Perl 6 - the future is here, just unevenly distributed

IRC log for #puppet-openstack, 2015-08-18

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

All times shown according to UTC.

Time Nick Message
00:07 ducttape_ joined #puppet-openstack
00:12 xarses joined #puppet-openstack
00:15 sergmelikyan joined #puppet-openstack
00:15 gildub crinkle, Hi, yesterday we talked about cherry picking across branches. What about if there are 2 changes (linked) that I would like to carry over? Is picking the second one is going to bring the first one in, I'd assume yes but the commit message is only the last one so I'm not sure. Or do I have to pick both separately, but how to link them then?
00:16 crinkle gildub: I'm not entirely sure how it works, I don't think cherrypicking just the second one will work
00:17 gildub crinkle, yeah. Ok I'll grab them separately then, I should be able to link them afterwards. Thanks
00:17 openstackgerrit Gilles Dubreuil proposed openstack/puppet-keystone: Purely cosmetic change on keystone provider  https://review.openstack.org/213938
00:19 gildub crinkle, well, I can't cherry pick the second one anyway, I get an error
00:20 crinkle gildub: yeah i think since it depends on the first one it can't be cherry-picked cleanly on its own
00:20 crinkle gildub: would have to wait for the first one to merge
00:21 crinkle gildub: i'm not sure if there's a better mechanism in gerrit to do that
00:21 gildub crinkle, that's fine, it makes sense
00:21 aimon joined #puppet-openstack
00:22 gildub crinkle, but the first one has beaker/upgrade issues, might be related to the liberty branch. The first one is only comestic so I shouldn't break anything
00:28 gildub EmilienM, ^
00:38 tiswanso joined #puppet-openstack
00:40 ducttape_ joined #puppet-openstack
00:41 tiswanso joined #puppet-openstack
00:41 ducttape_ joined #puppet-openstack
00:48 sergmelikyan joined #puppet-openstack
01:02 sergmelikyan joined #puppet-openstack
01:09 vinsh I found my bug crinkle and EmilienM.  When I evaluate "$service_name = split($name, 'swift-')" on centos.. it's different then ubuntu.
01:09 vinsh in centos it yields "[, object-server]"
01:09 zacdev joined #puppet-openstack
01:09 vinsh when in ubuntu its just "object-server"
01:09 vinsh I setup a centos vm and ran the beaker test locally.  Big help :)
01:10 ducttape_ joined #puppet-openstack
01:19 openstackgerrit Adam Vinsh proposed openstack/puppet-swift: WIP: Introduce swift_service class  https://review.openstack.org/203220
01:28 vinsh crinkle: EmilienM: I VOTE that we add --debug ontop of those puppet beaker runs... it's a lot of output.. but it would be a major help in debugging.
01:28 vinsh Maybe there is a reason we didn't?  I solved my problem by running it locally with debug.. I would have never seen the malformed string interpolation otherwise.
01:35 openstackgerrit Gilles Dubreuil proposed openstack/puppet-keystone: Endpoints are APIs version-less  https://review.openstack.org/213957
01:38 imcsk8 joined #puppet-openstack
01:38 crinkle vinsh: I think it just makes a ton of output, might make the log a lot harder to read
01:38 crinkle but I know this isn't the first time --debug has come in handy for that
01:39 openstackgerrit Gilles Dubreuil proposed openstack/puppet-keystone: Endpoints without API version and refactored auth  https://review.openstack.org/212523
01:46 vinsh crinkle: I suppose maybe the user could add --debug to their commit somehow to debug if needed.  Dunno.
03:33 openstackgerrit Adam Vinsh proposed openstack/puppet-swift: WIP: Introduce swift_service class  https://review.openstack.org/203220
03:34 logan2 joined #puppet-openstack
03:47 xingchao joined #puppet-openstack
03:47 fedexo joined #puppet-openstack
03:48 vinsh nailed it.
04:04 openstackgerrit Gilles Dubreuil proposed openstack/puppet-keystone: Authentication URLs and endpoint clarity re-factor  https://review.openstack.org/213603
04:29 openstackgerrit Merged openstack/puppet-tuskar: Rely on autorequire for config resource ordering  https://review.openstack.org/211917
04:33 openstackgerrit Gilles Dubreuil proposed openstack/puppet-keystone: Resource keystone_trust for Keystone V3 API  https://review.openstack.org/200996
04:37 openstackgerrit Merged openstack/puppet-swift: Rely on autorequire for config resource ordering  https://review.openstack.org/211929
04:42 markvoelker joined #puppet-openstack
05:06 _nick joined #puppet-openstack
05:15 xingchao joined #puppet-openstack
05:40 LimorStotland joined #puppet-openstack
05:40 LimorStotland Hi all
05:41 LimorStotland I need just one more +2 to finish with the puppet-mistral. can anyone form the core team can help? pleas
05:46 aimon joined #puppet-openstack
05:47 LimorStotland https://review.openstack.org/#/c/208457/ if you want to review :-)
06:03 sergmelikyan joined #puppet-openstack
06:05 serg_melikyan joined #puppet-openstack
06:13 nibalizer glxgears
06:13 nibalizer whoops
06:26 gildub joined #puppet-openstack
06:28 jtomasek joined #puppet-openstack
06:33 openstackgerrit Denis Egorenko proposed openstack/puppet-sahara: Update parameters for Sahara  https://review.openstack.org/205988
06:43 openstackgerrit_ joined #puppet-openstack
06:43 markvoelker joined #puppet-openstack
06:49 openstackgerrit Yanis Guenane proposed openstack/puppet-manila: Rely on autorequire for config resource ordering  https://review.openstack.org/211902
06:57 rochaporto joined #puppet-openstack
07:01 dgurtner joined #puppet-openstack
07:01 dgurtner joined #puppet-openstack
07:08 openstackgerrit Gilles Dubreuil proposed openstack/puppet-keystone: Authentication URLs and endpoint clarity re-factor  https://review.openstack.org/213603
07:09 openstackgerrit Gilles Dubreuil proposed openstack/puppet-keystone: Endpoints are APIs version-less  https://review.openstack.org/213957
07:22 fedexo joined #puppet-openstack
07:40 michchap LimorStotland: are there no packages so we have to have the systemd unit files in puppet?
07:40 openstackgerrit Yanis Guenane proposed openstack/puppet-openstacklib: Allow the use of an ensure_absent_val param  https://review.openstack.org/202574
07:42 LimorStotland I am not sure i understand your question (I am new at puppet)....there aren't any RPMs (or what ever) for mistral
07:43 michchap LimorStotland: yep that was the question, I see in init.pp there is a package {'mistral'} and I would expect the unit files to come from there, not the module
07:44 michchap LimorStotland: in fact won't this module fail because that package doesn't exist in rdo/uca?
07:44 LimorStotland michchap  : I created for myself an rpm for installing the mistral
07:45 michchap LimorStotland: are you at redhat or do you know if they have plans to release a mistral rpm in the near future?
07:45 LimorStotland I am not from redhat and as far as i know thy will not support mistral in L
07:46 LimorStotland can we use the rpm i created for the testing?
07:46 michchap LimorStotland: I ask because it's a little odd to have a module that can't be used without building your own rpms. At the very least we should state as such in the readme. I think most users would expect the module to work with RDO at least
07:46 michchap LimorStotland: do you have a repo with the rpmspec or instructions for building?
07:47 openstackgerrit Yanis Guenane proposed openstack/puppet-openstacklib: Allow the use of an ensure_absent_val param  https://review.openstack.org/202574
07:47 openstackgerrit Yanis Guenane proposed openstack/puppet-openstacklib: Enable acceptance tests for openstack_config  https://review.openstack.org/213060
07:47 michchap LimorStotland: I think you should link to that from the readme so users can try the module - otherwise there is a big assumption that they have already built an appropriate mistral package.
07:48 LimorStotland I have an instructions for building the rpm (i build it using python ./setup.py bdist_rpm) and i can add to the readme
07:48 michchap LimorStotland: that would be great. thanks. I will +2 your initial patch and can you please add that as another patch?
07:49 LimorStotland sure Thanks
07:54 openstackgerrit Merged openstack/puppet-mistral: puppet-mistral: initial commit  https://review.openstack.org/208457
07:55 paramite joined #puppet-openstack
07:55 jistr joined #puppet-openstack
07:55 linkedinyou joined #puppet-openstack
07:58 social joined #puppet-openstack
08:01 gfidente joined #puppet-openstack
08:10 openstackgerrit Merged openstack/puppet-horizon: Make redirect type configurable  https://review.openstack.org/208462
08:12 derekh joined #puppet-openstack
08:13 derekh joined #puppet-openstack
08:20 openstackgerrit Merged openstack/puppet-neutron: Update Cisco ML2 classes and templates  https://review.openstack.org/197181
08:25 IBerezovskiy joined #puppet-openstack
08:25 openstackgerrit Ivan Berezovskiy proposed openstack/puppet-nova: Add nameservers (dns) parameters  https://review.openstack.org/202788
08:27 openstackgerrit Limor Stotland proposed openstack/puppet-mistral: Adding instruction for building RPM  https://review.openstack.org/214069
08:28 LimorStotland michchap I add to the README can you see if this it what you meant https://review.openstack.org/#/c/214069/?Thanks
08:28 michchap LimorStotland: thanks ;)
08:33 openstackgerrit Merged openstack/puppet-trove: Add api ratelimit options  https://review.openstack.org/200397
08:34 openstackgerrit Javier Peña proposed openstack/puppet-swift: Add support for swift-object-expirer service  https://review.openstack.org/211549
08:45 markvoelker joined #puppet-openstack
08:55 aimon_ joined #puppet-openstack
08:58 d0ugal joined #puppet-openstack
09:01 dgurtner joined #puppet-openstack
09:02 saneax joined #puppet-openstack
09:05 sh1kel joined #puppet-openstack
09:08 IBerezovskiy hello, everyone. Can you please review small change for nova nameservers https://review.openstack.org/#/c/202788/ ?
09:19 fvollero joined #puppet-openstack
09:20 fvollero joined #puppet-openstack
09:42 openstackgerrit David Gurtner proposed stackforge/puppet-ceph: Be more specific with CentOS7 package exclude  https://review.openstack.org/214093
09:50 openstackgerrit David Gurtner proposed stackforge/puppet-ceph: use full length APT key  https://review.openstack.org/209981
10:01 openstackgerrit David Gurtner proposed stackforge/puppet-ceph: Drop no longer supported releases  https://review.openstack.org/214104
10:31 saneax joined #puppet-openstack
10:38 openstackgerrit Michael Polenchuk proposed openstack/puppet-nova: Activate fping api extension  https://review.openstack.org/214132
10:45 markvoelker joined #puppet-openstack
10:47 openstackgerrit David Gurtner proposed stackforge/puppet-ceph: Fix RGW Apache syslog creation  https://review.openstack.org/214135
10:50 saneax joined #puppet-openstack
11:01 openstackgerrit Sebastien Badia proposed openstack/puppet-neutron: ml2: Fix typo with ml2_srvio/supported_pci_vendor_devs param.  https://review.openstack.org/214138
11:04 openstackgerrit David Gurtner proposed stackforge/puppet-ceph: Fix RGW Apache syslog creation  https://review.openstack.org/214135
11:05 openstackgerrit Sebastien Badia proposed openstack/puppet-neutron: ml2: Fix typo with ml2_srvio/supported_pci_vendor_devs param.  https://review.openstack.org/214141
11:32 cdelatte joined #puppet-openstack
11:40 bogdando joined #puppet-openstack
11:41 saneax joined #puppet-openstack
11:46 markvoelker joined #puppet-openstack
11:51 openstackgerrit Giulio Fidente proposed openstack/puppet-tripleo: Add generic local_mac_map function and ceph_osds class  https://review.openstack.org/211935
11:51 sergmelikyan joined #puppet-openstack
12:02 vinsh joined #puppet-openstack
12:04 dprince joined #puppet-openstack
12:07 dprince joined #puppet-openstack
12:08 EmilienM good morning
12:10 openstackgerrit Sebastien Badia proposed openstack/puppet-neutron: lint: Add and fix missing parameter documentation  https://review.openstack.org/210059
12:10 openstackgerrit Sebastien Badia proposed openstack/puppet-neutron: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/209122
12:12 markvoelker joined #puppet-openstack
12:19 EmilienM vinsh: re- logs/swift: yes I thought about that when I did a serie of patches to add debug in beaker jobs - afik puppet-swift does not support logs yet?
12:19 ducttape_ joined #puppet-openstack
12:20 iurygregory good morning
12:27 gildub joined #puppet-openstack
12:27 zacdev joined #puppet-openstack
12:30 mjblack joined #puppet-openstack
12:31 openstackgerrit Sebastien Badia proposed openstack/puppet-openstack_extras: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189876
12:31 openstackgerrit Sebastien Badia proposed openstack/puppet-nova: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189874
12:31 openstackgerrit Sebastien Badia proposed openstack/puppet-vswitch: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189883
12:31 openstackgerrit Sebastien Badia proposed openstack/puppet-tripleo: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189880
12:34 openstackgerrit Sebastien Badia proposed openstack/puppet-openstacklib: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189875
12:36 openstackgerrit Sebastien Badia proposed openstack/puppet-tuskar: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189882
12:37 openstackgerrit Sebastien Badia proposed openstack/puppet-heat: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189868
12:38 ducttape_ joined #puppet-openstack
12:39 openstackgerrit Sebastien Badia proposed openstack/puppet-glance: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189865
12:40 dgurtner_ joined #puppet-openstack
12:40 openstackgerrit Sebastien Badia proposed openstack/puppet-trove: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189881
12:41 mjblack joined #puppet-openstack
12:41 openstackgerrit Sebastien Badia proposed openstack/puppet-sahara: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189877
12:42 vinsh Hey there EmilienM.  Which logs do you mean? for puppet-swift ?
12:42 EmilienM vinsh: yeah, when I enabled logs for beaker jobs, I skipped swift because I did not know (for now) how to enable logs
12:42 vinsh I was asking about the actual puppet beaker run.  Lack of debug there makes it readable.. but not debugable ;)
12:43 vinsh So other puppet-modules are collecting openstack logs after beaker runs?
12:43 openstackgerrit Sergey Kolekonov proposed openstack/puppet-neutron: Add an ability to set up drop_flows_on_start option  https://review.openstack.org/211986
12:44 EmilienM vinsh: infra script does
12:44 vinsh Can you point me at an example for another service? I will work it out for swift.
12:45 EmilienM vinsh: https://github.com/openstack-infra/project-config/blob/master/jenkins/scripts/copy_puppet_logs.sh
12:45 vinsh Thank you.  I'll ping back when I have something. :)
12:45 EmilienM vinsh: https://github.com/openstack/puppet-nova/blob/master/manifests/init.pp#L166-L173
12:45 dgurtner joined #puppet-openstack
12:45 dgurtner joined #puppet-openstack
12:46 vinsh Perfect, this will work.
12:46 EmilienM vinsh: http://logs.openstack.org/12/213512/1/check/gate-puppet-trove-puppet-beaker-rspec-dsvm-centos7/dcdf1c5/logs/
12:46 EmilienM http://logs.openstack.org/12/213512/1/check/gate-puppet-trove-puppet-beaker-rspec-dsvm-centos7/dcdf1c5/logs/keystone/keystone.txt.gz
12:46 EmilienM vinsh: thx a lot
12:47 vinsh I wonder if we can log puppet debug output to log.. and normal to console
12:47 EmilienM vinsh: are we talking about swift logs or puppet logs?
12:47 vinsh Both right now.  My initial question yesterday was about puppet logs
12:47 vinsh but I am happy to enable swift logs also
12:47 EmilienM well, beaker jobs already have logs
12:48 vinsh just not with debug output... which makes failures hard to figure out unless you setup the respective vm locally to test
12:48 EmilienM while swift nothing
12:48 vinsh which I did for centos
12:48 openstackgerrit Sebastien Badia proposed openstack/puppet-manila: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189872
12:49 vinsh Yesterday you two had mentioned it looked like a service had failed to start.. and that was why my 3rd beaker puppet run found changes.. and posted failure.
12:49 vinsh I had to setup beaker on a centos vm and run puppet in debug there to even figure out the command puppet was running was invalid.
12:49 EmilienM spredzy: dropped a question on https://review.openstack.org/#/c/213060/
12:49 vinsh Which was a good and useful learning experience in the end :)
12:51 openstackgerrit Sebastien Badia proposed openstack/puppet-gnocchi: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189866
12:57 spredzy EmilienM, looking at https://review.openstack.org/#/c/213060/6/spec/acceptance/openstacklib_config_provider_spec.rb,cm it seems that is what I did first
12:58 EmilienM looking
12:58 spredzy I remember an issue with keystone trying to connect to mysql hence I did what I did but I can't refind it in the log ...
12:58 spredzy I prob. got confused
12:58 EmilienM spredzy: no, even without ::keystone
12:58 EmilienM is it possible to just use keystone_config?
12:58 EmilienM or do we have scope issue
12:59 spredzy EmilienM, not possible we need the package to be installed of the directory  to be created so the _config provider can work
12:59 spredzy the provider can create the file but not the directory it has to be in
12:59 EmilienM spredzy: yeah the mysql error you got is because you use mysql :)
13:00 EmilienM you should use sqlite
13:00 EmilienM spredzy: I suggest managing the directory manually
13:00 EmilienM and drop ::keystone
13:00 EmilienM it's way more lightweight
13:00 EmilienM spredzy: what do you think?
13:00 spredzy EmilienM, last week you posted a review telling me you didnt want that
13:00 spredzy but to use the package instead
13:01 EmilienM spredzy: sorry for that, I realize now we probably add useless tests here
13:01 spredzy :(
13:01 spredzy let me push it back again
13:01 EmilienM spredzy: wait
13:01 EmilienM spredzy: what do you think?
13:01 EmilienM the reason why I don't like having all these classes is because it's adding dependencies and complexity while we just want to test a configuration file
13:02 spredzy EmilienM, https://review.openstack.org/#/c/209882/4/spec/acceptance/heat_config_spec.rb,cm
13:02 spredzy so I should stick to the way it is there right ?
13:02 EmilienM I know and I was wrong
13:02 EmilienM yes
13:02 spredzy ok
13:02 EmilienM this is actually the right path I think
13:03 spredzy it is the simpler path in fact as it just test the provider logic and nothing else
13:03 EmilienM spredzy: even better, add a comment to explain why we do that
13:03 spredzy and that is what it was intended for
13:03 EmilienM spredzy: cool - thanks for this wor
13:03 EmilienM work*
13:04 EmilienM iurygregory: ping - re federation
13:04 iurygregory hi EmilienM
13:04 EmilienM iurygregory: I was reading the BP
13:05 EmilienM iurygregory: I have a few questions, maybe we can discuss here
13:05 iurygregory ok
13:05 xingchao joined #puppet-openstack
13:05 EmilienM iurygregory: first of all, is keystone wsgi required to run federation?
13:05 openstackgerrit Merged openstack/puppet-trove: acceptance: enable debug & verbosity for OpenStack logs  https://review.openstack.org/213512
13:06 iurygregory yes ^^
13:06 iurygregory we'll consider just using apache
13:06 EmilienM iurygregory: ok so please add it in the blueprint - this is an important information I suppose
13:07 openstackgerrit Merged openstack/puppet-neutron: acceptance: enable debug & verbosity for OpenStack logs  https://review.openstack.org/213510
13:07 EmilienM oh it's mentioned aready
13:07 EmilienM iurygregory: my second question is about packages
13:08 openstackgerrit Merged openstack/puppet-ironic: acceptance: enable debug & verbosity for OpenStack logs  https://review.openstack.org/213508
13:08 openstackgerrit Merged openstack/puppet-ironic: acceptance: install openstack-selinux on redhat plateforms  https://review.openstack.org/213348
13:08 iurygregory ok  go ahead
13:08 EmilienM when I configure keystone for federation but IDP is external (on a separate node) - do I still need extra packages?
13:09 iurygregory you mean IDP as "Google" for example? or other Keystone as IDP?
13:09 EmilienM iurygregory: yeah, like Google
13:09 EmilienM the actual OpenID server for example
13:10 EmilienM not IDP, but SP
13:10 openstackgerrit Sebastien Badia proposed openstack/puppet-swift: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189878
13:10 iurygregory yes o need to install the required packages to make it work, cause the packages aren't installed when you do "apt-get keystone"
13:11 EmilienM iurygregory: ok but my question is - do I need the packages even though the openid server is on a separate node, right?
13:11 openstackgerrit Giulio Fidente proposed openstack/puppet-tripleo: Add generic local_mac_map function and ceph_osds class  https://review.openstack.org/211935
13:11 EmilienM like libapache2-auth-openidc
13:12 openstackgerrit Sebastien Badia proposed openstack/puppet-swift: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189878
13:12 iurygregory wait a sec, you mean you have a Keystone in "node2" and "node1" have the openid?
13:13 EmilienM iurygregory: yes
13:14 iurygregory you need to install in the "node2" the openid cause keystone use enviroment variables when doing federated login
13:15 EmilienM iurygregory: sorry, it's not clear to me.
13:16 EmilienM Are we going to taking care of packages dependencies to install the SP backend?
13:16 EmilienM or just to make Keystone working
13:16 iurygregory the first option in my opinion
13:17 EmilienM I don't think this is the purpose of puppet-keystone
13:18 iurygregory you think we only need to add the settings related to the keystone?
13:19 iurygregory if we need to install and change the configuration for "openid" "saml" other module should do that?
13:20 dfisher joined #puppet-openstack
13:23 tiswanso joined #puppet-openstack
13:24 EmilienM iurygregory: puppet-keystone installs & configure keystone service
13:24 EmilienM that means: managing /etc/keystone/*, managing & running keystone service
13:24 EmilienM except for WSGI run, we manage keystone to work in apache by configuring apache vhost
13:25 EmilienM example: we use MySQL backend to store stuffs - we don't actually configure MySQL
13:25 EmilienM so if Federation aims to delegate auth to something else, we don't configure the something else, we just prepare keystone to work with the something else
13:25 EmilienM if someone is deploying openid as a service provider, good, we need to use puppet-openid to deploy openid
13:26 EmilienM (if there is)
13:26 EmilienM iurygregory: does that makes sense?
13:26 openstackgerrit Merged openstack/puppet-ceilometer: wsgi: fix typos in comments  https://review.openstack.org/213207
13:27 iurygregory yes EmilienM, the configuration for federation that resides in the openid or saml should be in the specific puppet module.
13:27 EmilienM iurygregory: this is really important to make the distinction
13:28 EmilienM iurygregory: because puppet modules should have a scope limit
13:28 iurygregory make sense EmilienM, i'll update and send other patch
13:30 sergmelikyan joined #puppet-openstack
13:32 saneax joined #puppet-openstack
13:32 saikrishna joined #puppet-openstack
13:33 iurygregory anything else EmilienM ?
13:33 EmilienM iurygregory: yes
13:33 EmilienM iurygregory: I'm finisihing the review
13:33 iurygregory ok ^^
13:34 EmilienM iurygregory: done
13:34 iurygregory thanks EmilienM
13:34 EmilienM iurygregory: please let me know if something is not clear - we can discuss here anytime
13:34 iurygregory ok
13:39 prad_ joined #puppet-openstack
13:39 openstackgerrit Sebastien Badia proposed openstack/puppet-tempest: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189879
13:40 ducttape_ joined #puppet-openstack
13:46 openstackgerrit Emilien Macchi proposed openstack/puppet-ironic: acceptance: bump to Liberty release  https://review.openstack.org/209292
13:47 openstackgerrit Sebastien Badia proposed openstack/puppet-swift: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189878
13:49 openstackgerrit Emilien Macchi proposed openstack/puppet-heat: acceptance: bump to Liberty release  https://review.openstack.org/209290
13:51 dfisher EmilienM: can I ask you to look at a review and help me decide what to do with a comment?
13:51 openstackgerrit Emilien Macchi proposed openstack/puppet-ceilometer: acceptance: enable debug & verbosity for OpenStack logs  https://review.openstack.org/213504
13:51 openstackgerrit Yanis Guenane proposed openstack/puppet-openstacklib: Enable acceptance tests for openstack_config  https://review.openstack.org/213060
13:51 EmilienM dfisher: shoot
13:51 dfisher https://review.openstack.org/#/c/211593/
13:51 dfisher in patch set 1
13:52 spredzy EmilienM, ^ (earlier conversation)
13:52 dfisher I'm trying to figure out if I should put some of the 'file' entries back in for glance-registry
13:52 EmilienM spredzy: ack
13:52 dfisher for db_sync
13:53 EmilienM dfisher: oh I see the problem
13:53 EmilienM IMHO we should not run db_sync at each File change
13:53 EmilienM rather at each package upgrade
13:54 dfisher i would agree with you.
13:54 EmilienM dfisher: replied
13:55 EmilienM spredzy: oh, I asked if we could put a comment to explain why we do that
13:55 openstackgerrit Giulio Fidente proposed openstack/puppet-tripleo: Add generic local_mac_map function and ceph_osds class  https://review.openstack.org/211935
13:55 dfisher ok, so at this point, I think the review should be in good shape then.
13:55 dfisher if you want to look at it, I would appreciate it.
13:55 EmilienM dfisher: yeah
13:55 EmilienM dfisher: after my daily scrum
13:55 * dfisher hattip
13:55 dfisher take your time :)
14:00 sergmelikyan joined #puppet-openstack
14:01 tiswanso joined #puppet-openstack
14:02 tchaypo joined #puppet-openstack
14:02 xingchao joined #puppet-openstack
14:10 openstackgerrit Yanis Guenane proposed openstack/puppet-openstacklib: Enable acceptance tests for openstack_config  https://review.openstack.org/213060
14:10 spredzy EmilienM, forgot about that. here ^
14:13 openstackgerrit Emilien Macchi proposed openstack/puppet-tempest: do not create tempest.conf from .sample file  https://review.openstack.org/210784
14:23 EmilienM spredzy: logs are very interesting
14:23 EmilienM http://logs.openstack.org/04/213504/3/check/gate-puppet-ceilometer-puppet-beaker-rspec-dsvm-centos7/e2989e5/logs/ceilometer/collector.txt.gz#_2015-08-18_14_07_28_829
14:24 EmilienM this kind of feedback will help us to update parameters
14:25 spredzy EmilienM, indeed
14:25 spredzy neat
14:29 EmilienM spredzy: look the manila issue: http://logs.openstack.org/09/213509/2/check/gate-puppet-manila-puppet-beaker-rspec-dsvm-centos7/79a771c/logs/manila/api.txt.gz#_2015-08-18_14_03_59_005
14:30 EmilienM spredzy: so I was wrong when I tried to fix orchestration
14:30 EmilienM it's really packagin ghere
14:34 spredzy :(
14:39 EmilienM clayton, mfisch, mgagne, sbadia: any core around can review https://review.openstack.org/#/c/209294/ ?
14:39 saneax joined #puppet-openstack
14:42 saikrishna joined #puppet-openstack
14:47 EmilienM gentle reminder: our weekly meeting will start in 15 min on #openstack-meeting-4
14:53 angdraug joined #puppet-openstack
14:57 iurygregory sorry EmilienM =X if the configuration for "openid" is set in the keystone virtual host file we can use the puppet-keystone right
14:57 jpena_ joined #puppet-openstack
14:58 michchap_ joined #puppet-openstack
14:58 EmilienM iurygregory: right !
14:58 iurygregory ok o/
15:00 bogdando joined #puppet-openstack
15:00 zino joined #puppet-openstack
15:02 skolekonov joined #puppet-openstack
15:03 fvollero joined #puppet-openstack
15:12 dprince joined #puppet-openstack
15:12 jtomasek joined #puppet-openstack
15:12 finchd joined #puppet-openstack
15:12 myatsenko joined #puppet-openstack
15:12 colby_ joined #puppet-openstack
15:13 saikrishna joined #puppet-openstack
15:34 openstackgerrit Adam Vinsh proposed openstack/puppet-swift: WIP: Introduce swift_service class  https://review.openstack.org/203220
15:38 openstackgerrit Merged openstack/puppet-nova: Add nameservers (dns) parameters  https://review.openstack.org/202788
15:40 openstackgerrit Merged openstack/puppet-neutron: Allow deployment when DVR is enabled and L2 pop and tunneling are disabled  https://review.openstack.org/202542
15:43 openstackgerrit Merged openstack/puppet-openstacklib: Allow the use of an ensure_absent_val param  https://review.openstack.org/202574
15:43 openstackgerrit Merged openstack/puppet-ceilometer: WSGI: make it work, and test it with acceptance  https://review.openstack.org/213146
15:43 openstackgerrit Merged openstack/puppet-tempest: do not create tempest.conf from .sample file  https://review.openstack.org/210784
15:44 openstackgerrit Sebastien Badia proposed openstack/puppet-tempest: Initial msync run for all Puppet OpenStack modules  https://review.openstack.org/189879
15:52 rgowrishankar joined #puppet-openstack
15:56 sergmelikyan joined #puppet-openstack
16:02 rgowrishankar Hello, long time lurker, trying to be first time committer.. I work for Nuage Networks and we have puppet configuration for the neutron plugin and puppet configuration for nuage agents. The files will go under puppet-neutron/manifests/plugins/ and puppet-neutron/manifests/agents. It will follow the existing patterns of vendor code (new folder named nuage
16:02 rgowrishankar similar to the cisco, nvp folders). Do I need a blue print for this or can I create a gerrit review directly or do I need to start a thread on the mailing list?
16:06 EmilienM rgowrishankar: https://github.com/openstack/puppet-neutron/commit/183541d3890cd212db5a581470bdf98c21c40ad7 or https://github.com/openstack/puppet-neutron/commit/4ff68894b3ad537f52c5ec860e8d1a5c7549d19b are good examples
16:07 EmilienM rgowrishankar: is nuage ml2 friendly?
16:11 tchaypo joined #puppet-openstack
16:33 ilbot3 joined #puppet-openstack
16:33 Topic for #puppet-openstack is now Puppet/OpenStack: Weekly meeting (every Tuesday 15:00 UTC): https://wiki.openstack.org/wiki/Meetings/PuppetOpenStack - Logs http://eavesdrop.openstack.org/irclogs/%23puppet-openstack/ - On going efforts: https://trello.com/puppetopenstack
16:40 dprince joined #puppet-openstack
16:50 saikrishna joined #puppet-openstack
16:57 aimon joined #puppet-openstack
17:05 v1k0d3n joined #puppet-openstack
17:13 rgowrishankar EmilienM: Nuage is still on a monolithic plugin but there are plans to add ml2 mechanism driver in the near future..
17:14 rgowrishankar EmilienM: Thanks for the links, I have followed the existing patterns, will look at these code samples also
17:20 aimon joined #puppet-openstack
17:21 aimon joined #puppet-openstack
17:30 angdraug joined #puppet-openstack
17:32 sergmelikyan joined #puppet-openstack
17:53 jpena|away joined #puppet-openstack
17:56 vinsh In and rspec beaker upgrade test... what might change on a centos vm.. to cause openstack services to be installed in /bin instead of /usr/bin ?
18:03 crinkle differences between kilo and liberty packages?
18:03 tiswanso joined #puppet-openstack
18:03 vinsh I thought so, but other changes have merged in this project without that failure.  So its got to be something silly i'm doing.
18:04 vinsh It may have to do with the manner in which I call the init script.  calling it by name vs its path/name
18:04 vinsh It might infer /bin if not called from /usr/bin
18:04 vinsh testing.
18:05 xarses joined #puppet-openstack
18:05 EmilienM crinkle: I know you're traveling, but have you 5 min to talk about integ work with pabelanger?
18:05 crinkle EmilienM: i'm at ops midcycle, so in and out
18:06 crinkle EmilienM: i can respond to comments on gerrit?
18:06 EmilienM crinkle: ok np
18:06 EmilienM crinkle: pabelanger is concerned about integration progress, so do I
18:06 pabelanger joined #puppet-openstack
18:06 vinsh crinkle: Cool, you can see a presentation of all of the 'fun' we had with the kilo upgrade :)
18:07 EmilienM crinkle: pabelanger is mentioning an eventual design error about cloning all modules when testing keystone
18:07 crinkle EmilienM: i don't see any new comments from pabelanger on those patches, pabelanger maybe you can elaborate here?
18:08 pabelanger EmilienM: going to pastebin our private conversation, cool?
18:08 pabelanger save me some typing
18:08 EmilienM pabelanger: sure
18:08 social joined #puppet-openstack
18:09 pabelanger http://paste.openstack.org/show/420816/
18:09 pabelanger some initial comments I just did with EmilienM. I was / am planning on composing a ML post for this
18:10 pabelanger TL;DR, feel we need a (better?)spec for the puppet-openstack-integration module since we are boncing back and forth on a few points
18:11 crinkle so this design that involves installing all the modules on integration tests was discussed at length in -infra, i can try to find logs and the etherpad
18:12 EmilienM I probably miss that information but I don't see how it's a blocker for us
18:12 EmilienM to clone all modules
18:13 pabelanger well, it is not a blocker, it just create a circular deps on both beaker / integration jobs.
18:14 pabelanger but, perhaps standing back further. Maybe somebody can comment on the status of msync for puppet module. Is the plan to use it or not?  Because a lot of things I am thinking about, really depend on that process.
18:14 EmilienM some people (sbadia, _ody, spredzy) use it but in a manual way
18:15 crinkle etherpad where we talked about how to get beaker working with zuul cloner https://etherpad.openstack.org/p/puppet-git-dependencies
18:15 EmilienM pabelanger: puppet-openstack-integration should contains some manifests that run in OpenStack infra, end I guess the best way is crinkle's proposal #3
18:16 EmilienM having fixtures/scenario001.pp
18:16 EmilienM and just run it
18:16 EmilienM crinkle: thx for the url
18:17 pabelanger I'm still not clear on msync. If we are running it manually, does that mean as a project we are good to use it?
18:17 EmilienM pabelanger: we have https://github.com/openstack/puppet-modulesync-configs
18:18 pabelanger crinkle: right, I am not talking about breaking zuul-clone support. I am talking but reducing the module depends for each beaker job.  I believe everybody agreed to a Puppetfile for this
18:18 pabelanger along with r10k?
18:19 EmilienM we did
18:19 chem joined #puppet-openstack
18:20 pabelanger so, if we are using msync, as a project. Why not define each module with a specific Puppetfile to define the dependencies. Breaking the need for beaker jobs to sync with puppet-openstack-integration.
18:20 pabelanger which then breaks to need to create new beaker gate jobs for puppet-openstack-integration
18:20 pabelanger since it is currently failing today, work needs to be done to make it pass
18:21 EmilienM it's not failing because of Puppetfile, isn't?
18:21 pabelanger no
18:21 zacdev joined #puppet-openstack
18:21 pabelanger it is failing because puppet-openstack-integration does not have a spec/acceptance file that beaker needs.
18:22 pabelanger and the gate-puppet-keystone jobs, excepts puppet-keystone to exists in the local file system.  Which it is not
18:22 EmilienM pabelanger: I expected the job to run acceptance from puppet-keystone repo
18:22 pabelanger no, that is the issue
18:22 pabelanger puppet-keystone is no where on the filesystem
18:22 pabelanger nothing is checking that repo out
18:22 EmilienM yeah
18:22 pabelanger So, a new gate needs to be added to do that
18:23 EmilienM but what is the relationship with the Puppetfile?
18:23 pabelanger there is none, the gate would need to be updated to then use the local puppet-openstack-integration report to install from
18:23 pabelanger which, means now updating puppet-keystone to handle this
18:24 pabelanger It is all doable, but we start getting more complex dependencies IMO
18:24 EmilienM I just want to run puppetk-keystone beaker job by using ZUUL_BRANCH from the puppet-openstack-integration patchset
18:25 pabelanger EmilienM: yes, but how do you install puppet-keystone?
18:28 EmilienM pabelanger: good question
18:29 pabelanger I know how to fix it.
18:29 EmilienM \o/
18:29 pabelanger yes and no
18:29 pabelanger will fix this issue, but expose another
18:30 crinkle pabelanger: as I said last time this came up, it would be good to discuss this in #openstack-infra or at least rope in jeblair and nibalizer since the design for this was worked out with the infra team and they are mimicing our design
18:30 pabelanger either way, before going down the path. I feel we need a better spec in place, to make sure everybody understand both sides of it. I'm hoping we can do that at the mid-cycle.
18:31 nibalizer I'm about to pop out to lunch
18:31 crinkle I was feeling like we were pretty close to having something working, I'd like to continue on that front
18:31 nibalizer and I am not at the mid cycle
18:31 pabelanger crinkle: right, this is the main reason I'm working _upstream_ now. Because I want -infra to roll the process too
18:31 nibalizer but I will read when I get back
18:33 pabelanger like I said, a lot revolves around msync
18:33 pabelanger if we are using it, we can extend our configs to handle some of this dependency
18:34 pabelanger https://review.openstack.org/#/c/210517/ and https://review.openstack.org/#/c/210528/ are 2 reviews if people don't know
18:38 crinkle iirc the main reason we moved toward installing from a global list of modules was that we didn't want to hard-code dependencies into the tests and we didn't have a good way of programmatically reading dependencies since they're not all in the metadata.json
18:38 crinkle I'm not entirely opposed to hardcoding in dependencies but I would like to work out a way that's compatible with how infra needs to do it
18:39 crinkle I don't think what we're doing now is so broken that we need to call a state of emergency to change it
18:39 pabelanger I agree, I don't think hardcording is the way either.
18:40 pabelanger I'm just asking for somebody to explain the reason for why puppet-keystone beaker jobs needs puppet-nova to be installed?
18:40 sanjayu joined #puppet-openstack
18:40 crinkle it doesn't
18:40 crinkle but automating the knowledge that it doesn't need puppet-nova was hard
18:41 EmilienM it just allows consistency when testing modules
18:41 pabelanger Now, this could be my misunderstanding on how msync works. But if we do / can use it, we could setup a template for each module with shared dependencies like mysql and such
18:41 EmilienM I don't think this is a blocker to make progress on actual testing jobs
18:41 crinkle my general feeling on msync btw http://lists.openstack.org/pipermail/openstack-dev/2015-July/071036.html
18:42 crinkle i would generally prefer to centralize over copy
18:42 nibalizer ok im back
18:43 chem joined #puppet-openstack
18:43 EmilienM pabelanger: other than this issue, wdyt about https://review.openstack.org/#/c/201298/ ?
18:47 ducttape_ joined #puppet-openstack
18:48 nibalizer pabelanger: so I'm having a hard time understanding what the issue is, and what your proposed solution is, can you try to explain it to me?
18:49 pabelanger EmilienM: I cannot speak to option 3, as I don't have much experience with that.  For me option 2 makes sense. I agree droping the testing namespace is fine. But my whole understanding of the puppet-openstack-integration module, was it was a puppet module / project (like system-config) that would be used by people to stand up a cloud environment for the purpose of testing / development (using puppet
18:49 pabelanger modules [not git]). Again, this goes back to the first set of code I pushed up, which had some transaction layer making things generic and reusable. So, right now, the main issue I have, 6 weeks in of development (which is not much right now) and we still haven't actually committed anything into the project yet. We are still having healty discussion about how this is going to work, but still
18:49 pabelanger confusing. At least on my side
18:50 pabelanger nibalizer: break beaker-jobs depending on puppet-openstack-integration is the most basic thing.  Either move install_modules.sh out into another module, and both sync / clone from it
18:50 pabelanger other wise, we have a circular dependency
18:50 pabelanger which, when we did our first commit. broke puppet-beaker jobs
18:50 EmilienM pabelanger: we never mentioned puppet-openstack-integration was a Puppet module
18:51 EmilienM our initial discussions was about putting basic manifest that deploys OpenStack
18:52 nibalizer the integration repo seems like the correct place to keep shared tests, scripts to help (install_modules.sh), and the cannonical list of dependencies
18:53 EmilienM pabelanger: the '6 weeks' things is only because we are not yet agree. We could have our basic structure in place
18:53 crinkle pabelanger: it is specifically not a puppet module, we do not want people to use as such
18:53 EmilienM if people wants to run tests, they use beaker tests on their laptop
18:54 nibalizer so install_modules.sh is just a script, I'm not sure why we would move that
18:54 EmilienM I would be in favor or doing #3 now, have the structure in place - and get the feedback at usage (and adapt the design accordingly)
18:54 nibalizer and It sound slike you want a Puppetfile per module, which I don't think is a good idea
18:55 ducttape_ joined #puppet-openstack
18:56 pabelanger can you explain why it is not a good idea?
18:56 openstackgerrit Adam Vinsh proposed openstack/puppet-swift: WIP: Introduce swift_service class  https://review.openstack.org/203220
18:57 crinkle pabelanger: isn't that the same as hardcoding dependencies, which you said above is not what you want?
18:57 pabelanger I don't understand what you mean hardcodeding dependencies.
18:57 pabelanger because the dependency exists in the puppet module, not out side?
18:58 nibalizer pabelanger: https://github.com/puppet-community/puppet-module-puppetboard/pull/26 is what I ran into
18:58 nibalizer librarian-puppet recurses puppetfiles I guess? and everything gets pretty broken
18:59 pabelanger Right, in the past I used librarian-puppet-simplke
18:59 pabelanger or what ever it was called
18:59 * crinkle afk for lunch
18:59 nibalizer so at a /really/ high level: puppet is shit at dependencies
18:59 nibalizer there aren't any tools that do it well, the standards for experessing dependncies are bad
18:59 nibalizer there isn't runtime inspection
19:00 nibalizer so simply writing down a list of what you need, in one place, is about as good as we are going to do
19:00 tiswanso joined #puppet-openstack
19:00 pabelanger I agree, I am just suggesting we move that location to msync
19:00 nibalizer when users come to this channel like 'hey does X version of mysql work?' we can point them to one file in puppet-integration and that's the end of it
19:00 pabelanger we then either sync all projects against it
19:00 pabelanger or update clones to pull from it
19:01 pabelanger either way, it breaks our beaker-jobs / integration dependency
19:01 pabelanger look at it this way
19:01 pabelanger _what_ if we needed to update the acceptance today to change how install_modules works
19:01 pabelanger how are we expecting to do that with out breaking our gates
19:02 nibalizer ?
19:02 nibalizer install_moudles.sh is in puppet-openstack-integration so we would change it there
19:02 pabelanger we moved it into tools
19:02 pabelanger for example
19:03 pabelanger or changed the file name
19:03 nibalizer you'd have to change spec_helper_acceptance in every module
19:03 nibalizer which would be like 40 reviews
19:04 nibalizer we've discussed moving the blob in s_h_a to a centralized repo or ruby gem
19:04 pabelanger we couldn't even do that today. If puppet-keystone beaker-job was a gate for puppet-openstack-integration
19:04 pabelanger that is the circular dep
19:04 pabelanger it is the same issue we have migrating to puppet-httpd and puppet apply jobs
19:04 nibalizer right and thats the one time you're alloowed to ask infra for help unwedging you
19:04 nibalizer if you have an asymetric but wedged gate
19:05 pabelanger Right
19:05 nibalizer but install_modules.sh is 40 lines
19:05 pabelanger but again, this all goes back to msync
19:05 nibalizer and I don't really expect to have to change it ever
19:05 nibalizer maybe a little bit
19:05 pabelanger if we are using it, there is a way to get around the wedge
19:05 nibalizer but not commonly
19:05 nibalizer I think you're conflating 'growing pains' with 'fundamental flaw'
19:06 nibalizer yes getting off the ground is hard
19:06 nibalizer people need to learn the new system, it has some unexpected behaviors
19:06 nibalizer but it can work most of the time once it gets going
19:06 nibalizer and msying every repo every time just creates review churn and slows the project down
19:07 EmilienM the msync thing just does not scale when having tons of modules I would say
19:08 pabelanger I'm not saying it is a fundamental flaw, I am saying there might be a better way to do some of this. But more to the point, it looks like I'm not on the same page as most people here. Which is fine, but causing me fustrations due to things slowing down more then I am used too
19:09 pabelanger To me, it is a clear plan in my head how it works. Take devstack, rewrite it using puppet.
19:11 pabelanger something I basically did already 18 months ago. So, I am clinging to something in the past. Re-writing it to work different is a hard thing for me to grasp right now. Mostly because I don't know how other people see all this working
19:11 pabelanger the main reason I signed up to attend mid-cycle
19:11 nibalizer the mid-cycle is now virtual ya?
19:12 EmilienM nibalizer: y
19:12 pabelanger is, but MTL is 90mins away
19:12 EmilienM from 2 to 4
19:12 EmilienM nibalizer: paul and I will meet in MTL
19:13 nibalizer ok
19:13 ianw joined #puppet-openstack
19:15 nibalizer pabelanger: I believe the initial impetus for puppet-openstack-integration was only to centralize the Puppetfile and install_modules.sh
19:16 EmilienM and then a basic manifest that is run to validate the whole stack
19:16 EmilienM and then run tempest
19:19 nibalizer uh sure
19:19 pabelanger well, that is part of the confusion. I don't think everybody understand what the repo will do
19:20 nibalizer ya
19:20 pabelanger so, I think for now, I'm going to step aside on work until the mid-cycle then see if I can get some spec in place.  Because, for the most part, I am just spining my wheels right now.
19:20 nibalizer so a commit adding a README.md to that repo could be a thing
19:20 nibalizer or a spec
19:20 nibalizer pabelanger: ok
19:21 nibalizer right now I am not seeing the keystone tests gating puppet-openstackci
19:21 nibalizer er puppet-openstack-integration, sorry
19:21 nibalizer https://review.openstack.org/#/c/207070/ for example
19:22 pabelanger nibalizer: https://review.openstack.org/#/c/213502/
19:24 nibalizer so thats a misconfigure
19:24 nibalizer right so its called the keystone job
19:24 nibalizer but its just trying to run bundle exec rspec spec/acceptance in the puppet-openstack-integration repo
19:24 nibalizer which won't work
19:24 pabelanger correct
19:25 nibalizer basically the beaker jobs would have to get smarter
19:25 pabelanger what they want it to do, is checkout puppet-keystone, then have it run keystone acceptance
19:25 nibalizer or a new job could be created that is smart enough to run the spec tests that it is assigned
19:27 nibalizer is there a patch up to project-config to pull out the keystone test?
19:27 EmilienM no
19:27 nibalizer okay I will spin that up then
19:27 nibalizer its a good idea to do that I think
19:27 EmilienM I just don't know how to do that
19:27 nibalizer but it simply can't ever work
19:27 nibalizer as written
19:27 EmilienM nibalizer: yeah I did that patchn you +1'ed and it failed
19:28 nibalizer ya
19:28 nibalizer go us
19:28 nibalizer :)
19:28 EmilienM :P
19:28 EmilienM ahah
19:28 nibalizer I didn't realize what I now know
19:28 nibalizer which is that test is more adaptive than descriptive
19:28 nibalizer can you propose the patch to undo it, or propose a revert
19:29 EmilienM nibalizer: ok
19:30 EmilienM nibalizer: https://review.openstack.org/#/c/213582/
19:30 EmilienM nibalizer: look how I'm fast, I came back in the past and did it
19:30 nibalizer oh okay it was already done
19:30 nibalizer sweet
19:31 EmilienM nibalizer: my time machine is up!
19:32 EmilienM nibalizer: what kind of patch do you think?
19:32 angdraug joined #puppet-openstack
19:32 nibalizer EmilienM: ?
19:32 EmilienM pabelanger: do that mean we won't push code in integ repo before midcycle??
19:32 EmilienM nibalizer: for project-config
19:33 EmilienM "I will spin that up then"
19:34 nibalizer I was going to spin up 213582
19:34 nibalizer but you alraedy did it
19:35 pabelanger EmilienM: Ya, until I have a better understanding what people want to use the repo for.
19:36 EmilienM pabelanger: I think and hope we can solve this issue *before* the midcycle, honestly
19:36 EmilienM we were so close to have something
19:36 EmilienM tempest is running etc
19:37 * crinkle is back
19:37 crinkle I agree, I thought we were generally on the right track, I'm not sure what postponing will do
19:39 pabelanger So, if we merge the code today
19:39 pabelanger what are we going to do with the repo?
19:41 EmilienM run the integ jobs in puppet-keystone as non voting
19:41 EmilienM this is an idea^
19:41 EmilienM see how it behaves and iterate
19:43 EmilienM crinkle, pabelanger, nibalizer ^
19:44 openstackgerrit Iury Gregory Melo Ferreira proposed openstack/puppet-openstack-specs: Enabling-federation  https://review.openstack.org/190361
19:46 pabelanger right, but currently we have coverage from beaker-jobs, since it is a copy and paste for the most part
19:46 EmilienM pabelanger: for now
19:46 EmilienM pabelanger: once we have this merged, I'll push glance support
19:47 EmilienM and run tempest glance
19:47 EmilienM and add glance to puppe-glance
19:47 EmilienM ant so on
19:47 EmilienM at the end, we have a full OpenStack cloud
19:47 pabelanger EmilienM: well, I think we should address the apt-get problems we are having first
19:47 pabelanger since the jobs are bascially bouncing
19:47 EmilienM the apt-get problem?
19:47 pabelanger error 100 from apt-get?
19:47 EmilienM which error sorry, I feel dumb here
19:48 pabelanger http://logs.openstack.org/70/207070/12/check/gate-puppet-openstack-integration-dsvm-trusty/f5bf4b6/console.html.gz#_2015-08-14_19_06_42_892
19:48 pabelanger right now, it is unstable
19:49 EmilienM http://logs.openstack.org/70/207070/13/check/gate-puppet-openstack-integration-dsvm-trusty/8221fda/console.html#_2015-08-17_16_06_17_857
19:49 EmilienM we're using this puppet code everywhere in our acceptance
19:49 EmilienM and it never failing like this
19:49 EmilienM I think this is a random, rare issue
19:50 pabelanger still needs addressed no?
19:50 EmilienM sure but not a blocker for us
19:50 EmilienM I closely watch beaker jobs and I saw that once
19:50 EmilienM it's less problematic than EPEL downtimes
19:50 EmilienM that happen every day by the way
19:56 pabelanger Ya, so if somebody want to take over my patchsets for integration, I'll stand aside. I have some other work with -infra I need to finish up in the next 2 weeks.
19:57 pabelanger Only other questions are for msync. If we are _not_ going to use it, then I'll drop https://review.openstack.org/#/c/210517/
19:57 pabelanger however, I would not be in favor of people manually running it them
19:57 pabelanger since the patch set just automates the process
19:59 EmilienM pabelanger: me neither, I would like automation too
20:11 EmilienM pabelanger: I'll take over patches
20:11 iurygregory thanks EmilienM \o/
20:11 EmilienM iurygregory: cool, thx - I think it's good now, I might have missed something
20:12 iurygregory No problem ^^
20:12 iurygregory crinkle, if you have time you can review https://review.openstack.org/190361 ?
20:15 pabelanger EmilienM: crinkle: I am currently looking at: https://review.openstack.org/#/c/207070/13/modules/openstack_integration/manifests/testing/scenario001.pp
20:15 iurygregory There is a way to verify if keystone is running over apache?
20:15 pabelanger if we keep this as a simple manifest (option 3). How would we move this common logic out into scenario002, if we wanted?
20:16 EmilienM pabelanger: copy/paste for now I guess
20:16 pabelanger s/into/share with
20:16 pabelanger EmilienM: okay, other then copy and paste
20:16 EmilienM pabelanger: common.pp + include?
20:17 EmilienM iurygregory: I think yeah, you can check the vhost resource in the catalog
20:17 iurygregory ok i'll keep this in mind =)
20:17 pabelanger EmilienM: would puppet know how to resolve common.pp if we moved it into fixtures/ ?
20:18 EmilienM for now, no
20:19 pabelanger I'd be in favor of condensing to manifest folder for this purpose
20:21 crinkle that wouldn't help, it would still need to be in a modulepath or add .. to the modulepath
20:22 crinkle so that is a reasonable argument for having that directory layout
20:22 EmilienM crinkle: having fixtures is reasonable?
20:23 pabelanger right. I don't know any other way, other then keeping puppet module structure, to use the include functionality
20:23 crinkle EmilienM: no I'm saying pabelanger's directory layout is reasonable
20:23 crinkle that was something I hadn't considered
20:24 crinkle I asked for the reason on the review, this is a good answer to me
20:25 EmilienM pabelanger, crinkle: so do we agree to have manifests/scenario001.pp and manifests/common.pp ?
20:26 EmilienM err, modules/openstack_integration/manifests/scenario001.pp and common.pp
20:26 pabelanger What ever layout allows use to do include ::common from scenario001.pp
20:26 pabelanger the current was just 1 example way to do it
20:27 EmilienM pabelanger: are you git pushing it?
20:27 pabelanger EmilienM: no, in meeting right now
20:27 EmilienM pabelanger: ok
20:28 crinkle so if we do it that way we still need some kind of common manifest that includes the class, because just defining the class isn't enough
20:28 EmilienM crinkle: having modules/openstack_integration/manifests/common.pp and modules/openstack_integration/manifests/scenario001.pp that includes it is not enough?
20:30 crinkle EmilienM: we could puppet apply -e "include scenario001" if we don't anticipate having any parameters
20:30 crinkle EmilienM: but puppet apply modules/openstack_integration/manifests/scenario001.pp will have no effect
20:31 pabelanger Ya, this was another reason for using --certname for the node, and including the manifest
20:31 openstackgerrit Merged openstack/puppet-ceilometer: acceptance: enable debug & verbosity for OpenStack logs  https://review.openstack.org/213504
20:37 EmilienM pabelanger, crinkle: honestly, I don't think we should block on that now. If we have a second scenario later, we will just change the folder layout if needed
20:37 EmilienM I don't see major issue here
20:38 EmilienM we might want a first scenario up and running and see how it works
20:38 vinsh after hours of riding the struggle buss.  I find that centos sets /bin before /usr/bin in its path.. thus busting my code.
20:38 EmilienM that's why we have iteration after all
20:38 crinkle EmilienM: sounds good to me, we can always fix later
20:38 EmilienM design is not really critical here
20:38 EmilienM it's not like designing an API or...
20:38 EmilienM vinsh: stay with us
20:39 vinsh ;) deep breath
20:39 EmilienM pabelanger: wdyt?
20:41 angdraug joined #puppet-openstack
20:42 EmilienM I vote for having fixtures/scenario001.pp now, make it work, run tempest, merge it, gate puppet-keystone with that as non voting and see how it behaves
20:43 EmilienM then, during midcycle, pabelanger and I will see how we can scale this out and solve our issues
20:43 EmilienM but at least we will have great feedback in the meantime
20:43 EmilienM I don't want to wait until september to have a job workin
20:45 pabelanger EmilienM: rather fix it now, other then kicking it down the road. But that is just me. Like I said, going on sidelines until summit, so feel free to continue on your changes.
20:49 dgurtner joined #puppet-openstack
20:58 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: Create puppet folder layout  https://review.openstack.org/201298
20:58 EmilienM crinkle: ^
21:11 vinsh I need to find a way to set the PATH variable used in a puppet run.  Anyone messed with something like that before?
21:16 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: scenario001: deploy MySQL and Keystone (WSGI)  https://review.openstack.org/207070
21:17 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: Create puppet folder layout  https://review.openstack.org/201298
21:18 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: scenario001: deploy MySQL and Keystone (WSGI)  https://review.openstack.org/207070
21:20 openstackgerrit Adam Vinsh proposed openstack/puppet-swift: WIP: Introduce swift_service class  https://review.openstack.org/203220
21:27 ducttape_ joined #puppet-openstack
21:41 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: Create puppet folder layout  https://review.openstack.org/201298
21:42 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: scenario001: deploy MySQL and Keystone (WSGI)  https://review.openstack.org/207070
21:44 openstackgerrit Merged openstack/puppet-neutron: Prepare Liberty release  https://review.openstack.org/209294
21:50 tiswanso joined #puppet-openstack
22:05 tiswanso joined #puppet-openstack
22:10 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: scenario001: configure and run tempest  https://review.openstack.org/207078
22:11 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: scenario001: configure and run tempest  https://review.openstack.org/207078
22:15 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: scenario001: deploy MySQL and Keystone (WSGI)  https://review.openstack.org/207070
22:16 openstackgerrit Emilien Macchi proposed openstack/puppet-openstack-integration: scenario001: configure and run tempest  https://review.openstack.org/207078
22:19 aimon joined #puppet-openstack
22:34 chem joined #puppet-openstack
23:06 tiswanso joined #puppet-openstack
23:19 zacdev joined #puppet-openstack
23:28 openstackgerrit Adam Vinsh proposed openstack/puppet-swift: WIP: Introduce swift_service class  https://review.openstack.org/203220
23:32 xingchao joined #puppet-openstack
23:59 sergmelikyan joined #puppet-openstack

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