Perl 6 - the future is here, just unevenly distributed

IRC log for #puppet-openstack, 2017-01-18

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

All times shown according to UTC.

Time Nick Message
00:01 michapma_alt joined #puppet-openstack
00:09 alee joined #puppet-openstack
00:11 mwhahaha gah the ubuntu ceilometer package drops a vhost config file in place that messes with beaker
00:17 danpawli1 joined #puppet-openstack
00:17 mwhahaha hmm actually i think it's a dependency ordering issue
00:22 openstackgerrit Alex Schultz proposed openstack/puppet-ceilometer: Fix Ubuntu Ocata issues  https://review.openstack.org/421561
00:28 social_ joined #puppet-openstack
00:30 michapma_alt joined #puppet-openstack
00:58 michapma_alt joined #puppet-openstack
01:21 michapma_alt joined #puppet-openstack
01:25 ducttape_ joined #puppet-openstack
01:28 basilAB joined #puppet-openstack
01:29 stevemar joined #puppet-openstack
01:30 acormier joined #puppet-openstack
01:49 michchap_alt joined #puppet-openstack
02:14 michapma_alt joined #puppet-openstack
02:20 acormier joined #puppet-openstack
02:29 ducttape_ joined #puppet-openstack
02:37 michapma_alt joined #puppet-openstack
02:49 ilbot3 joined #puppet-openstack
02:49 Topic for #puppet-openstack is now CI status: green - Puppet OpenStack modules http://docs.openstack.org/deve​loper/puppet-openstack-guide/
03:07 michapma_alt joined #puppet-openstack
03:17 EmilienM mwhahaha: iirc, canonical was about adding vhost for telemetry services
03:19 mwhahaha Well it's breaking things so we'll see
03:22 EmilienM mwhahaha: jobs are non voting in a few min FYI
03:22 mwhahaha Cool. I'll try and figure out the ceilometer thing tomorrow.
03:22 agrebennikov joined #puppet-openstack
03:27 acormier_ joined #puppet-openstack
03:30 EmilienM mwhahaha: cool, I'm off now. Good night
03:31 mwhahaha Later
03:35 michapma_alt joined #puppet-openstack
03:36 openstackgerrit Merged openstack/puppet-manila: Prepare 9.5.0 release  https://review.openstack.org/421498
03:36 openstackgerrit Merged openstack/puppet-murano: Prepare 9.5.0 release  https://review.openstack.org/421500
03:36 openstackgerrit Merged openstack/puppet-ovn: Prepare 9.5.0 release  https://review.openstack.org/421506
03:36 openstackgerrit Merged openstack/puppet-magnum: Prepare 9.5.0 release  https://review.openstack.org/421497
03:41 acormier joined #puppet-openstack
04:05 openstackgerrit Merged openstack/puppet-aodh: Prepare 9.5.0 release  https://review.openstack.org/421487
04:05 openstackgerrit Merged openstack/puppet-barbican: Prepare 9.5.0 release  https://review.openstack.org/421488
04:06 openstackgerrit Merged openstack/puppet-ceilometer: Prepare 9.5.0 release  https://review.openstack.org/421513
04:06 openstackgerrit Merged openstack/puppet-cinder: Prepare 9.5.0 release  https://review.openstack.org/421489
04:06 openstackgerrit Merged openstack/puppet-designate: Prepare 9.5.0 release  https://review.openstack.org/421490
04:06 openstackgerrit Merged openstack/puppet-glance: Prepare 9.5.0 release  https://review.openstack.org/421491
04:06 openstackgerrit Merged openstack/puppet-gnocchi: Prepare 9.5.0 release  https://review.openstack.org/421492
04:06 openstackgerrit Merged openstack/puppet-heat: Prepare 9.5.0 release  https://review.openstack.org/421493
04:06 openstackgerrit Merged openstack/puppet-horizon: Prepare 9.5.0 release  https://review.openstack.org/421494
04:06 openstackgerrit Merged openstack/puppet-ironic: Prepare 9.5.0 release  https://review.openstack.org/421495
04:06 openstackgerrit Merged openstack/puppet-keystone: Prepare 9.5.0 release  https://review.openstack.org/421496
04:07 openstackgerrit Merged openstack/puppet-mistral: Prepare 9.5.0 release  https://review.openstack.org/421499
04:07 openstackgerrit Merged openstack/puppet-neutron: Prepare 9.5.0 release  https://review.openstack.org/421501
04:07 openstackgerrit Merged openstack/puppet-nova: Prepare 9.5.0 release  https://review.openstack.org/421502
04:07 openstackgerrit Merged openstack/puppet-openstack_extras: Prepare 9.5.0 release  https://review.openstack.org/421503
04:07 openstackgerrit Merged openstack/puppet-openstacklib: Prepare 9.5.0 release  https://review.openstack.org/421504
04:07 openstackgerrit Merged openstack/puppet-oslo: Prepare 9.5.0 release  https://review.openstack.org/421505
04:07 openstackgerrit Merged openstack/puppet-sahara: Prepare 9.5.0 release  https://review.openstack.org/421507
04:07 openstackgerrit Merged openstack/puppet-swift: Prepare 9.5.0 release  https://review.openstack.org/421508
04:07 openstackgerrit Merged openstack/puppet-tempest: Prepare 9.5.0 release  https://review.openstack.org/421509
04:07 openstackgerrit Merged openstack/puppet-trove: Prepare 9.5.0 release  https://review.openstack.org/421510
04:07 openstackgerrit Merged openstack/puppet-vswitch: Prepare 9.5.0 release  https://review.openstack.org/421511
04:08 openstackgerrit Merged openstack/puppet-zaqar: Prepare 9.5.0 release  https://review.openstack.org/421512
04:29 agrebennikov joined #puppet-openstack
04:37 udesale joined #puppet-openstack
04:49 agrebennikov_ joined #puppet-openstack
04:53 xarses joined #puppet-openstack
04:53 xarses joined #puppet-openstack
05:04 dgurtner joined #puppet-openstack
05:04 dgurtner joined #puppet-openstack
05:16 nmathew joined #puppet-openstack
05:21 skramaja joined #puppet-openstack
05:27 agrebennikov_ joined #puppet-openstack
07:18 bapalm joined #puppet-openstack
07:21 dgurtner joined #puppet-openstack
07:21 dgurtner joined #puppet-openstack
07:25 tesseract joined #puppet-openstack
07:25 rasca joined #puppet-openstack
07:26 tesseract joined #puppet-openstack
08:04 yprokule joined #puppet-openstack
08:08 hjensas joined #puppet-openstack
08:08 hjensas joined #puppet-openstack
08:10 openstackgerrit OpenStack Proposal Bot proposed openstack/puppet-openstack-integration: Updated from Puppet OpenStack modules constraints  https://review.openstack.org/421723
08:12 shardy joined #puppet-openstack
08:19 dgurtner joined #puppet-openstack
08:19 dgurtner joined #puppet-openstack
08:26 markvoelker joined #puppet-openstack
08:35 rmart04 joined #puppet-openstack
08:45 chem joined #puppet-openstack
08:52 jpich joined #puppet-openstack
08:56 rmart04 joined #puppet-openstack
09:00 gfidente joined #puppet-openstack
09:23 derekh joined #puppet-openstack
09:34 dgurtner joined #puppet-openstack
09:34 dgurtner joined #puppet-openstack
09:43 salmankhan joined #puppet-openstack
09:45 dgurtner joined #puppet-openstack
09:45 dgurtner joined #puppet-openstack
09:57 tobasco joined #puppet-openstack
09:58 owalsh_ joined #puppet-openstack
10:32 hjensas joined #puppet-openstack
10:32 hjensas joined #puppet-openstack
10:33 openstackgerrit Luke Hinds proposed openstack/puppet-horizon: Manage password_validator with puppet manifest  https://review.openstack.org/413653
10:38 dgurtner joined #puppet-openstack
10:54 paramite joined #puppet-openstack
11:05 salmankhan joined #puppet-openstack
11:11 richm joined #puppet-openstack
11:13 dgurtner joined #puppet-openstack
11:13 dgurtner joined #puppet-openstack
11:18 openstackgerrit Marios Andreou proposed openstack/puppet-nova: Add a class to run the db online_data_migrations  https://review.openstack.org/421249
11:39 openstackgerrit Alex Ruiz Estradera proposed openstack/puppet-neutron: Add new MidoNet plugin package  https://review.openstack.org/421847
11:45 hjensas joined #puppet-openstack
11:45 dprince joined #puppet-openstack
11:59 salmankhan joined #puppet-openstack
12:14 salmankhan joined #puppet-openstack
12:14 michapma_alt joined #puppet-openstack
12:15 cdelatte joined #puppet-openstack
12:21 acabot joined #puppet-openstack
12:28 chem joined #puppet-openstack
12:29 EmilienM hello
12:32 chem joined #puppet-openstack
12:33 iurygregory good morning
12:44 ducttape_ joined #puppet-openstack
12:58 mwhahaha or something
13:00 salmankhan joined #puppet-openstack
13:03 social joined #puppet-openstack
13:05 morazi joined #puppet-openstack
13:06 udesale joined #puppet-openstack
13:14 openstackgerrit Evgeny Antyshev proposed openstack/puppet-glance: Add ploop to supported disk formats  https://review.openstack.org/421889
13:15 ansmith joined #puppet-openstack
13:17 lhinds joined #puppet-openstack
13:24 iurygregory mwhahaha, we put scenarios is ubuntu for no-voting so we can approve https://review.openstack.org/#/c/420994/ ?
13:25 mwhahaha iurygregory: not sure, let me see. I might want to split the changes apart
13:25 iurygregory mwhahaha, ack
13:27 openstackgerrit Evgeny Antyshev proposed openstack/puppet-glance: Add ploop to supported disk formats  https://review.openstack.org/421889
13:27 morazi joined #puppet-openstack
13:37 ccamacho joined #puppet-openstack
13:37 fc__ joined #puppet-openstack
13:39 pradk joined #puppet-openstack
13:40 salmankhan joined #puppet-openstack
13:41 agrebennikov_ joined #puppet-openstack
13:46 openstackgerrit Alex Schultz proposed openstack/puppet-nova: Stop doing cell v2 in beaker  https://review.openstack.org/420994
13:53 ducttape_ joined #puppet-openstack
13:57 openstackgerrit Alex Schultz proposed openstack/puppet-nova: Fix libvirt for ubuntu  https://review.openstack.org/421917
14:02 openstackgerrit Alex Schultz proposed openstack/puppet-nova: Fixed libvirt package name for parallels virt_type  https://review.openstack.org/420032
14:04 fultonj joined #puppet-openstack
14:16 owalsh-brb joined #puppet-openstack
14:17 udesale joined #puppet-openstack
14:19 udesale joined #puppet-openstack
14:20 openstackgerrit David Moreau Simard proposed openstack/puppet-openstack-integration: Test: do not merge  https://review.openstack.org/421528
14:20 jperry joined #puppet-openstack
14:21 EmilienM wth https://review.openstack.o​rg/#/c/421723/1/Puppetfile
14:21 udesale joined #puppet-openstack
14:23 dmsimard wtf ?
14:23 dmsimard lol
14:23 dmsimard sounds like a bug in the update constraints script
14:25 jperry joined #puppet-openstack
14:29 dprince joined #puppet-openstack
14:30 shardy joined #puppet-openstack
14:31 hjensas joined #puppet-openstack
14:31 hjensas joined #puppet-openstack
14:42 chem joined #puppet-openstack
14:42 ccamacho joined #puppet-openstack
14:49 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: logs: add keystone-resources for debug  https://review.openstack.org/421980
14:49 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: logs: add keystone-resources for debug  https://review.openstack.org/421980
14:54 dfisher joined #puppet-openstack
14:58 bklei joined #puppet-openstack
15:02 nmathew joined #puppet-openstack
15:09 mwhahaha gah i guess i need to go back and align the stars, forgot the xenial beaker tests
15:10 openstackgerrit Alex Schultz proposed openstack/puppet-nova: Align stars to fix puppet-ci  https://review.openstack.org/420994
15:11 openstackgerrit Alex Schultz proposed openstack/puppet-nova: Fixed libvirt package name for parallels virt_type  https://review.openstack.org/420032
15:12 openstackgerrit Emilien Macchi proposed openstack/puppet-ironic: Allow to configure glance_api_version  https://review.openstack.org/422005
15:17 dmsimard mitaka scenario001 is not passing in RDO CI for trunk but it's passing in upstream gate
15:17 dmsimard (╯°□°)╯︵ ┻━┻
15:17 yprokule_ joined #puppet-openstack
15:17 dmsimard There has to be something different ..
15:17 EmilienM our stars are aligned
15:18 EmilienM yours are probably not.
15:18 furlongm_ joined #puppet-openstack
15:18 owalsh_ joined #puppet-openstack
15:18 pradk joined #puppet-openstack
15:18 yprokule joined #puppet-openstack
15:18 pradk joined #puppet-openstack
15:18 dmsimard it's failing on cirros image upload
15:18 openstackgerrit Emilien Macchi proposed openstack/puppet-ironic: Allow to configure glance_api_version  https://review.openstack.org/422005
15:18 dmsimard timing out after 170 seconds
15:18 dmsimard but I don't see anything in particular in logs
15:18 dmsimard going to compare more ..
15:18 blkperl_ joined #puppet-openstack
15:18 rochapor1o joined #puppet-openstack
15:18 LarsErik1 joined #puppet-openstack
15:19 fc__ joined #puppet-openstack
15:19 openstack joined #puppet-openstack
15:22 alee joined #puppet-openstack
15:22 mgagne joined #puppet-openstack
15:22 trown|mtg joined #puppet-openstack
15:22 radez joined #puppet-openstack
15:22 degorenko joined #puppet-openstack
15:22 khappone_ joined #puppet-openstack
15:22 jtomasek joined #puppet-openstack
15:22 zeus joined #puppet-openstack
15:22 acormier joined #puppet-openstack
15:23 mgagne joined #puppet-openstack
15:24 fvollero joined #puppet-openstack
15:27 mwhahaha joined #puppet-openstack
15:28 mwhahaha now that my irc client is back, i vaguely remember fixing such an issue in the mitaka timeframe around glance
15:29 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: logs: add keystone-resources for debug  https://review.openstack.org/421980
15:31 dmsimard mwhahaha: left OpenStack gate, right RDO CI https://www.diffchecker.com/LfTNqt4Z
15:31 dmsimard I'll check more later.. in two meetings simultaneously :(
15:32 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: Re-enable Designate on CentOS7  https://review.openstack.org/393738
15:34 petems joined #puppet-openstack
15:36 simonmcc joined #puppet-openstack
15:41 openstackgerrit Brent Eagles proposed openstack/puppet-octavia: Move db include from ::octavia to ::octavia::api  https://review.openstack.org/422031
15:41 davidmichaelkarr joined #puppet-openstack
15:41 jaosorior joined #puppet-openstack
15:47 EmilienM mwhahaha: can we make an etherpad to list blockers we currently have in ubuntu?
15:47 EmilienM unless everything is all under control
15:48 ducttape_ joined #puppet-openstack
15:48 dmsimard EmilienM: CI status is green so everything is cool
15:48 EmilienM dmsimard: where?
15:48 dmsimard /topic
15:49 EmilienM are you trolling us?
15:49 EmilienM :P
15:49 dmsimard dude it's in the topic
15:49 dmsimard lol
15:49 EmilienM I've lost admin
15:49 morazi joined #puppet-openstack
15:57 mwhahaha yea i'll get the list today, been meeting it up this morning so i'm not sure of the complete status
15:57 xarses joined #puppet-openstack
15:58 _ducttape_ joined #puppet-openstack
15:58 xarses_ joined #puppet-openstack
15:59 xarses_ joined #puppet-openstack
16:00 rajinir joined #puppet-openstack
16:03 EmilienM let's dress a list so we can kill it together
16:06 mwhahaha k let me start one
16:09 mwhahaha https://etherpad.openstack.o​rg/p/puppet-ubuntu-ocata-m2
16:09 dmsimard mwhahaha: only thing I see that's relevant in the difference between rdo and upstream CI for mitaka is KVM version which is 2.6 (from virt SIG repo) in upstream whereas RDO CI uses base OS 1.5 .. I don't see how that could cause a glance image upload timeout :(
16:09 mwhahaha it's what i've got so far
16:10 * dmsimard keeps looking
16:10 mwhahaha dmsimard: link to logs?
16:10 dmsimard mwhahaha: http://logs.openstack.org/28/421528/2/c​heck/gate-puppet-openstack-integration-​3-scenario001-tempest-centos-7/652b133/
16:11 dmsimard and https://ci.centos.org/artifacts/rdo/weirdo-generic​-puppet-openstack-scenario001/846/weirdo-project/
16:11 dmsimard that hardware it ran on is fairly slow AMD .. but it shouldn't take more than 170s to upload cirros :(
16:12 mwhahaha it's a ceph backend
16:12 mwhahaha so maybe it's a ceph issue
16:12 dmsimard I don't see any stack traces or errors anywhere, it seems like the upload eventually finishes but I don't see the ceph status update to take account for image being uploaded
16:12 dmsimard https://ci.centos.org/artifacts/rdo/weirdo-​generic-puppet-openstack-scenario001/846/we​irdo-project/logs/ceph/ceph-mon.n21.txt.gz
16:13 dmsimard vs http://logs.openstack.org/28/421528/2/check/g​ate-puppet-openstack-integration-3-scenario00​1-tempest-centos-7/652b133/logs/ceph/ceph-mon​.centos-7-osic-cloud1-s3700-6743322.txt.gz
16:14 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: logs: add keystone-resources for debug  https://review.openstack.org/421980
16:16 dmsimard I retried the job FWIW -- twice failed on the cirros image issue
16:18 mwhahaha seems to be hanging on the ceph upload on the backend
16:19 mwhahaha but not sure why
16:19 shardy mwhahaha: Hey, when you get a sec can we pls discuss https://review.openstack.org/#/c/421341/?
16:19 shardy mwhahaha: it's a blocker for tripleo upgrades, and I'm not sure on the best way to proceed
16:19 shardy basically there's a mismatch between puppet-swift and the swift config file defaults
16:20 mwhahaha shardy: sure, what's up
16:20 mwhahaha puppet swift doesn't really get much attention so if something's changed it's probably really out of date
16:21 shardy mwhahaha: So, the swift default pipeline contains entries like container-quotas and account-quotas
16:21 shardy but puppet-swift has hard-coded assumptions that it's container_quotas and account_quotas
16:21 shardy this means I've got to either explicitly configure the pipeline via puppet, or we need to align the values in the module with those in swift
16:22 shardy the latter is what I posted, but it breaks puppet CI because that does configure the pipeline AFAICS
16:22 shardy https://github.com/openstack/swift/blob/​master/etc/proxy-server.conf-sample#L94
16:23 openstackgerrit Marios Andreou proposed openstack/puppet-nova: Add a class to run the db online_data_migrations  https://review.openstack.org/421249
16:23 mwhahaha i don't think we configure the pipeline
16:23 mwhahaha we don't use the site.pp
16:23 mwhahaha so either we don't have a package that udnerstand the - vs _, or there's something else
16:24 shardy http://logs.openstack.org/41/421341/1/che​ck/gate-puppet-openstack-integration-4-sc​enario002-tempest-ubuntu-xenial-nv/f26f2c​9/logs/etc/swift/proxy-server.conf.txt.gz
16:24 mwhahaha http://logs.openstack.org/41/421341/1/ch​eck/gate-puppet-openstack-integration-4-​scenario002-tempest-centos-7/2ec4741/log​s/swift/proxy.txt.gz#_Jan_18_11_09_35
16:24 mwhahaha container_quotas
16:24 mwhahaha it's expecting that
16:24 shardy This is where I got confused - that has container_quotas, but my local deploys via tripleo have container-quotas
16:24 shardy e.g the same as the swift defaults
16:25 mwhahaha so let's look at the scenario, maybe that's getting passed in
16:25 shardy https://github.com/openstack/puppet-sw​ift/blob/master/manifests/proxy.pp#L18
16:25 shardy so basically it looks like "pipeline" there is mandatory, and it's got to match the assumptions in the module
16:26 mwhahaha https://github.com/openstack/puppet-openstack-in​tegration/blob/master/manifests/swift.pp#L41-L45
16:26 mwhahaha the test needs to be updated
16:26 mwhahaha or rather our scenario config
16:27 shardy ah, that explains it
16:27 * shardy ponders chicken/egg scenario with the two repos
16:27 mwhahaha drop quota config?
16:27 mwhahaha update, and replace?
16:27 shardy yeah I guess it'll need to be a series of three patches?
16:27 mwhahaha i'm thinking yes
16:27 openstackgerrit Ricardo Noriega proposed openstack/puppet-neutron: Add support for BGPVPN plugin  https://review.openstack.org/422051
16:27 shardy mwhahaha: Ok, thanks, let me give that a try :)
16:28 mwhahaha patch to drop the quota, fix swift, add back into pipeline
16:28 openstackgerrit Marios Andreou proposed openstack/puppet-nova: Add a class to run the db online_data_migrations  https://review.openstack.org/421249
16:30 ducttape_ joined #puppet-openstack
16:31 openstackgerrit Steven Hardy proposed openstack/puppet-openstack-integration: Temporarily remove *_quotas from swift pipeline  https://review.openstack.org/422084
16:32 openstackgerrit Steven Hardy proposed openstack/puppet-swift: Make account/container quotas syntax match default pipeline  https://review.openstack.org/421341
16:33 * mwhahaha brb
16:33 EmilienM mwhahaha: sounds like https://review.openstack.org/#/c/421561/2 fails to run beaker jobs (apache is not starting)
16:35 openstackgerrit Steven Hardy proposed openstack/puppet-openstack-integration: Temporarily remove *_quotas from swift pipeline  https://review.openstack.org/422084
16:35 openstackgerrit Steven Hardy proposed openstack/puppet-openstack-integration: Re-add container/account quotas to swift pipeline  https://review.openstack.org/422087
16:36 _ducttape_ joined #puppet-openstack
16:38 openstackgerrit Michael Bayer proposed openstack/puppet-neutron: Allow db sync timeouts to be configurable  https://review.openstack.org/422091
16:39 openstackgerrit Michael Bayer proposed openstack/puppet-neutron: Allow db sync timeouts to be configurable  https://review.openstack.org/422091
16:40 openstackgerrit Matthew J Black proposed openstack/puppet-ovn: Northd class include the vswitch::ovs class  https://review.openstack.org/422093
16:41 openstackgerrit Michael Bayer proposed openstack/puppet-neutron: Allow db sync timeouts to be configurable  https://review.openstack.org/422091
16:46 openstackgerrit Merged openstack/puppet-cinder: Deprecate host option  https://review.openstack.org/420768
16:55 salmankhan joined #puppet-openstack
16:56 ducttape_ joined #puppet-openstack
17:00 mwhahaha EmilienM: ok i hadn't looked at that one today yet
17:09 rmart04 joined #puppet-openstack
17:10 * mjblack grumbles about download.ceph.com errors
17:38 markvoelker joined #puppet-openstack
17:44 markvoelker_ joined #puppet-openstack
17:58 markvoelker joined #puppet-openstack
18:08 openstackgerrit Michael Bayer proposed openstack/puppet-neutron: Allow db sync timeouts to be configurable  https://review.openstack.org/422091
18:09 owalsh-afk joined #puppet-openstack
18:16 shardy joined #puppet-openstack
18:43 openstackgerrit Luke Hinds proposed openstack/puppet-horizon: Manage password_validator with puppet manifest  https://review.openstack.org/413653
18:48 pradk hmm seeing issues with puppet-ceilometer ci .. https://review.openstack.org/#/c/420908/
19:00 EmilienM pradk: yes, ubuntu broke us
19:00 EmilienM pradk: https://etherpad.openstack.o​rg/p/puppet-ubuntu-ocata-m2
19:00 EmilienM mwhahaha has it covered
19:00 pradk ah
19:03 openstackgerrit David Critch proposed openstack/puppet-ceph: Adding support to force zapping of ceph OSDs and journals  https://review.openstack.org/420992
19:20 iurygregory yay merged  https://review.openstack.org/420994
19:39 EmilienM mwhahaha: for https://review.openstack.org/#/c/421561/
19:40 EmilienM it's an idempotency issue I think
19:40 mwhahaha it is
19:40 mwhahaha i  just haven't had time to figure it out yet :D
19:40 mwhahaha working on stable/newton release at the moment
19:40 EmilienM oh sorry
19:40 mwhahaha the problem is that the package is delivering a vhost config after we've already run our apache config which clears out the vhost
19:43 mwhahaha amoralej|off, EmilienM: FYI, https://review.openstack.org/422183
19:44 EmilienM already +1 :D
19:44 EmilienM you're damn slow
19:44 * EmilienM runs
19:44 mwhahaha :o
19:44 * mwhahaha makes RDO uses hashes in the future.
19:44 * EmilienM hits the door mwhahaha just closed
19:45 mwhahaha ok back to figuring out what happened to ceilometer and aodh
19:53 mwhahaha https://github.com/openstack/puppet-openstack-inte​gration/blob/master/manifests/keystone.pp#L80-L88
19:53 mwhahaha so we do that for keystone in p-o-i
19:53 mwhahaha should we continue this pattern for ceilometer?
19:54 EmilienM mwhahaha: yeah
19:54 mwhahaha k
19:54 mwhahaha i mean basically we just need to ensure the package was installed prior to apache
19:54 EmilienM mwhahaha: AFIK we haven't got any negative feedback about it
19:54 mwhahaha which i think is the more reasonable thing to do than patch in p-o-i
19:54 EmilienM we can do it also
19:54 mwhahaha let me try the package before apache thing
19:55 openstackgerrit Ade Lee proposed openstack/puppet-nova: Set correct ownership for nova keytab  https://review.openstack.org/422200
19:55 alee joined #puppet-openstack
19:56 alee mwhahaha, EmilienM  https://review.openstack.org/422200  should be a quick one
19:57 mwhahaha alee: make sure to rebase, we just landed some ci fixes
19:57 mwhahaha alee: so it may fail if you aren't on head
19:57 EmilienM alee: no tests?
19:57 alee well - there is the test to show that the command is run as expected ..
19:57 EmilienM ah
19:57 EmilienM without details
19:58 openstackgerrit Alex Schultz proposed openstack/puppet-ceilometer: Fix Ubuntu Ocata issues  https://review.openstack.org/421561
19:59 alee EmilienM, you want me to change the test to be is_expected.to get service user keytab and set ownership  ?  or something like that?
19:59 openstackgerrit Ade Lee proposed openstack/puppet-nova: Set correct ownership for nova keytab  https://review.openstack.org/422200
19:59 EmilienM alee: i'm fine with it
20:00 alee EmilienM, ok
20:00 amoralej|off thanks mwhahaha
20:03 markvoelker_ joined #puppet-openstack
20:14 mwhahaha Could not set 'present' on ensure: Network is unreachable - connect(2) for "download.ceph.com"
20:14 * mwhahaha sighs
20:15 EmilienM mwhahaha: wait, we're supposed to use mirror
20:15 EmilienM mwhahaha: where did you see that
20:15 mwhahaha beaker
20:15 EmilienM do we use repos.pp?
20:15 mwhahaha http://logs.openstack.org/61/421561/3/che​ck/gate-puppet-ceilometer-puppet-beaker-r​spec-ubuntu-xenial/57d3241/console.html
20:15 mwhahaha yes
20:15 EmilienM yes we do, ::openstack_integration::repos
20:15 EmilienM mhh
20:15 mwhahaha http://logs.openstack.org/61/421561/3/check/gate-p​uppet-ceilometer-puppet-beaker-rspec-ubuntu-xenial​/57d3241/console.html#_2017-01-18_20_05_00_505908
20:15 mwhahaha unknown variable $::ceph_mirror_host
20:16 EmilienM http://logs.openstack.org/61/421561/3/check/ga​te-puppet-ceilometer-puppet-beaker-rspec-ubunt​u-xenial/57d3241/logs/apt-cache-policy.txt.gz
20:16 EmilienM wth
20:16 mwhahaha Warning: Unknown variable: '::ceph_mirror_host'. at /etc/puppetlabs/code/modules/openstack​_integration/manifests/repos.pp:59:20
20:17 openstackgerrit David Critch proposed openstack/puppet-ceph: Adding support to force zapping of ceph OSDs and journals  https://review.openstack.org/420992
20:17 EmilienM I know why
20:17 EmilienM because in run_tests.sh we do:
20:17 EmilienM export FACTER_ceph_mirror_host="h​ttp://${CEPH_MIRROR_HOST}"
20:17 EmilienM but beaker doesn't run this script
20:21 acormier joined #puppet-openstack
20:21 openstackgerrit Merged openstack/puppet-nova: Fixed libvirt package name for parallels virt_type  https://review.openstack.org/420032
20:36 _ducttape_ joined #puppet-openstack
20:39 mwhahaha not that anyone cares, but if you ever want a quick way to run a beaker test on a xenial throwaway vm -> https://gist.github.com/mwhahaha/​35c7a63de0910cf8ba76df0d6cef2b7e
20:40 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: CI test - never merge  https://review.openstack.org/422228
20:51 openstackgerrit Michael Bayer proposed openstack/puppet-neutron: Allow db sync timeouts to be configurable  https://review.openstack.org/422091
20:53 alee mwhahaha, going to do the ensure_resource() line you suggested -- what does the spec test for that look like?
20:53 mwhahaha alee: a normal file check
20:54 alee mwhahaha, ok
20:54 mwhahaha ensure_file(<keyfile>).with( :owner => nova'
20:54 mwhahaha ensure_file(<keyfile>).with( :owner => <nova user> )
20:54 mwhahaha something to that effect
20:56 mwhahaha sorry contain_file
21:09 openstackgerrit Luke Hinds proposed openstack/puppet-horizon: Manage password_validator with puppet manifest  https://review.openstack.org/413653
21:21 dmsimard mwhahaha: "makes RDO uses hashes in the future."
21:21 dmsimard what do you mean ?
21:22 mwhahaha something about not releasing actual versions anymore :D
21:22 mwhahaha git hashes for all
21:22 mwhahaha no more tags/branches
21:24 dmsimard you can deploy from source if you want
21:24 * dmsimard pouts
21:25 * mwhahaha replaces dmsimard with a tiny container
21:25 dmsimard that's good, containers are the future and solves every problem
21:25 mwhahaha correct. we don't actually need configuration management anymore
21:28 ducttape_ joined #puppet-openstack
21:28 dgurtner joined #puppet-openstack
21:28 dgurtner joined #puppet-openstack
21:29 openstackgerrit Oliver Walsh proposed openstack/puppet-nova: Fix cell_v2 setup  https://review.openstack.org/422248
21:33 openstackgerrit Ade Lee proposed openstack/puppet-nova: Set correct ownership for nova keytab  https://review.openstack.org/422200
21:33 alee mwhahaha, ^^
21:34 openstackgerrit Luke Hinds proposed openstack/puppet-horizon: Manage password_validator with puppet manifest  https://review.openstack.org/413653
21:40 openstackgerrit Luke Hinds proposed openstack/puppet-horizon: Manage password_validator with puppet manifest  https://review.openstack.org/413653
21:51 fultonj joined #puppet-openstack
21:59 owalsh mwhahaha: thanks for the review - I was kinda expecting that to stand out
21:59 mwhahaha there's probably some more since we backported the setup stuff we might need to deprecate some of the items
22:00 mwhahaha owalsh: this is the problem with implementing things and them not being done
22:00 owalsh mwhahaha: amen
22:01 owalsh mwhahaha: not sure what I can do about create_cell...
22:01 mwhahaha is list cell not landed yet?
22:01 mwhahaha because it should be be a list cell | grep -q <cellname>
22:02 owalsh mwhahaha: name is optional
22:02 owalsh it's transport_url & database_connection tuple that is significant
22:02 mwhahaha and this is why you guys need to deploy stuff
22:02 mwhahaha cause that's not how the world works
22:02 owalsh I ain't a nova dev :-)
22:03 mwhahaha yea but you can relay such feedback ;) no one listens to us anyway
22:03 mwhahaha so for us, name should not be optional
22:03 mwhahaha so our expectation is that it'll always be there
22:04 mwhahaha is cell0 some special thing? i thought the whole point was we were going to be deploying just cell0 not cell0 and cell1
22:04 owalsh yea, cell0 is the graveyard
22:05 mwhahaha i am here, https://imgflip.com/s/meme​/Computer-Guy-Facepalm.jpg
22:05 owalsh actually, in this case there should be exactly 2 cells
22:06 owalsh ... so I could add an unless param
22:07 mwhahaha let me review all that stuff again
22:07 mwhahaha we need the unless
22:07 mwhahaha the way it currently exists is not ok
22:07 mwhahaha but i also thing we might want to extract the create out into it's own thing so it's reusable
22:14 owalsh would need to create a db too I believe
22:14 * mwhahaha flips tables
22:15 mwhahaha and this is why this is a terrible thing
22:15 owalsh "If you think you understand quantum mechanics, you don't understand quantum mechanics."
22:16 owalsh s/quantum mechianics/cell_v2/g
22:16 mwhahaha if you don't understand your users, don't write code.
22:16 mwhahaha that's my current mantra as of late
22:17 mwhahaha so we can drop the db in the cell creation, that's probably not  a huge problem
22:17 owalsh right now I don't think we need to worry about creating multiple cell_v2 cells, so I' abandoned creating a resource provider for this...
22:18 mwhahaha right but i'm not looking to rewrite this next cycle
22:18 mwhahaha so i'm not pushing for the provider
22:18 mwhahaha but the setup needs to be isolated and done cleanly so we  don't have to do this  again
22:18 mwhahaha plus if we can't understand how to do it, how will our users be able to
22:19 owalsh simple_cell_setup :-)
22:19 mwhahaha which is clearly not so simple
22:19 owalsh joking... we are basically doing the initial part in puppet
22:22 mwhahaha replied, so I think we need to move the cell1 create out of that class
22:23 mwhahaha the issue with the db creation thing is that I know for a fact some people need that to be separate from the nova manage code as they may need to run on completely different hosts
22:24 * owalsh looking
22:25 owalsh ok, so if an Newton -> Ocata upgrade was performed it will have run simple_cell_setup and created "cell1" minus the name
22:26 mwhahaha it's possible
22:27 mwhahaha but that might not be the case
22:30 owalsh but I can't use nova-manage cell_v2 list_cell | grep -q ${title}
22:30 mwhahaha why?
22:30 openstackgerrit Alex Schultz proposed openstack/puppet-nova: Revert "Add cell_v2 simple_cell_setup"  https://review.openstack.org/422285
22:30 owalsh it's the transport_url, database_connection tuple that's the primary key here
22:30 mwhahaha is name not returned?
22:30 mwhahaha in the list?
22:31 owalsh if it is set
22:31 mwhahaha we will set it.
22:31 owalsh we can only have 1 cell using the defualt nova db & transport url, if simple_cell_setup created it then we can't
22:32 openstackgerrit Merged openstack/puppet-openstack-integration: logs: add keystone-resources for debug  https://review.openstack.org/421980
22:32 mwhahaha why would you not create the name in simple_cell_setup
22:32 mwhahaha anyway, so assume no simple_cell_setup
22:33 mwhahaha cause we've turned that off, i'm reverting the change for newton so let's assume it's not been done
22:33 owalsh shardy is doing it in the tripleo upgade
22:34 mwhahaha that code has not been written yet
22:34 mwhahaha so we shouldn't do that
22:34 mwhahaha specifically because of this
22:34 mwhahaha so we need to pick the direction we go with this
22:34 mwhahaha brb gotta pick up my kid from the bus. back in like 15. but the point is we need to do the upgrade and this consitently
22:35 mwhahaha if we're doing simple cell setup for the default, then we need to do it for our case as well
22:35 mwhahaha but i think it better to do it the non simple fashion for both
22:35 owalsh https://review.openstack.org/#/c/4​05241/10/puppet/services/nova-api.yaml@227
22:36 agrebennikov_ joined #puppet-openstack
22:36 mwhahaha that's still in progress so that needs to be updated to be consistent with what you're proposing in puppet
22:36 mwhahaha because once the upgrade is done, puppet takes over
22:36 mwhahaha so they must match
22:50 mwhahaha owalsh: ok so. i guess the thing is we should implement simple setup (optional) and the ability to manually create cells
22:51 owalsh ok... simple setup is what this is doing, actually if I drop the cell1 name it should be consistent
22:51 mwhahaha well no
22:51 owalsh we just don't require any hosts to map
22:51 mwhahaha i mean we should actually runs imple setup
22:51 mwhahaha or not
22:52 mwhahaha like we need to implement both ways
22:52 mwhahaha so my thought would be implement a toggle for simple setup in puppet-nova and implement the manual actions
22:52 owalsh but we can't run simple setup without hosts
22:53 mwhahaha why is everything always followed with a 'but it doesn't work like that because of X'
22:54 mwhahaha this entire process is ridiculous
22:54 owalsh this *just* implement the simple setup steps without the host/instance mapping
22:55 owalsh create_cell without a transport_url/database is more like a db sync
22:55 mwhahaha yea sorry, i'm just really frustrated at this entire thing because it's totally not user friendly
22:56 mwhahaha like if you could pick the worst way to handle things and implemented it, that's this
22:58 owalsh I've made that point... having to reverse the order of the db syncs for upgrade vs fresh install is horrible for config mgmt tools
22:58 dmsimard ಠ_ಠ
22:58 dmsimard mwhahaha: I don't have a stdin pipe to my brain yet
22:58 dmsimard (I wish)
22:58 mwhahaha dmsimard: need tiny container v2
22:59 mwhahaha owalsh: so basically you're trying to implement http://docs.openstack.org/develope​r/nova/cells.html#first-time-setup right?
23:00 owalsh mwhahaha: yes
23:00 mwhahaha owalsh: ok so, i would say that we should follow the same procedure in the upgrade process as well
23:01 mwhahaha so we need a map_cell0, create a cell and then a discover hosts
23:01 owalsh well... we do by running simple setup
23:01 owalsh no, discover_hosts is for newly deploy compute hosts
23:01 mwhahaha right i'm talking overall process
23:01 owalsh for upgrade we need map_hosts_and_cells + map_instances
23:01 mwhahaha oh jesus
23:02 mwhahaha so no more talk of simple_cell_setup, it doesn't exist as far as i'm concerned
23:03 owalsh well, for upgrades it does all this for us so why not let it?
23:03 mwhahaha because it is neither simple nor a setup
23:03 mwhahaha because you can't use both together
23:03 xarses_ joined #puppet-openstack
23:03 mwhahaha because of optional name
23:03 owalsh hence why I'm using exit code...
23:03 openstackgerrit Merged openstack/puppet-openstack-integration: Sort the output of rpm -qa for easier comparisons  https://review.openstack.org/421366
23:04 mwhahaha owalsh: which hides failures and is not acceptable
23:04 owalsh no, exit code 2 is specific
23:04 owalsh exit code 1 would be failures
23:05 owalsh https://github.com/openstack/nova/b​lob/master/nova/cmd/manage.py#L1428
23:05 openstackgerrit David Moreau Simard proposed openstack/puppet-openstack-integration: Sort the output of rpm -qa for easier comparisons  https://review.openstack.org/422301
23:05 openstackgerrit David Moreau Simard proposed openstack/puppet-openstack-integration: logs: add keystone-resources for debug  https://review.openstack.org/422302
23:07 mwhahaha ok fine don't use unless, use the exit code but move the cell create to it's own resource
23:07 owalsh but grr, simple_cell_setup does hide failures as it returns 1 when there are no hosts or if *anything else goes wrong*
23:07 mwhahaha but you can't because that's not idempotent which is why that's a problem
23:07 owalsh not idempotent in what way?
23:08 mwhahaha i guess it is if you leave it refresh only
23:08 mwhahaha the whole point of the unless/onlyif is to know when not to run it
23:08 mwhahaha which is why having name be optional is a poor choice
23:09 openstackgerrit Ade Lee proposed openstack/puppet-nova: Set correct ownership for nova keytab  https://review.openstack.org/422200
23:10 acormier joined #puppet-openstack
23:13 owalsh point is we can only create one cell using the default db/transport... so we could add an unless/only_if but the easier check here might be to count the cells
23:14 mwhahaha let me pull down this patch and think on it
23:15 mwhahaha owalsh: the problem i'm having with this entire thing is being able to  continue to manage and scale the cloud
23:15 mwhahaha owalsh: much of the assumptions being made break down if someone was to come along and want to say decom the original cell or create more
23:16 mwhahaha owalsh: so i'm trying to figure out a way that scales because the use case is beyond just tripleo which may only have one cell
23:16 owalsh mwhahaha: that all a WIP I believe, I don't think cell_v2 is landing in Ocata, just some pieces
23:16 mwhahaha owalsh: then it should be optional
23:16 mwhahaha so we don't have this problem
23:17 mwhahaha that's another one of my frustrations around this entire thing that it's being make mandatory without quality thought going into how someone would actually useit
23:17 owalsh .placement api was optional in Newton, but not... I think this is happening now for a reason
23:18 mwhahaha i'd like to know the reasons :/ anyway
23:21 mwhahaha owalsh: what was the specific reasoning behind flipping the api/db sync?
23:23 owalsh map_cell0/create_cell modify the api db so it should be synced already
23:24 owalsh and another syncdb is required after create_cell, so instead of running it twice just run after
23:24 mwhahaha k
23:24 owalsh (sigh except for upgrades
23:24 * mwhahaha shares his pain with owalsh
23:24 furlongm_ joined #puppet-openstack
23:24 acabot joined #puppet-openstack
23:24 stevemar joined #puppet-openstack
23:28 owalsh mwhahaha: while I'm stealing your time... how to deal with discover_hosts?
23:28 mwhahaha i was just thinking about that
23:28 mwhahaha we can't really
23:28 mwhahaha because it requires cross node coordination
23:28 mwhahaha that's not something we traditionally handle in puppet-nova
23:28 mwhahaha so i was writing a class that could be included but i'm not sure it's usefulness
23:29 owalsh already have one :-)
23:29 mwhahaha well that should be included in this setup thing
23:30 owalsh I guess we could have deployed the computes beforehand
23:30 mwhahaha owalsh: no we traditionally do not
23:30 owalsh then there is nothing to discover
23:30 mwhahaha owalsh: what does discover_hosts actually return?
23:31 mwhahaha because we could cheat and always run it but do it via unless
23:31 owalsh it adds hosts to the cell, after they've been deployed
23:31 mwhahaha http://paste.openstack.org/show/595489/
23:33 mwhahaha owalsh: no i mean what return codes are there
23:33 owalsh none
23:33 owalsh well 0 or something terrible happened
23:35 owalsh if we never have computes deployed at this stage then it's a noop
23:35 mwhahaha well basically that would run it whenever puppet is run but not show up as non-idempotent unless it failed
23:36 mwhahaha it's cheating the system :D
23:36 mwhahaha cause it'll techincally run all the time but not show up as that
23:39 owalsh hmm, so we care because puppet would report that changes were applied?
23:39 mwhahaha yea
23:39 mwhahaha so in order to be idempotent you have to know when to run and not run something
23:40 mwhahaha rather than blindly run every time
23:40 mwhahaha technically ansible also has a similar concept for many of it's modules, but i'm not sure if people are really aware of this concept
23:42 owalsh ah... so now it make sense why you're suggesting list_cells & grep... I'm quite new, and my puppet is a bit rusty... I'll get there eventually :-)
23:43 mwhahaha providers also use a similar concept but you can do  more complex things in providers than just execs
23:43 mwhahaha so for something like cells, a provider is a better choice because then you can pull down all the stuff and do that complex comparison
23:43 owalsh yea, I remember providers :-)
23:44 mwhahaha it's all about current state vs expected state :)
23:44 mwhahaha anyway
23:45 owalsh yea, I've been knee deep in ansible/chef/salt for a while but it all began with puppet
23:47 owalsh ok, it's getting pretty late here and I've got to catch a train to the Dublin office in (arrrrrh) six hours...
23:47 mwhahaha owalsh: ok i'll propose something that you can review tomorrow
23:47 mwhahaha owalsh: you don't happen to have a list_cells output somewhere do you?
23:48 owalsh mwhahaha: sure...
23:50 owalsh mwhahaha: http://paste.openstack.org/show/595492/
23:51 mwhahaha owalsh: cool thanks
23:53 owalsh mwhahaha: big thanks for helping with this. I'll catch up with you tomorrow. Goodnight/evening/afternoon
23:53 mwhahaha owalsh: sounds good

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