Perl 6 - the future is here, just unevenly distributed

IRC log for #puppet-openstack, 2015-06-16

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

All times shown according to UTC.

Time Nick Message
00:15 rcallawa_ joined #puppet-openstack
00:18 stevemar joined #puppet-openstack
00:25 rcallawa joined #puppet-openstack
00:30 rcallawa_ joined #puppet-openstack
00:48 xarses joined #puppet-openstack
00:55 ducttape_ joined #puppet-openstack
00:56 ducttape_ joined #puppet-openstack
01:01 gildub joined #puppet-openstack
01:17 openstackgerrit Gilles Dubreuil proposed openstack/puppet-keystone: support for keystone v3 api - v3 domain utility code  https://review.openstack.org/178385
01:27 openstackgerrit Clayton O'Neill proposed openstack/puppet-designate: Add support for RabbitMQ clustering  https://review.openstack.org/192033
01:31 cwolferh joined #puppet-openstack
01:33 openstackgerrit Gilles Dubreuil proposed openstack/puppet-keystone: support for keystone v3 api - the domain resource  https://review.openstack.org/174439
01:39 openstackgerrit Merged openstack/puppet-neutron: Create file /etc/default/neutron-server  https://review.openstack.org/130937
02:06 openstackgerrit Gilles Dubreuil proposed openstack/puppet-keystone: keystone_service resource using API version 3  https://review.openstack.org/184195
02:20 ducttape_ joined #puppet-openstack
02:28 richm gildub: https://bugs.launchpad.net/puppet-keystone/+bug/1465465
02:28 openstack Launchpad bug 1465465 in puppet-keystone "keystone_user password does not work" [Undecided,New]
02:28 richm The auth restructure code broke this
02:31 fedexo joined #puppet-openstack
02:49 sanjayu joined #puppet-openstack
03:32 blentz joined #puppet-openstack
03:33 reed joined #puppet-openstack
03:36 mfisch joined #puppet-openstack
05:20 openstackgerrit Colleen Murphy proposed openstack/puppet-keystone: Use module install script for acceptance tests  https://review.openstack.org/190839
05:30 stevemar joined #puppet-openstack
05:36 gfidente joined #puppet-openstack
05:36 gfidente joined #puppet-openstack
05:57 jtomasek joined #puppet-openstack
06:05 hostranger joined #puppet-openstack
06:05 hostranger left #puppet-openstack
06:45 xarses joined #puppet-openstack
06:47 xarses joined #puppet-openstack
07:02 tshtilma joined #puppet-openstack
07:06 dgurtner joined #puppet-openstack
07:06 dgurtner joined #puppet-openstack
07:10 paramite joined #puppet-openstack
07:14 jpena joined #puppet-openstack
07:35 openstackgerrit Colleen Murphy proposed openstack/puppet-keystone: Use module install script for acceptance tests  https://review.openstack.org/190839
08:10 derekh joined #puppet-openstack
08:23 danieru joined #puppet-openstack
08:24 dgurtner joined #puppet-openstack
08:24 dgurtner joined #puppet-openstack
08:36 openstackgerrit Javier Peña proposed openstack/puppet-horizon: Add support for WEBROOT in local_settings  https://review.openstack.org/173327
08:48 xarses joined #puppet-openstack
08:49 xarses joined #puppet-openstack
08:51 aarefiev joined #puppet-openstack
08:59 openstackgerrit Denis Egorenko proposed openstack/puppet-sahara: Enable tests for oslo_messaging_rabbit/rabbit_hosts parameter  https://review.openstack.org/191913
09:10 cdent joined #puppet-openstack
09:42 barthalion https://review.openstack.org/#/c/191913/ some discussion desired here
09:43 barthalion because we're not entirely sure if there is some principle to follow about explicit/implicit oslo_messaging_rabbit/rabbit_hosts
09:46 xavpaice joined #puppet-openstack
10:40 degorenko EmilienM, hi, can you merge this change https://review.openstack.org/#/c/176924/ please?
10:42 skolekonov joined #puppet-openstack
10:43 skolekonov EmilienM, hi, could you please take a look at this patch https://review.openstack.org/#/c/168848/ ? We discussed it in comments some time ago and it's really old
10:59 rcallawa joined #puppet-openstack
11:29 dprince joined #puppet-openstack
11:58 ducttape_ joined #puppet-openstack
11:59 morazi joined #puppet-openstack
12:09 EmilienM good morning
12:14 degorenko morning :)
12:15 EmilienM degorenko: I'm doing a 'recheck' on your patch, because beaker jobs are red
12:15 degorenko EmilienM, yes, i saw
12:16 social hmm crinkle is not around yet I guess?
12:17 EmilienM social: nop, too early for her now
12:17 social Error: Could not autoload puppet/type/vs_port: Could not autoload puppet/provider/vs_port/ovs_redhat: Could not find parent provider ovs of ovs_redhat
12:18 social https://review.openstack.org/#/c/189301/1/lib/puppet/provider/vs_port/ovs_redhat.rb
12:18 social I'd like to know if I'm not just fixing package/puppet bug
12:19 barthalion EmilienM: hi there, could you take a look at our comments here: https://review.openstack.org/#/c/191913/ ?
12:19 social becuse turns out we might need to make juno work with puppet 4 as we have packstack juno in fedora 22 :/
12:20 EmilienM barthalion: will look
12:20 barthalion thank you :)
12:34 openstackgerrit Merged openstack/puppet-cinder: Add cinder::scheduler::filter for managing scheduler.filter  https://review.openstack.org/176924
12:36 openstackgerrit Emilien Macchi proposed openstack/puppet-neutron: acceptance: test OVS bridge listing  https://review.openstack.org/192176
12:37 EmilienM skolekonov: ^
12:38 EmilienM I did this patch on top of yours, to check if br-int is really created
12:38 skolekonov EmilienM, thank you, let's wait for test results
12:43 Madkiss EmilienM: hey.
12:43 Madkiss EmilienM: what is the plan for kilo?
12:43 EmilienM Madkiss: the plan for a release?
12:43 Madkiss yes
12:44 EmilienM we're still waiting for a few patches to merge, that are blocking this release
12:44 EmilienM otherwise it's happening soon
12:46 Madkiss i see, thanks
12:48 rcallawa joined #puppet-openstack
12:49 rcallawa joined #puppet-openstack
12:51 openstackgerrit Lukas Bezdicka proposed openstack/puppet-vswitch: Explicitly say that ovs_redhat parent is ovs  https://review.openstack.org/189301
12:52 social EmilienM: crinkle: now I understand it, it worked with older pupept as it isn't that picky about what to load, one must specify that we are loading from puppet type vs_port ovs provider now
12:58 openstackgerrit Merged openstack/puppet-neutron: Drop OVS & LB monolitic plugins  https://review.openstack.org/190395
12:58 ferest joined #puppet-openstack
13:03 EmilienM skolekonov: can you rebase https://review.openstack.org/#/c/168848 please? We finally dropped monolotic drivers
13:03 EmilienM (ovs+lb only)
13:05 bdemers joined #puppet-openstack
13:17 dgurtner joined #puppet-openstack
13:19 richm1 joined #puppet-openstack
13:21 imcsk8|z1z joined #puppet-openstack
13:25 openstackgerrit Sergey Kolekonov proposed openstack/puppet-neutron: Do not create tun and int bridges manually  https://review.openstack.org/168848
13:25 openstackgerrit Sergey Kolekonov proposed openstack/puppet-neutron: acceptance: test OVS bridge listing  https://review.openstack.org/192176
13:26 skolekonov EmilienM, ^ I've rebased both patches
13:26 EmilienM cool, will see results :)
13:44 RichardRaseley joined #puppet-openstack
13:54 RichardRaseley EmilienM: I went ahead and added a discussion note about the mailing list to the weekly meeting etherpad.
13:54 EmilienM RichardRaseley: cool
13:55 RichardRaseley In the original email I should've said 'some of us want to move forward' vs. 'we want to move forward'. 'We' implied that there was an agreement on this team, vs. just some people in it scoping out the change.
13:55 degorenko EmilienM, i have question about puppet-sahara. There is only master branch. I suggest update master to current kilo features and create stable/kilo and freeze them. What do you think?
13:56 EmilienM degorenko: stable/kilo branch is already in our mind, we're just waiting for some patches to land
13:58 degorenko EmilienM, that's great. Also, puppet-sahara have some hardcode for broken ubuntu packages. I think it's wrong, we don't care about bugs in package specs.
14:00 EmilienM degorenko: that's true, I did that code in the meantime it's fixed in UCA
14:00 EmilienM degorenko: jamespage is already working on this for Liberty I think
14:00 EmilienM but we still have Kilo broken
14:00 EmilienM degorenko: this is not an example of what we like, but at least it brings beaker jobs green
14:01 EmilienM degorenko: don't follow this example :)
14:01 degorenko EmilienM, i can help with update puppet-sahara :)
14:02 degorenko EmilienM, actually, puppet-sahara is outdated now, because from kilo sahara uses distributed mode (api, engine), but here in puppet - all in one mode now
14:03 degorenko EmilienM, i can start working on it, and also remove hardcode
14:03 xarses joined #puppet-openstack
14:04 xarses joined #puppet-openstack
14:04 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - the user resource  https://review.openstack.org/174976
14:06 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - the user_role resource  https://review.openstack.org/174999
14:06 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - keystone and keystone::roles::admin  https://review.openstack.org/175060
14:07 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - add domains to service_identity  https://review.openstack.org/180586
14:18 RichardRaseley joined #puppet-openstack
14:18 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - the project resource  https://review.openstack.org/174468
14:19 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - the user resource  https://review.openstack.org/174976
14:19 RichardRaseley EmilienM: I am a little confused https://review.openstack.org/#/c/70458/ , it seems like the patch which was merged had the inline template scheme vs. the scheme we all said we preferred in patch set 8...
14:19 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - the user_role resource  https://review.openstack.org/174999
14:20 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - keystone and keystone::roles::admin  https://review.openstack.org/175060
14:20 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - add domains to service_identity  https://review.openstack.org/180586
14:21 RichardRaseley 'we all' being Hunner, _ody, and myself.
14:21 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - beaker tests  https://review.openstack.org/178828
14:22 RichardRaseley EmilienM: Also, one of the voting tests still seems to be failing (tripleo).
14:22 RichardRaseley Err... maybe it isn't voting, but it isn't marked as non-voting.
14:23 ducttape_ joined #puppet-openstack
14:24 social anyone hit this? http://paste.fedoraproject.org/232602/43446462
14:29 social paramite: ^^
14:29 rcallawa_ joined #puppet-openstack
14:33 openstackgerrit Richard Megginson proposed openstack/puppet-openstack-specs: Support Keystone v3 API in openstack puppet modules  https://review.openstack.org/150108
14:38 crinkle morning
14:42 vinsh joined #puppet-openstack
14:43 RichardRaseley crinkle: morning.
14:43 social crinkle: could you look at https://review.openstack.org/#/c/189301/ and any hints for http://paste.fedoraproject.org/232602/43446462
14:43 xarses_ joined #puppet-openstack
14:44 kkirkpatrick joined #puppet-openstack
14:44 xarses_ joined #puppet-openstack
14:47 dgurtner joined #puppet-openstack
14:47 dgurtner joined #puppet-openstack
14:49 stevemar joined #puppet-openstack
14:50 saneax joined #puppet-openstack
14:50 mdorman joined #puppet-openstack
14:50 crinkle social: is that paste using puppet 4?
14:51 crinkle social: wrt 189301 i wasn't giving a -1 i was just curious if there was something documented that we could refer to
14:52 social crinkle: but I was wrong in 189301 the change is imho more correct
14:53 social crinkle: yes the paste is puppet 4, if I downgrade run it and rerun with updated puppet 4 it works
14:53 EmilienM weekly meeting in 8 min!
14:53 xingchao joined #puppet-openstack
14:54 RichardRaseley EmilienM: I might be able to make the first few mins of this one! =X
14:56 crinkle social: looks like `model` isn't a thing in puppet 4, we'll have to figure out a better way to look up catalogue resources from the provider
14:56 mfisch joined #puppet-openstack
14:59 cwolferh joined #puppet-openstack
15:00 RichardRaseley Weekly meeting starting now in #openstack-meeting-4
15:01 mfisch joined #puppet-openstack
15:04 Hunner _ody: You around for meeting?
15:07 social crinkle: I'll try to look into that further but I'm open to suggestions, maybe just git greping other modules for catalog might help me :)
15:09 openstackgerrit Mathieu Gagné proposed openstack/puppet-nova: Introduce public_url, internal_url and admin_url (and v3/ec2)  https://review.openstack.org/70458
15:36 mgagne EmilienM: I need we need those patches so we can deprecate the old parameters ASAP
15:36 EmilienM mgagne: I +2 them
15:36 EmilienM mgagne: please raise them here, so other reviewers can have a look and maybe land it
15:37 _ody sbadia: Your inlline comment.  Yeah they have to be set for Puppet 4 because we do comparisions against them and Puppet 4 doesn't let you compare Strings to undef.  I'll implement your comment, I was going to do a one over for something mgagne and I were talking about yesterday.
15:38 sbadia _ody: ack, thanks :)
15:38 sbadia ok for the precision
15:40 openstackgerrit Cody Herriges proposed openstack/puppet-neutron: Sync up module with latest msync commit.  https://review.openstack.org/189873
15:44 _ody sbadia: You think it is alright to move default fact to spec_helper.rb across all our modules, i.e. adding it to the msync configs?
15:45 openstackgerrit Merged openstack/puppet-designate: Update spec_helper_acceptance for project renames  https://review.openstack.org/191328
15:45 openstackgerrit Merged openstack/puppet-neutron: Fix support for auth_uri setting in neutron provider  https://review.openstack.org/190634
15:46 fedexo joined #puppet-openstack
15:46 sbadia _ody: seems legit for operatingsystem and release no?
15:46 * sbadia take a look
15:48 sbadia hum, it may be painful :-s
15:49 _ody It seems logical that all modules should be tested with the same default facts...but...
15:50 _ody I suspect we'd have edits to most all the rspec-puppet tests to account for it.
15:50 sbadia yep :-s
15:50 sbadia we can maybe keep this for a msync iteration 2 :)
15:51 _ody Ok.  I'll see if there is another way to quickly do testwide default facts, else it'll just remain individually defined in each test for now.
15:54 openstackgerrit Merged openstack/puppet-ceilometer: Update spec_helper_acceptance for project renames  https://review.openstack.org/191326
15:55 EmilienM sbadia: puppet-gnocchi requires some work now, packaging is in Fedora: https://repos.fedorapeople.org/repos/openstack/openstack-kilo/f21/
15:55 EmilienM sbadia: I'll have a look later
15:55 sbadia EmilienM: héhé :-D i'm a debian packager :p
15:55 sbadia EmilienM: I can help ?
15:56 EmilienM sbadia: yes, but no
15:56 EmilienM sbadia: we don't test on debian, only ubuntu UCA
15:56 EmilienM sbadia: and afik you're not UCA maintainer :)
15:56 sbadia yes I know ;-)
15:56 EmilienM sbadia: though maybe you can talk with james page
15:57 EmilienM sbadia: so I can focus on bringing functional testing on this modules and also fix the module to work with fedora packaging
15:57 sbadia ack, I take a look and I ask you if questions/problems
16:08 RichardRaseley EmilienM: Sorry I had to cut out on the meeting - I just wanted to follow-up with you on the two topics: puppet-zaqar and the openstack-operators mailing list.
16:08 rcallawa joined #puppet-openstack
16:09 RichardRaseley With regard to puppet-zaqar, are you OK with how things are progressing and with the plan to use cookie-cutter to initially populate that module?
16:10 xarses joined #puppet-openstack
16:12 EmilienM RichardRaseley: I'm good with the general work done here. Though I have no clue about the new repo creation, if your patch will be enough
16:13 RichardRaseley EmilienM: OK, good to know. I will be sure to work with the openstack-infra team to push that through to completion, making any necessary tactical adjustments along the way.
16:13 EmilienM RichardRaseley: cool, thanks for working on that
16:14 EmilienM RichardRaseley: make sure to talk with sbadia & spredzy|afk - they are leaders on the cookiebutter work
16:14 RichardRaseley +1
16:14 RichardRaseley EmilienM: On the second item (mailing list for usage questions) it seems there is general consensus that this is a good path forward. My thoughts were that we would (1) send out a notice to puppet-openstack noting the proposed change and soliciting feedback (2) assuming no blockers, set a date (1 or 2 weeks in the future) when we will mark the list as read-only archives and add an auto-responder directin
16:14 RichardRaseley g people to the dev and operators lists.
16:15 RichardRaseley Does that generally sound like the right approach?
16:15 EmilienM I'm good with that, if we are coherent with what openstack-operators use this ML
16:15 EmilienM consistency looks good here
16:16 RichardRaseley The feedback I've gotten from that very question on that list seems positive in that regard.
16:16 EmilienM RichardRaseley: do you know who manages the puppet-openstack google ML ?
16:16 RichardRaseley I don't know, but I can 'grab' control via Google Groups (part of Puppet's Google Apps organization).
16:16 RichardRaseley EmilienM ^
16:19 openstackgerrit Merged openstack/puppet-nova: move setting of novncproxy_base_url  https://review.openstack.org/190464
16:34 xarses_ joined #puppet-openstack
16:37 degorenko EmilienM, can i start working on puppet-sahara or not? :)
16:38 EmilienM degorenko: depends of what you want to do
16:38 EmilienM degorenko: my goal of today is to figure out the failing beaker tests
16:38 degorenko EmilienM, as i say above: i can help with update puppet-sahara :) actually, puppet-sahara is outdated now, because from kilo sahara uses distributed mode (api, engine), but here in puppet - all in one mode now. i can start working on it, and also remove hardcode
16:38 EmilienM but I truly think we need to keep the ugly code that does packaging work because UCA packaging is broken fo rsahara
16:39 EmilienM degorenko: if you remove hardcode, beaker will never pass and the patch will *never* merge
16:40 degorenko EmilienM, yes, i know. But in one time ubuntu specs will be fixed
16:40 EmilienM degorenko: I'm not sure jamespage will backport the changes done in Liberty to Kilo
16:41 degorenko EmilienM, what exactly  jamespage do now?
16:41 RichardRaseley EmilienM: When convenient, we need your explicit PTL blessing on our addition of puppet-zaqar. (https://review.openstack.org/#/c/191942/)
16:42 EmilienM degorenko: what I mean is: if you drop the hard code now, you'll break puppet-designate CI job on trusty
16:43 EmilienM degorenko: so you can do whatever patch you like but if the patch is not passing beaker CI, it will never land
16:43 EmilienM RichardRaseley: will do
16:43 RichardRaseley EmilienM: Thank you.
16:43 degorenko EmilienM, i got it, thanks
16:44 EmilienM RichardRaseley: done
16:44 RichardRaseley +1
16:44 EmilienM degorenko: so today, i'm investigating the failure logs, and try to figure out what's wrong. In the meantime, feel free to patch the module to add your distributed mode thing
16:45 degorenko EmilienM, great! Thanks!
16:47 ferest joined #puppet-openstack
16:57 openstackgerrit Cody Herriges proposed openstack/puppet-neutron: Make tests pass on Puppet 4.x  https://review.openstack.org/184646
16:58 rcallawa_ joined #puppet-openstack
16:59 EmilienM crinkle: beaker jobs are about to vote now: https://review.openstack.org/#/c/190778/
16:59 EmilienM *boom*
16:59 _ody mgagne: https://review.openstack.org/184646 removes the ugliness of testing for empty string.  Testing for it was a fault of the spec test, after another look.  Users aren't actually ever going to set those variables to empty string.
17:00 mgagne _ody: glad to hear it, thanks!
17:00 _ody (or rather if they do they are doing so on purpose and not what we are really testing for.)
17:15 EmilienM degorenko: if you want to know what breaks beaker/trusty: http://logs.openstack.org/28/191328/1/check/gate-puppet-designate-puppet-beaker-rspec-dsvm-trusty/669627f/console.html#_2015-06-16_00_12_04_390
17:19 vinsh joined #puppet-openstack
17:20 EmilienM degorenko: in fact, puppet-sahara works now with ubuntu, it's a centos issue. Though I'm running issues on trusty with designate
17:21 xarses So how well do we support working with multiple regions? (back to the remote keystone host from yesterday)
17:21 xarses I don't see anything explicit adding the region to the openstacklib auth calls
17:21 degorenko EmilienM, is it test for sahara?
17:22 EmilienM bug reported here https://bugs.launchpad.net/puppet-designate/+bug/1465780
17:22 openstack Launchpad bug 1465780 in puppet-designate "Beaker job is broken on Trusty" [Undecided,New]
17:22 EmilienM afk 30min now
17:22 EmilienM degorenko: yes they are in repo already
17:22 degorenko EmilienM, actually, as i know, in trusty problem with sahara-db-manage which try to use sqlite3 which sahara doesnt support
17:29 mfisch I can't seem to vote on reviews anymore
17:30 mfisch Invalid Authentication method...
17:30 mfisch nm, reload fixes it
17:41 EmilienM degorenko: well, if you read the commit message of why I did that in Sahara, you'll see the exact same reason, let me find the link
17:42 EmilienM degorenko: https://github.com/openstack/puppet-sahara/commit/e12840fe780624b99449718c3cde9443efaeb078
17:42 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: keystone_user password does not work  https://review.openstack.org/192330
17:46 dgurtner joined #puppet-openstack
17:57 rcallawa joined #puppet-openstack
18:09 aimon joined #puppet-openstack
18:11 openstackgerrit Cody Herriges proposed openstack/puppet-modulesync-configs: Manage default configs  https://review.openstack.org/177959
18:17 openstackgerrit Cody Herriges proposed openstack/puppet-modulesync-configs: Move to rspec-puppet 2.2.0  https://review.openstack.org/189886
18:29 _ody sbadia: I didn't include the disabling of documentation related lint checks in latest puppet-modulesync-configs patch set.  I'll update all modules first and see what the failure rate it.  I already know neutron doesn't pass.
18:37 EmilienM degorenko: so there is no bug on sahara, it was an inconsistent issue with packaging download
18:38 EmilienM degorenko: that means only designate has an issue, I'm on it now. So you can send a patch for sahara, it should pass CI
18:52 dmsimard Have you guys bumped into any rspec related issues due to the recent version bump in a couple rspec gems ?
19:00 openstackgerrit Cody Herriges proposed openstack/puppet-nova: Enable nova service by default  https://review.openstack.org/184656
19:03 _ody dmsimard: We moved to rspec-puppet 2.1.0 a few months back and we're about to go to 2.2.0 and I have haven't seen any issue.  Also have not tested any newer non-puppet rspec packages as of late.
19:10 _ody https://review.openstack.org/#/c/186887 https://review.openstack.org/#/c/186955 https://review.openstack.org/#/c/186933 are all dups of https://review.openstack.org/#/c/184656
19:10 _ody Author isn't on IRC.
19:11 _ody ...and https://review.openstack.org/#/c/184570
19:12 * _ody will email him
19:38 RichardRaseley joined #puppet-openstack
20:38 soundboy45 joined #puppet-openstack
20:39 rcallawa joined #puppet-openstack
20:40 mdorman is there any way to use a collector in a conditional?  something like:   if Package<| name == $package_name |> { … }  ?
20:40 mdorman i want do to something like defined(), except be able to search for the package name, rather than the resource title
20:43 rcallawa joined #puppet-openstack
20:43 _ody mdorman: I doubt it since a collector actually collects resources and not data.
20:48 mdorman right.  i’m thinking this is not going to happen
21:03 mgagne how about we abstract package installation in classes and depend on them instead?
21:03 mgagne and we make sure there is no containment issue, etc.
21:04 mdorman trying to get a picture in my brain of what that looks like
21:04 mgagne and add alias to package resource which installs 2 "services" and hope that relationship will recognize them: https://docs.puppetlabs.com/references/latest/metaparameter.html#alias
21:05 mgagne mdorman: put the package resource declaration in a class and hide the logic in there
21:06 mgagne mdorman: and have others depends on it instead of the actual package resource
21:06 mdorman that’s kind of where i was going with this, but just containing the logic within the generic_service class instead
21:07 mgagne some people made the request to remove generic_service in nova
21:07 mgagne as it introduces some technical limitations, don't know what kind tbh
21:07 mdorman i mean it seems like this generic_service case is really the only place where we run into this package definition problem
21:07 mgagne mdorman: no, see puppet-glance
21:08 mgagne or puppet-neutron
21:08 mgagne https://github.com/openstack/puppet-glance/blob/master/manifests/registry.pp#L165-L172
21:08 mgagne https://github.com/openstack/puppet-neutron/blob/master/manifests/plugins/ml2.pp#L153-L167
21:08 mgagne those kinds of exceptions are everywhere
21:12 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - add domains to service_identity  https://review.openstack.org/180586
21:12 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - the user_role resource  https://review.openstack.org/174999
21:12 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - v3 domain utility code  https://review.openstack.org/178385
21:12 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - keystone and keystone::roles::admin  https://review.openstack.org/175060
21:12 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - the user resource  https://review.openstack.org/174976
21:12 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - the project resource  https://review.openstack.org/174468
21:12 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - beaker tests  https://review.openstack.org/178828
21:12 openstackgerrit Richard Megginson proposed openstack/puppet-keystone: support for keystone v3 api - the domain resource  https://review.openstack.org/174439
21:13 mdorman mgagne:   kk, thanks
21:23 openstackgerrit Cody Herriges proposed openstack/puppet-nova: Enable nova service by default  https://review.openstack.org/184656
21:27 openstackgerrit Mike Dorman proposed openstack/puppet-nova: Better handling of package dependencies in nova generic_service  https://review.openstack.org/192009
21:27 mdorman mgagne:  your thoughts on ^^
21:28 _ody crinkle: If I have an array in config_defaults.yml for msync and I want to add a single value to the array in a .sync.yml file, locally in a modules...Can I or am I copy and pasting the entire array?
21:28 mgagne mdorman: what if service_name is undef?
21:28 mgagne mdorman: otherwise it looks like a good compromise
21:29 crinkle _ody: you can't do that exactly, but you could have the template concatenate a var from config_defaults.yml with a differently named var from .sync.yml
21:30 crinkle _ody: sort of like https://github.com/puppetlabs/modulesync_configs/blob/master/moduleroot/Gemfile#L14
21:31 crinkle _ody: otherwise, yeah you'd have to copy the whole array into .sync.yml
21:32 _ody ah.  That might be ok.
21:32 _ody mgagne: re: "The purpose of having $nova_title instead of $package_name is to generate a predictable package resource title so other manifests can reference it."
21:33 _ody I've see that pattern in other Puppet OpenStack modules.  Do you remember the reasoning for that pattern over just creating dependencies on entire classes?
21:33 mgagne _ody: [...] without having to include nova::params and use variables from there
21:33 mgagne _ody: we just didn't structure the module that one in the first place
21:34 mgagne _ody: I guess it's just a matter of time and planning the refactor
21:34 _ody Ok.  Just wondering if that was the case.
21:34 mgagne _ody: which I wouldn't be against
21:35 mgagne _ody: and would oddly ease support for other installation methods such as venv
21:36 mdorman mgagne:   thanks i will look at it some more
21:37 mgagne _ody: what I hope for is: abstract installation in class, add ability to switch installation methods without major refactor, use anchors to help people write new installation methods and/or hook themselves in-between/before/after installation steps
21:37 mgagne _ody: and add the ability for our users to override values which are currently hardcoded in params
21:38 mgagne _ody: same with service management
21:38 mgagne it might take a year to complete the refactor in a backward compatible way but it's not an impossible task
21:41 _ody ;) that last statement is why I often say, once you open up a puppet code refactor it never closes...you just keep opening them.
21:42 mdorman mgagne:   not sure if it matters if $service_name is not set?  the title of the service resource will be $nova_title regardless of that
21:42 mgagne mdorman: catalog compilation will fail if it's not defined which happens to be a supported use case as per the conditional below
21:43 mgagne mdorman: just wrap your new relationship in if $service_name and be done
21:44 mdorman oh gottcha gottcha, i see it now
21:50 tfz joined #puppet-openstack
21:53 openstackgerrit Mike Dorman proposed openstack/puppet-nova: Better handling of package dependencies in nova generic_service  https://review.openstack.org/192009
22:07 crinkle how would people feel if beaker-rspec installed external modules (apache, mysql, etc) from git instead of with `puppet module install`?
22:12 nibalizer I think thats the way to go
22:12 nibalizer especially if/when combined with one shared list of modules
22:12 nibalizer that way that list becomes the list you hand to users "install this, it will work" instead of any number of other combinations that could come out of puppet module install commands
22:13 crinkle right - it'd be nice to maybe have a general puppetfile that could be distributed as well as tested against
22:14 nibalizer I'm also okay with puppet module install --no-dependencies or just curling the tarball at version from the forge
22:14 nibalizer ya a Puppetfile would be great
22:14 nibalizer crinkle: what about this: 1) r10k runs against shared Puppetfile 2) cat Puppetfile | grep openstack/puppet | xargs zuul-cloner
22:15 crinkle nibalizer: yeah i'm in support of something like that
22:15 nibalizer since zuul cloner will, if needed, wipe out any directory that r10k alreayd laid down
22:16 nibalizer would cat Puppetfile | grep openstack/puppet grab all the INTEGRATION_MODULES ?
22:16 nibalizer put another way, can we infer from name if a module is under INTEGRATION
22:17 crinkle it would also grab the openstack_extras module which doesn't really need to be integrated but also doesn't hurt to pull from zuul
22:17 openstackgerrit Cody Herriges proposed openstack/puppet-ceilometer: Sync up module with 4f259b6 of puppet-modulesync-configs  https://review.openstack.org/189862
22:18 crinkle an argument against moving away from puppet module install is that we wouldn't catch issues with unpinned dependencies in metadata.json, like we did with puppet-openstacklib and rabbitmq/apt
22:18 crinkle but counterargument to that is that calling puppet module install one at a time isn't really exercising dependnecy resolution properly anyway
22:39 openstackgerrit joined #puppet-openstack
22:47 nibalizer crinkle: yea I think an end user can call puppet module install on any module, or multiples in any order
22:47 nibalizer and figuring out whatever path it might take is a bit crazy
22:55 mdorman how can i test a dependency relationship created via collector in rspec?  it fails on with(require => whatever)  is there some other way i can state that such that the collector piece will get picked up?
22:56 xavpaice anyone else having issues with apt.puppetlabs.com?
22:56 xavpaice no files... http://apt.puppetlabs.com/pool/trusty/main/p/puppet/
23:00 saneax joined #puppet-openstack
23:13 rcallawa_ joined #puppet-openstack
23:14 xavpaice and now it's back....
23:16 openstackgerrit Mike Dorman proposed openstack/puppet-nova: Better handling of package dependencies in nova generic_service  https://review.openstack.org/192009
23:17 _ody xavpaice: Because i just fixed it...
23:17 _ody ;)
23:17 xavpaice <3
23:18 xavpaice you just unblocked a day's worth of CI builds in our office, many thanks
23:18 Hunner anyone else having issues with http://apt.puppetlabs.com/pool/jessie/main/p/puppet/ not having files?
23:18 Hunner :o
23:19 * Hunner is just trolling
23:20 dalees joined #puppet-openstack
23:20 mdorman hehe
23:21 _ody xavpaice: Let me know if you see other issues with the repo.  I had to do an ad-hock failover of our apt/yum repos from an unknown state to a different box last week.
23:23 xavpaice no worries, thanks for fixing it
23:31 gildub joined #puppet-openstack
23:33 _ody joined #puppet-openstack
23:33 openstackgerrit Merged openstack/puppet-neutron: Sync up module with latest msync commit.  https://review.openstack.org/189873
23:35 openstackgerrit Merged openstack/puppet-neutron: Make tests pass on Puppet 4.x  https://review.openstack.org/184646
23:37 gildub richm, hi - regarding [Bug 1465465] Re: keystone_user password does not work, why "is that still a problem, because puppet-keystone keystone_user password does not catch and handle this error", again there is no need to handle a wrong password because if it's wrong we can't get a token anyway
23:37 openstack bug 1465465 in puppet-keystone "keystone_user password does not work" [Critical,In progress] https://launchpad.net/bugs/1465465 - Assigned to Richard Megginson (rmeggins)
23:39 gildub richm, that said handling other errors would be needed though :)
23:42 gildub richm, ok I see the v3 different message, never mind
23:44 stevemar joined #puppet-openstack
23:46 openstackgerrit Sebastien Badia proposed openstack/puppet-designate: Introduce public_url, internal_url and admin_url  https://review.openstack.org/70393
23:49 aimon joined #puppet-openstack
23:51 aimon_ joined #puppet-openstack
23:51 openstackgerrit Merged openstack/puppet-sahara: Introduce public_url, internal_url and admin_url  https://review.openstack.org/190816
23:51 openstackgerrit Merged openstack/puppet-neutron: Introduce public_url, internal_url and admin_url  https://review.openstack.org/70411
23:52 EmilienM crinkle: about your question (puppet module install). +1 for a Puppetfile (if possible common). I guess you're facing issues with zuul cloner thing?
23:52 openstackgerrit Merged openstack/puppet-nova: Introduce public_url, internal_url and admin_url (and v3/ec2)  https://review.openstack.org/70458
23:53 ducttape_ joined #puppet-openstack
23:54 openstackgerrit Merged openstack/puppet-ironic: Introduce public_url, internal_url and admin_url  https://review.openstack.org/70160
23:58 xingchao joined #puppet-openstack
23:59 openstackgerrit Merged openstack/puppet-manila: Introduce public_url, internal_url and admin_url  https://review.openstack.org/190802
23:59 openstackgerrit Merged openstack/puppet-trove: Introduce public_url, internal_url and admin_url  https://review.openstack.org/190822

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