12:45 openstackgerrit Charlie Llewellyn proposed openstack/puppet-cinder: Fixed alginment and added test  https://review.openstack.org/356359
12:46 iberezovskiy EmilienM, unfortunately, Fuel Ci is broken. so we can't merge our adapt patch. I think we need to unblock you first. but let's please wait for mwhahaha. I want to notify him about this problem
12:46 EmilienM ok
12:46 EmilienM we don't want to break you more :)
12:46 fultonj_ joined #puppet-openstack
12:47 degorenko EmilienM, fuel ci is already broken by our internal problems, so, it doesnt matter if we will merge your patch
12:47 degorenko but agree, lets wait for Alex
12:48 openstackgerrit Sven Anderson proposed openstack/puppet-openstack-integration: scenario003: Enable ec2api tempest tests  https://review.openstack.org/356442
12:48 EmilienM ok
12:52 iberezovskiy EmilienM, jfyi https://review.openstack.org/#/c/356007/
12:53 beagles and the winning recheck goes to iurygregory !
12:54 beagles thanks man
12:54 beagles did they disable osic or something?
12:55 EmilienM iberezovskiy: oh nice
12:55 EmilienM beagles: your patch got merged?
12:55 beagles EmilienM, yup
12:57 EmilienM beagles: congrats \o/
12:57 beagles EmilienM, it was a team effort :)
12:57 EmilienM and stars
12:59 beagles yup it looks like they disabled osic
13:00 ducttape_ joined #puppet-openstack
13:03 iurygregory beagles, haha tks \o/
13:06 mkarpin joined #puppet-openstack
13:08 EmilienM chem: what should we do with https://review.openstack.org/#/c/354872/ ?
13:09 EmilienM merge it until next version of puppet?
13:09 EmilienM it's blocking multiple patches
13:10 chem EmilienM: they have release it master and backported it, but yes, in the meantime we could merge it.  I'm just afraid that we just forget to remove it.
13:10 EmilienM chem: do they have released it?
13:11 chem EmilienM: merged to stable: 05a7ced, and merged to master 2 days ago
13:11 EmilienM ah, so it's not released
13:12 chem nope
13:14 chem EmilienM: I've just added a comment asking for a quick release.
13:16 fultonj_ joined #puppet-openstack
13:16 EmilienM nice!
13:17 chem in the meantime, we can merge it and let an note on the next weekling meeting to revert it if released
13:17 chem EmilienM: ^ ?
13:17 EmilienM chem: yes
13:17 EmilienM I did a recheck
13:18 EmilienM +2 now
13:18 chem ack, I'm updating https://etherpad.openstack.org/p/pupp​et-openstack-weekly-meeting-20160823
13:19 openstackgerrit Erno Kuvaja proposed openstack/puppet-ceph: Expose osd crush update on start option  https://review.openstack.org/356376
13:20 EmilienM chem: thx
13:22 iurygregory EmilienM, can you +A https://review.openstack.org/#/c/356014/ ? ^^
13:22 fultonj joined #puppet-openstack
13:23 EmilienM iurygregory: done!
13:23 EmilienM almost 100 meetings!
13:23 iurygregory yeah
13:23 iurygregory :D
13:24 iurygregory did you have sucess in know the TripleO problem in nova authtoken?
13:25 openstackgerrit Merged openstack/puppet-openstack-guide: Update meeting #90  https://review.openstack.org/356014
13:27 EmilienM iurygregory: not yet
13:28 iurygregory EmilienM, ok, i was about to try the recheck lol
13:28 EmilienM iurygregory: I did it, because the tripleo error was not related
13:28 iurygregory EmilienM, awesome :D
13:43 beagles EmilienM++
13:43 beagles "recheck this is insane" lol
13:43 EmilienM lol
13:48 xarses_ joined #puppet-openstack
13:54 danpawlik joined #puppet-openstack
13:55 ducttape_ joined #puppet-openstack
14:05 mwhahaha iberezovskiy: is it still broken?
14:08 dfisher joined #puppet-openstack
14:11 openstackgerrit Emilien Macchi proposed openstack/puppet-tempest: trove: add db_flavor_ref parameter  https://review.openstack.org/356491
14:11 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: scenario003: re-enable trove  https://review.openstack.org/351953
14:13 EmilienM tripleo CI is currently blocked to promote to latest trunk and we might need https://review.openstack.org/#/c/355632/ asap
14:14 EmilienM mwhahaha, iberezovskiy: would it be ok to land the puppet patch?
14:15 mwhahaha yea gimme a sec to try and catch up this morning, we can probably land it
14:15 iberezovskiy mwhahaha, as I can see it was fixed some minutes ago, but I didn't check
14:16 mwhahaha yea it looks like it's fixed now
14:29 xarses_ joined #puppet-openstack
14:38 EmilienM mwhahaha: should we recheck fuel CI?
14:41 iberezovskiy not yet
14:41 EmilienM ok
14:42 EmilienM iberezovskiy: it would be great if we could merge it today, just fyi
14:42 iberezovskiy yep, I understand
14:42 iberezovskiy I'm on it
14:45 karthiks EmilienM, regarding https://review.openstack.org/#/c/3​43505/3/manifests/agents/ml2/sriov.pp@78, if we need to set a fixed value in puppet, wouldn't it be appropriate to set it as default.
14:46 EmilienM karthiks: the default value in neutron doesn't work?
14:50 karthiks EmilienM, As far as I understand from  http://docs.openstack.org/mitaka/net​working-guide/adv-config-sriov.html, (section Enable neutron sriov-agent) this configuration needs to be given explicitly
14:50 EmilienM karthiks: ok then
14:50 EmilienM karthiks: +2
14:51 EmilienM and we're waiting for tripleo/fuel CI to be green
14:51 karthiks Thanks EmilienM :)
14:51 openstackgerrit Pradeep Kilambi proposed openstack/puppet-tripleo: Clean up unused params from aodh profiles  https://review.openstack.org/356523
15:04 openstackgerrit joined #puppet-openstack
15:10 openstackgerrit Keith Schincke proposed openstack/puppet-ceph: Update Jewel Ceph RGW to support Keystone V3  https://review.openstack.org/347956
15:18 dhollinger joined #puppet-openstack
15:27 colonwq xarses_, If I hit the patch enough, it will work and I will have a flat spot on my head.
15:40 ducttape_ joined #puppet-openstack
15:51 beagles EmilienM (and anyone else who wants to chime in) while I don't have much puppet experience - nor ruby for that matter - I do like to apply what I learn as I go. The reason I ask is I'm wondering how rigorous I should go on reviewing puppet patches. My feeling is be as rigorous as I know how and if I'm wrong somebody who knows better will correct me.
15:52 beagles thoughts?
15:52 EmilienM beagles: I started the review process by looking how others used to review and what were common use
15:53 EmilienM just pick some patches under review, and try to see if the patch has the doc, is tested, is following conventions: http://docs.openstack.org/developer/pu​ppet-openstack-guide/coding-style.html
15:53 beagles EmilienM, ack
15:53 EmilienM if it's passing CI, if the commit message is good
15:53 EmilienM for the code itself, it comes by reading existing code and be aware about how to write good puppet manifests
15:54 beagles the usual :)
15:57 beagles it's a bit interesting as each community has - even within the mundane things like commit messages - different standards
15:57 beagles so the reviewing the reviewers is the big thing I need to work at
15:57 ebalduf joined #puppet-openstack
15:57 beagles EmilienM, thanks
16:14 EmilienM yeah ! https://review.openstack.org/#/c/356491/ really helps to run trove tests
16:14 EmilienM if anyone can review it ^
16:14 EmilienM there is one more failure but I think it's a bug in trove/tempest tests, I'm on it
16:14 EmilienM http://logs.openstack.org/53/351953/8/ch​eck/gate-puppet-openstack-integration-3-​scenario003-tempest-centos-7/6bfc661/con​sole.html#_2016-08-17_14_48_58_392634
16:14 EmilienM but the 2 other failures we had are gone with my puppet-tempest patch
16:27 dprince joined #puppet-openstack
16:28 EmilienM not sure I'm doing right: https://review.openstack.org/356587
16:36 beagles degorenko, iberezovskiy, the reason I asked that question of EmilienM above is I was in the middle of reviewing https://review.openstack.org/#/c/344918/ and I got to wondering "what's too far?". Maybe you can advise me a bit here?
16:37 openstackgerrit Carlos Camacho proposed openstack/puppet-nova: Removing WARNING: line has more than 140 characters in puppet-nova profiles  https://review.openstack.org/355409
17:26 rcernin joined #puppet-openstack
17:29 iberezovskiy beagles, good comments I think
17:31 iberezovskiy beagles, the main idea on review for me is to do not say 'this is wrong', the idea is to describe what's exactly wrong and suggest how to fix, attach links for additional information if possible
17:35 beagles iberezovskiy, ack
17:35 openstackgerrit Merged openstack/puppet-openstack-integration: Promote RDO repository to latest consistent URL  https://review.openstack.org/356245
17:36 openstackgerrit Merged openstack/puppet-tempest: trove: add db_flavor_ref parameter  https://review.openstack.org/356491
17:37 beagles iberezovskiy, what are your thoughts on two resources conflicting with each other, ie 'eth0:4', 'eth0:1'
17:38 iberezovskiy it won't conflict because of different title
17:38 iberezovskiy but they will manage the same configuration file
17:38 beagles iberezovskiy, in puppet they will but they end up hitting the same file yeah
17:38 iberezovskiy I think it's the user issue, because he provided incorrect input
17:38 beagles iberezovskiy, it just ends up being a user issue
17:38 beagles yeah what you said :)
17:38 iberezovskiy :)
18:02 openstackgerrit Pradeep Kilambi proposed openstack/puppet-tripleo: Move ceilometer api to run under apache wsgi  https://review.openstack.org/352476
18:07 EmilienM iberezovskiy: thx a lot
18:09 iberezovskiy EmilienM, thx for the patience. mwhahaha can land both patches then
18:10 iberezovskiy I'm off for today
18:16 dmsimard EmilienM: hi
18:16 dmsimard This is sort of weird: https://github.com/openstack/puppet-openst​ack-integration/blob/master/Puppetfile#L88
18:17 dmsimard Since the module includes itself (rather than copy itself?), it installs the master branch of puppet-openstack-integration in /etc/puppet/modules instead of whatever version I'm executing run_tests.sh from
18:17 dmsimard so for instance I'm trying to reproduce the designate issues outside the gate to troubleshoot and it's unable to find openstack_integration::designate because the master branch doesn't have it
18:18 dmsimard Shouldn't p-o-i do a puppet module install of itself or something like that instead ?
18:22 mwhahaha zuul cloner overwrites it i think
18:22 mwhahaha in the gate
18:26 laurosn2 joined #puppet-openstack
18:27 _ducttape_ joined #puppet-openstack
18:27 dmsimard I'm testing outside the gate, though :p
18:28 EmilienM dmsimard: hi
18:28 EmilienM I see your problem
18:29 EmilienM you'll need to hack something in functions
18:32 dmsimard EmilienM: here? https://github.com/openstack/puppet-openst​ack-integration/blob/master/functions#L67
18:33 EmilienM maybe yeah
18:33 dmsimard and the hack would be to delete the openstack_integration module from /etc/puppet/modules and copy itself to /etc/puppet/modules ?
18:33 dmsimard :(
18:48 openstackgerrit Lars Kellogg-Stedman proposed openstack/puppet-tripleo: [WIP] updates to fluentd support  https://review.openstack.org/353507
18:51 EmilienM mwhahaha: we good for https://review.openstack.org/#/c/355632/ ?
18:51 mwhahaha not yet, trying to land the fuel fix (waiting on ci)
18:51 mwhahaha if it doesn't come through i'll just land that one anyway in a bit
18:52 mwhahaha https://review.openstack.org/#/c/356336/ if you're following along at home :D
18:52 EmilienM sorry to push
18:52 EmilienM the reason is tripleo CI blocked for 3 days
18:53 mwhahaha no it's ok, i'm hoping in the next hour
18:53 EmilienM kk
18:54 iurygregory EmilienM, can you ping when TripleO be unblocked? =)
18:57 ducttape_ joined #puppet-openstack
18:57 EmilienM iurygregory: when the puppet-ironic is merged :)
18:57 iurygregory EmilienM, ok ;D
18:59 paramite joined #puppet-openstack
19:15 danpawlik joined #puppet-openstack
19:20 mwhahaha life in openstack, watching paint dry aka waiting for CI
19:21 EmilienM lol
19:21 iurygregory mwhahaha, go watch the olympic games XD
19:21 danpawlik mwhahaha: xD
19:21 iurygregory while you wait for CI
19:22 EmilienM I don't have a TV :(
19:22 iurygregory EmilienM, internet LOL
19:22 EmilienM what is that?
19:22 mwhahaha nbc has a lock on the coverage in the US (and it's awful)
19:22 EmilienM I'm really old school
19:22 EmilienM my laptop crash when I run youtube video
19:23 iurygregory EmilienM, LOL
19:23 EmilienM I have a tablet
19:23 iurygregory you have a mac no?
19:23 EmilienM no
19:23 iurygregory sad =(
19:23 EmilienM old dell xps :P
19:23 mwhahaha he runs linux i think so that's probably the problem ;)
19:24 iurygregory linux is not the problem hahah we watch olympic games in the second monitor here at the lab :D
19:25 EmilienM I'm watching zuul
19:25 EmilienM which is more than olympic game
19:25 iurygregory Yeah
19:26 EmilienM "trying to make CI working" is like olympic sometimes
19:26 EmilienM but we'll survive
19:27 EmilienM if mwhahaha merges my patch
19:27 EmilienM :-P
19:27 EmilienM /rent
19:27 mwhahaha sooon
19:27 EmilienM /rant* even
19:27 iurygregory EmilienM, the sab pat is that you don't recive a gold medal for have the  CI working haha
19:27 iurygregory sad*
19:28 EmilienM we all receive a medal
19:28 EmilienM the medal will be Newton release
19:28 _ducttape_ joined #puppet-openstack
19:29 xarses_ joined #puppet-openstack
19:29 * mwhahaha grumbles about memory utilization
19:30 iurygregory EmilienM, good one :D
19:30 mwhahaha almost there on fuel ci, deploy passed just waiting for health checks so i can merge that patch (hopefully in the next 5-10 mins)
19:33 EmilienM ok
19:44 xarses_ joined #puppet-openstack
19:44 mwhahaha yay it passed come on jenkins
19:45 danpawlik joined #puppet-openstack
19:46 openstackstatus NOTICE: The volume for logs.openstack.org filled up rather suddenly, causing a number of jobs to fail with a POST_FAILURE result and no logs; we're manually expiring some logs now to buy breathing room, but any changes which hit that in the past few minutes will need to be rechecked and/or approved again
19:46 mwhahaha booo logs.openstack.org
19:47 EmilienM yay
19:49 mwhahaha hmm
19:50 ducttape_ joined #puppet-openstack
19:51 mwhahaha i guess i was looking at the wrong tab, it failed let me double check it's  not related and i might just force it
19:52 mwhahaha EmilienM: just merge your change to unblock tripleo and i'll work on getting the fuel patch landed soon
19:52 mwhahaha pesky CI shenanigans today
19:52 EmilienM mwhahaha: i don't want to break your CI too
19:52 mwhahaha it's already hosed for probably completely unrelated reasons
19:54 mwhahaha i've retriggered the fuel-ci for it so it'll be another 2 hours before it comes back
19:58 EmilienM dang
19:58 EmilienM we really need to land this patch in puppet-ironic
19:58 mwhahaha EmilienM: did you see https://review.openstack.org/#/c/355203/
19:58 EmilienM mwhahaha: no
19:58 EmilienM thanks
19:58 EmilienM looking
19:59 EmilienM mwhahaha: -2
19:59 mwhahaha :o
20:00 EmilienM -6
20:00 EmilienM I wish I could -6 sometimes
20:01 EmilienM mwhahaha: do we wait 2 more hours? lol
20:02 mwhahaha just land it
20:02 rwsu joined #puppet-openstack
20:02 mwhahaha it doesn't break noop or anything and we don't test ironic in CI
20:02 mwhahaha i'll get our fix landed later today
20:02 mwhahaha or i will
20:03 EmilienM ok thx
20:12 xarses_ joined #puppet-openstack
20:17 danpawlik joined #puppet-openstack
20:18 ayoung joined #puppet-openstack
20:21 EmilienM mwhahaha: wdyt about https://review.openstack.org/#/c/354872/ until puppetlabs release > 4.6.0 ?
20:21 mwhahaha i guess as a temporary thing
20:22 mwhahaha but the number of places we need to do that for various modules is quite annoying
20:22 EmilienM puppet-keystone is kind of prio
20:22 mwhahaha yea
20:22 EmilienM as it breaks beaker jobs
20:22 EmilienM other projects *only* have unit-latest broken
20:23 EmilienM which is non voting
20:23 mwhahaha better to pin to puppet agent 1.5.2?
20:23 mwhahaha cause that's ultimately what's doing it
20:23 mwhahaha i think we'll end up with weird parse order issues if we have to go include everything
20:26 mfisch chem: ping
20:27 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-guide: CI: updates about current jobs  https://review.openstack.org/354122
20:27 EmilienM mfisch: what's up
20:28 mfisch working on this weird domain Id bug with franza
20:28 mfisch we can see the default domain Id changing after a resource is made
20:28 mfisch kinda strange
20:29 mfisch http://paste.openstack.org/show/559094/
20:29 mfisch after we create the heat domain the default domain ID changes
20:29 EmilienM you running master?
20:29 mfisch which is wrong
20:29 EmilienM fork?
20:29 mfisch no
20:29 mfisch you think this is fixed?
20:29 mfisch we're a bit behind but thise code looks the same from what I saw
20:29 EmilienM no, just making sure we test the same code
20:29 EmilienM what is your manifest?
20:30 EmilienM can we reproduce in CI? our scripts test idempotency
20:30 mfisch We're from March 16
20:30 mfisch do we create a domain in any of our tests?
20:31 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: nova: switch libvirt_virt_type from qemu to kvm  https://review.openstack.org/356695
20:31 EmilienM dmsimard: ^
20:31 dmsimard EmilienM: http://logs.openstack.org/99/355599/1/ch​eck/gate-tempest-dsvm-full-ubuntu-xenial​/fbe5f9b/logs/etc/nova/nova.conf.txt.gz
20:31 EmilienM mfisch: lol
20:31 EmilienM is it a real question ahah
20:31 dmsimard virt_type = qemu
20:32 EmilienM dmsimard: ah so I can cancelt it
20:32 mfisch hah I dont even remember
20:32 EmilienM mfisch: https://github.com/openstack/puppet-keystone/blo​b/master/spec/acceptance/default_domain_spec.rb
20:32 EmilienM mfisch: look in https://github.com/openstack/puppet-​keystone/tree/master/spec/acceptance :)
20:32 dmsimard EmilienM: well, no
20:32 dmsimard EmilienM: feel free to let it go
20:33 EmilienM mfisch: we also test here: https://github.com/openstack/puppe​t-keystone/blob/master/spec/accept​ance/keystone_wsgi_apache_spec.rb
20:33 EmilienM dmsimard: let's see how it works :)
20:34 mfisch EmilienM: I bet I could get this to fail by changing the intergration tests
20:35 EmilienM sure, please try
20:35 EmilienM mfisch: again, what is your manifest?
20:35 mfisch EmilienM: adding to the bug
20:36 mfisch LP 1614141
20:36 openstack Launchpad bug 1614141 in puppet-keystone "Need to create heat domain without changing the default_domain_id" [Undecided,New] https://launchpad.net/bugs/1614141
20:36 mfisch pretty basic just some ensure resources
20:36 mfisch on keystone_domain
20:37 Jokke_ joined #puppet-openstack
20:38 openstackgerrit Merged openstack/puppet-ironic: conductor: add api_url parameter  https://review.openstack.org/355632
20:38 danpawlik joined #puppet-openstack
20:38 EmilienM mfisch: why don't use use https://github.com/openstack/puppet-heat/​blob/master/manifests/keystone/domain.pp ?
20:39 mfisch lol because of the other bug I filed earlier that you commented on
20:39 mfisch heat and Keystone on different nodes
20:39 mfisch thats where the code came from
20:39 EmilienM ah
20:40 EmilienM indeed
20:40 EmilienM what I would do is :
20:41 EmilienM 1) patch heat::keystone::domain to add a new conditional if $manage_heat_config
20:41 EmilienM 2) use heat::keystone::domain on keystone server and set manage_domain, manage_user and manage_role to True
20:42 EmilienM 3) use heat::keystone::domain and set $manage_heat_config to True and the other to False
20:43 EmilienM and honestly, I don't understand 100% the bug description
20:43 EmilienM "Even if I set the identity/default_domain_id to default, the mistake persists."
20:43 EmilienM from keystone.conf?
20:43 EmilienM did you configure something into heat.conf?
20:44 EmilienM and why did you mention the idempotency problem, which isn't mentionned in the bug report
20:47 mfisch EmilienM: I think I found it
20:47 mfisch weird corner case
20:47 mfisch will update the bug
20:47 EmilienM bug set to incomplete
20:47 EmilienM yeah and please make it a bit more clear
20:48 EmilienM mfisch: and also please submit a patch into puppet-heat
20:48 EmilienM do not duplicate code like this in your manifests, this is really durty
20:48 mfisch agreed on dirty
20:48 EmilienM your use case is really valid (containers)
20:48 mfisch EmilienM: this bug happens when the default domain is NOT in your catalog and a non-default domain IS
20:49 EmilienM mhh i'm not sure to get it :)
20:49 mfisch confirming now
20:49 mfisch you can make a catalog that does not define the default domain
20:49 mfisch but that does define a domain called "matt"
20:49 mfisch then this code fails
20:51 openstackgerrit David Moreau Simard proposed openstack/puppet-openstack-integration: Parameterize the libvirt_virt_type parameter so it can be overridden  https://review.openstack.org/356708
20:52 EmilienM mfisch: so in our CI we do not define a default domain
20:52 EmilienM https://github.com/openstack/puppet-openstack-in​tegration/blob/master/manifests/keystone.pp#L52
20:52 mfisch openning all your CI link spam again mon ami
20:52 EmilienM you say, if I add keystone_domain { 'mfish' } it will fail?
20:53 mfisch I can make it not fail if you set is_default to false on the mfish
20:53 mfisch the failure is just a warning
20:53 EmilienM mfisch: can you look https://github.com/openstack/puppet-openstack-​integration/blob/master/manifests/keystone.pp and try to submit a patch that test this corner case?
20:54 EmilienM I'm really unsure about how to reproduce it
20:54 EmilienM and you seem the best person to do it
20:54 mfisch EmilienM: can I just fork this test and push a review up?
20:54 mfisch will that show it?
20:54 iurygregory what error you get mfisch ?
20:54 mfisch iurygregory: a warning
20:54 mfisch let me update the bug since nothing is clear right now
20:55 mfisch but I could break this test easily
20:55 EmilienM mfisch: fork what?
20:55 EmilienM this is not a test
20:55 EmilienM this is a manifest
20:55 mfisch sure
20:55 mfisch an integraton manifest
20:55 mfisch I can make one that breaks by changing about 8 lines
20:55 EmilienM we have 3 scenarios, right?
20:55 EmilienM they run keystone
20:56 EmilienM so modify this keystone.pp to make it fail and we'll see what is the problem
20:56 mfisch yep ok
20:57 EmilienM thanks!
20:59 mfisch iurygregory: updated
20:59 mfisch https://bugs.launchpad.net/p​uppet-keystone/+bug/1614141
20:59 openstack Launchpad bug 1614141 in puppet-keystone "creating a keystone_domain can make it the default even though is_default is false" [Undecided,Incomplete]
20:59 mfisch see if that makes sense to you ^
20:59 mfisch I'm going to work on the integration for slave driver EmilienM ;)
21:00 openstackgerrit Charlie Llewellyn proposed openstack/puppet-cinder: Fixed alginment and added test  https://review.openstack.org/356359
21:03 iurygregory mfisch, tks I'll look at home =)
21:03 EmilienM mfisch: lol
21:04 mfisch iurygregory: can I have my guy re-file the bug in portuguese?
21:05 iurygregory mfisch, oh is the laurosn problem?
21:05 iurygregory O.o
21:05 mfisch iurygregory: this guy went to your school
21:05 mfisch he lives here now
21:05 danpawlik joined #puppet-openstack
21:05 iurygregory mfisch, oh awesome XD , there is no need to be in portugiese =)
21:07 mfisch I was j/k
21:07 mfisch just cool that 2 guys fro the same school working on this
21:08 iurygregory I'll try to help be sure o/ just need to finish internal things for tomorrow deadline
21:09 openstackgerrit Matt Fischer proposed openstack/puppet-openstack-integration: Break tests for 1614141  https://review.openstack.org/356726
21:10 mfisch EmilienM: I think thats all you need for a corner case ^
21:10 mfisch let me make a better commit msg
21:11 openstackgerrit Matt Fischer proposed openstack/puppet-openstack-integration: Break tests for 1614141  https://review.openstack.org/356726
21:13 mfisch iurygregory: so this was originally filed by a guy you met, Franza
21:14 mfisch Franciraldo
21:14 iurygregory mfisch, yeah i saw the comments in bug :D
21:17 EmilienM oh wow
21:17 EmilienM https://review.openstack.org/#/c/356695/
21:17 EmilienM dmsimard: ^ it pass CI
21:17 openstackgerrit David Moreau Simard proposed openstack/puppet-openstack-integration: Parameterize the libvirt_virt_type parameter so it can be overridden  https://review.openstack.org/356708
21:17 dmsimard EmilienM: lol wtf
21:17 dmsimard EmilienM: well
21:18 dmsimard EmilienM: I /think/ nested KVM works, even without kvm_nested=1
21:18 dmsimard It'll just be slow because hardware extensions aren't exposed
21:18 EmilienM virt_type=kvm
21:18 bkero It will fall back to qemu
21:18 dmsimard bkero: you think ?
21:18 bkero That's how I've seen it behave
21:18 EmilienM can you see in our logs? http://logs.openstack.org/95/356695/1/che​ck/gate-puppet-openstack-integration-3-sc​enario001-tempest-centos-7/1741c97/logs/
21:18 EmilienM if we actually run kvm?
21:18 bkero dmsimard: Yeah, I've seen whole clouds accidentally not kvm_nested=1, and everything worked. VMs were just slow.
21:19 EmilienM I see in http://logs.openstack.org/95/356695/1/check/g​ate-puppet-openstack-integration-3-scenario00​1-tempest-centos-7/1741c97/logs/dstat.txt.gz
21:19 dmsimard EmilienM: we'd probably see it here if tempest wouldn't kill the VMs http://logs.openstack.org/95/356695/1/check/​gate-puppet-openstack-integration-3-scenario​001-tempest-centos-7/1741c97/logs/ps.txt.gz
21:19 EmilienM qemu-kvm
21:19 EmilienM you can see it in dstat
21:19 dmsimard EmilienM: oh, and otherwise you'd just have qemu ?
21:19 EmilienM ah no
21:19 EmilienM qemu-kvm is also started on other jobs without this patch
21:20 EmilienM how can we check?
21:20 bkero That's just the binary name I think.
21:20 dmsimard yeah, you kinda need the whole process line and args to check what hypervisor it runs
21:20 EmilienM in libvirt logs? http://logs.openstack.org/95/356695/1/check/g​ate-puppet-openstack-integration-3-scenario00​1-tempest-centos-7/1741c97/logs/libvirt/qemu/
21:20 EmilienM I have one:
21:20 bkero The option is -enable-kvm
21:20 EmilienM http://logs.openstack.org/95/356695/1/ch​eck/gate-puppet-openstack-integration-3-​scenario001-tempest-centos-7/1741c97/log​s/libvirt/qemu/instance-0000000b.txt.gz
21:20 dmsimard like http://paste.openstack.org/show/559102/
21:20 bkero or I guess passing -machine=blah,accel=kvm
21:21 dmsimard http://logs.openstack.org/95/356695/1/ch​eck/gate-puppet-openstack-integration-3-​scenario001-tempest-centos-7/1741c97/log​s/libvirt/qemu/instance-0000000b.txt.gz looks right to me
21:21 EmilienM I have accel=kvm
21:21 bkero If it didn't throw a "no kvm support" message, it is probably enabled
21:21 EmilienM oh ben is right
21:21 bkero qemu throws that to stdout/stderr
21:21 EmilienM looks this is a job without kvm:
21:21 EmilienM http://logs.openstack.org/45/356245/1/ga​te/gate-puppet-openstack-integration-3-s​cenario001-tempest-centos-7/528a28f/logs​/libvirt/qemu/instance-0000000a.txt.gz
21:21 EmilienM and there is no accel=kvm
21:22 bkero accel=tcg. Wonder what tcg is
21:22 EmilienM let's see in syslog
21:22 EmilienM http://logs.openstack.org/95/356695/1/​check/gate-puppet-openstack-integratio​n-3-scenario001-tempest-centos-7/1741c​97/logs/syslog.txt.gz#_Aug_17_20_44_32
21:22 dmsimard that ran on centos-7-osic-cloud1-3573896
21:22 EmilienM Hypervisor detected: KVM
21:22 EmilienM ah nevermind
21:22 bkero TCG = tiny code generator, probably some HLE
21:22 EmilienM it's the VM itself
21:22 dmsimard I'd be curious to check something off of Rackspace, which runs Xen
21:23 bkero Do you have the output of /proc/cpuinfo?
21:23 bkero Can check for 'vmx' flag
21:23 dmsimard lol, all 6 jobs ran from osic
21:23 dmsimard T_T
21:23 bkero http://logs.openstack.org/45/356245/1/gate/ga​te-puppet-openstack-integration-3-scenario001​-tempest-centos-7/528a28f/logs/cpuinfo.txt.gz looks so
21:23 dmsimard Maybe osic exposes nested and stuff
21:23 dmsimard I would expect the job to fail on other providers
21:23 EmilienM we can do recheck to see if it works again
21:24 dmsimard EmilienM: can we do a recheck until we hit something else than osic
21:25 EmilienM maybe we can use facter to determine if nested virt is enabled
21:25 EmilienM and whether or not set qemu as value
21:25 EmilienM but yeah, it's weird kvm is not tested
21:26 EmilienM dmsimard: osic is not like other rackspace clouds FYI
21:26 EmilienM it's not rackspace cloud servers, etc
21:26 EmilienM it's another project afik
21:26 * dmsimard shrugs
21:27 bkero Different clouds pass different settings to their hypervisors. I wouldn't necessarily expect nested=1 everywhere
21:27 dmsimard yeah, exactly
21:27 gfidente joined #puppet-openstack
21:27 dmsimard especially since it's not a nodepool provider requirement
21:28 dmsimard (although I would've somehow expected otherwise)
21:28 bkero You could add a new label to zuul and tell it which providers have nested kvm, then have the jobs run on that.
21:28 dmsimard seriously, we're testing openstack on openstack clouds without nested virt as a requirement
21:28 dmsimard that sounds pretty awkward
21:28 dmsimard bkero: not a bad idea
21:31 openstackgerrit Merged openstack/puppet-openstack-guide: CI: updates about current jobs  https://review.openstack.org/354122
21:31 bkero dmsimard: That might not be accepted as an undue burden on a single provider
21:31 bkero What about doing multinode testing with ironic?
21:33 dmsimard yeah they're not interested in any way that would facilitate targetting nested virt deployment scenarios
21:34 bkero clarkb says it's unreliable
21:42 openstackgerrit Emilien Macchi proposed openstack/puppet-tripleo: (mitaka only) loadbalancer: manage ipv6 sysctl  https://review.openstack.org/356755
21:45 EmilienM couldn't we use an existing fact?
21:45 EmilienM to find out if kvm is available?
21:46 dmsimard EmilienM: but then you'd be sort of randomly testing nested or non-nested
21:46 EmilienM yeah
21:46 EmilienM I'll abandon my patch
21:46 EmilienM it's risky
21:47 EmilienM I'm sad
21:47 EmilienM we were the first one doing that :P
21:47 dmsimard EmilienM: I submitted https://review.openstack.org/#/c/356708/
21:47 dmsimard I'll leverage it with weirdo.
21:47 EmilienM I saw
21:47 EmilienM -6
21:47 dmsimard ?
21:47 EmilienM it's my new score
21:48 EmilienM ok I go running now, please don't break CI
21:51 dmsimard SPACESHIP OPERATOR
21:51 dmsimard I love the name of that operator.
21:51 iurygregory O.o
21:51 iurygregory spaceship?
21:51 iurygregory hahah
21:51 dmsimard iurygregory: https://docs.puppet.com/puppet/lat​est/reference/lang_collectors.html
21:52 dmsimard Resource collectors (AKA the spaceship operator)
21:52 iurygregory dmsimard, ohhhh
21:52 dmsimard <| |> looks like a spaceship :)
21:53 EmilienM <| |>      --  --  --  -- <| |>
21:53 EmilienM it's a battle of spaceships
21:53 EmilienM ok I go now
21:54 iurygregory hahahah awesome name
21:56 dmsimard Is it possible to check the value of another fact within a fact ?
21:57 dmsimard i.e, a custom fact that'd check the value of another fact in order to return an appropriate value
21:57 dmsimard or would I need to run a literal "facter" command within the custom fact to retrieve the value
21:59 dmsimard Ah, you can check with Facter.value(:fact)
21:59 dmsimard Cool.
22:26 dmsimard EmilienM: fyi, hacky but it works :p https://review.openstack.org/#/c/356​774/2/packstack/puppet/modules/packs​tack/lib/facter/default_hypervisor.rb
22:27 dmsimard EmilienM: I can submit it to openstacklib in a more generic format if you like the approach
22:27 dmsimard puppet-nova defaults to kvm
22:30 amoralej joined #puppet-openstack
22:30 openstackgerrit Daniel Pawlik proposed openstack/puppet-watcher: Change authtoken to class  https://review.openstack.org/356792
22:42 EmilienM dmsimard: yes it would work
22:42 EmilienM well I wouldn't have written a fact for that
22:42 EmilienM but it works
22:42 dmsimard EmilienM: it's a required hack in the context of packstack
22:42 EmilienM i met a spaceship when I ran
22:42 EmilienM not sure if it's /me tired
22:43 EmilienM dmsimard: if $::is_virtual { $hypervisor = 'qemu' } else { $hypervisor = 'kvm' } would also work
22:44 dmsimard That's exactly what Packstack has and it doesn't allow for someone to run qemu on baremetal or kvm on virtual.
22:44 dmsimard So I made the hypervisor configurable while keeping the same default/backwards compat
22:44 EmilienM excellent
22:44 EmilienM +6
23:10 EmilienM mfisch: your patch is working well https://review.openstack.org/#/c/356726/
23:10 EmilienM I don't see where could be the bug
23:11 EmilienM at least the puppet runs are idempotent
23:18 ayoung joined #puppet-openstack

| Channels | #puppet-openstack index | Today | | Search | Google Search | Plain-Text | summary