Perl 6 - the future is here, just unevenly distributed

IRC log for #puppet-openstack, 2014-08-28

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

All times shown according to UTC.

Time Nick Message
00:09 openstackgerrit Gilles Dubreuil proposed a change to stackforge/puppet-vswitch: Refacfored a more suitable ovs_redhat provider  https://review.openstack.org/94504
00:09 rmoe joined #puppet-openstack
00:12 cjellick joined #puppet-openstack
00:17 topshare joined #puppet-openstack
00:29 sputnik13 joined #puppet-openstack
00:39 aimon joined #puppet-openstack
00:50 gildub joined #puppet-openstack
01:31 aimon joined #puppet-openstack
01:32 xingchao joined #puppet-openstack
01:34 xarses joined #puppet-openstack
02:03 openstackgerrit joined #puppet-openstack
02:03 hogepodge joined #puppet-openstack
02:06 sputnik13 joined #puppet-openstack
02:25 topshare joined #puppet-openstack
02:54 topshare joined #puppet-openstack
03:03 thumpba_ joined #puppet-openstack
03:07 tdb joined #puppet-openstack
03:17 hogepodge joined #puppet-openstack
03:22 topshare joined #puppet-openstack
03:24 mmaglana joined #puppet-openstack
04:00 sanjayu joined #puppet-openstack
04:08 topshare joined #puppet-openstack
04:12 sputnik13 joined #puppet-openstack
04:18 openstackgerrit Ricardo Rocha proposed a change to stackforge/puppet-ceph: define ceph::rgw, ceph::rgw::apache.  https://review.openstack.org/96060
04:18 openstackgerrit Ricardo Rocha proposed a change to stackforge/puppet-ceph: add ceph-extras and fast-cgi repositories.  https://review.openstack.org/96065
05:27 topshare joined #puppet-openstack
05:32 sputnik13 joined #puppet-openstack
05:34 openstackgerrit Ricardo Rocha proposed a change to stackforge/puppet-ceph: define ceph::rgw::keystone.  https://review.openstack.org/117422
05:51 tfz joined #puppet-openstack
06:20 qba73 joined #puppet-openstack
06:42 ddmitriev1 joined #puppet-openstack
06:44 ddmitriev joined #puppet-openstack
07:08 giorgiodinapoli joined #puppet-openstack
07:14 EmilienM good morning
07:23 topshare joined #puppet-openstack
07:29 michchap EmilienM: morning
07:29 michchap EmilienM: https://github.com/puppetlabs/hiera/pull/210
07:35 topshare joined #puppet-openstack
07:56 topshare joined #puppet-openstack
08:20 mmagr joined #puppet-openstack
08:20 derekh joined #puppet-openstack
09:14 qba73 joined #puppet-openstack
09:14 mtl1 joined #puppet-openstack
09:17 fvollero EmilienM: good morning :)
09:18 EmilienM michchap: isn't what spredzy was trying to do somehow?
09:20 michchap EmilienM: Everybody is trying to do it. It means you can have one value like 'debug: false' and have other things like 'nova::debug: "%{alias('debug')}". If you use hiera() instead of alias() it will convert to a string, not a bool.
09:20 fvollero lol :)
09:21 fvollero everyone thought: challenge accepted
09:21 michchap EmilienM: same problem for a list of rabbit servers (array) or anything where you want to set something in one place and have it propagate to multiple places.
09:22 EmilienM michchap: http://paste.openstack.org/show/101578/
09:22 EmilienM I was about showing you this example
09:22 michchap fvollero: I have seen a few solutions. One is yaml anchors, which is what I use at the moment. Another is the data_mapper that bodepd wrote.
09:23 EmilienM michchap: spredzy did like this: https://github.com/enovance/edeploy-roles/blob/master/files/backend.rb.patch
09:23 michchap EmilienM: There was a similar proposal upstream but it was rejected because it changes behavior
09:23 EmilienM yeah
09:24 michchap https://github.com/puppetlabs/hiera/pull/188
09:24 michchap I really just took that patch and made it a new function instead of changing the old behavior.
09:34 gcha joined #puppet-openstack
10:03 michchap joined #puppet-openstack
10:09 openstackgerrit Lukas Bezdicka proposed a change to stackforge/puppet-neutron: Unescape value in parse_allocation_pool  https://review.openstack.org/110913
10:48 openstackgerrit A change was merged to stackforge/puppet-heat: Require Class['mysql::server'], not Service['mysqld'] when creating DB.  https://review.openstack.org/114308
11:24 openstackgerrit Richard Arends proposed a change to stackforge/puppet-ceph: Added hook 'ceph::profile::params::manage_repo'.  https://review.openstack.org/116465
11:35 mattymo joined #puppet-openstack
12:04 topshare joined #puppet-openstack
12:04 richm joined #puppet-openstack
12:09 markvoelker joined #puppet-openstack
12:11 dprince joined #puppet-openstack
12:14 sputnik13 joined #puppet-openstack
12:30 topshare joined #puppet-openstack
12:38 mmaglana joined #puppet-openstack
12:40 mmaglana_ joined #puppet-openstack
12:42 mmaglan__ joined #puppet-openstack
12:45 rharrison joined #puppet-openstack
13:13 qba73 joined #puppet-openstack
13:21 qba73 joined #puppet-openstack
13:25 qba73 joined #puppet-openstack
13:32 qba73 joined #puppet-openstack
13:37 openstackgerrit Mathieu Gagné proposed a change to stackforge/puppet-neutron: Add kombu_reconnect_delay  https://review.openstack.org/106116
13:37 qba73 joined #puppet-openstack
13:38 mmagr joined #puppet-openstack
13:40 mmaglana joined #puppet-openstack
13:49 xarses joined #puppet-openstack
13:57 openstackgerrit Colleen Murphy proposed a change to stackforge/puppet-keystone: Migrate mysql backend to use openstacklib::db::mysql  https://review.openstack.org/111075
13:59 pradk joined #puppet-openstack
14:01 crinkle Could we revisit 111036, 111037, 111072, 111073, 111074, 111075, 111076, and 111077? They have been sitting there for a while
14:28 michchap crinkle: morning :)
14:29 michchap crinkle: I will try to do a swathe of review tomorrow to recude the queue.
14:30 crinkle michchap: good morning
14:30 crinkle michchap: actually I think you've already reviewed most of these
14:30 crinkle just needs another approval
14:30 michchap crinkle: I have a favor to ask: could you ping someone appropriate about https://github.com/puppetlabs/hiera/pull/210? I'm sure it could be improved, but it's something that's been bugging me for a very long time. I think I've addressed the issue from the previous PR that was rejected.
14:31 crinkle michchap: I can try to get some eyes on it
14:35 crinkle michchap: you can also ask in #puppet-dev
14:36 michchap crinkle: thanks. Anyone in particular I should ping there?
14:36 EmilienM michchap: Hunner knows I think
14:37 mmaglana joined #puppet-openstack
14:42 s_albtraum joined #puppet-openstack
14:46 michchap EmilienM: Good idea. I'll see if I can get a hold of him in his afternoon.
14:52 cjellick joined #puppet-openstack
14:56 dmsimard dachary: having a hard time with gerritexec? :)
15:00 cjellick joined #puppet-openstack
15:13 mdorman joined #puppet-openstack
15:25 mmaglana joined #puppet-openstack
15:26 dmsimard dachary, xarses: Outstanding reviews for puppet-ceph (some very good ones). Please give me a hand :)
15:37 cjellick joined #puppet-openstack
15:40 sanjayu joined #puppet-openstack
15:47 markvoelker1 joined #puppet-openstack
15:48 sputnik13 joined #puppet-openstack
15:52 saju_m joined #puppet-openstack
16:10 mgagne won't get much merged this week, gate is backed up for hours...
16:36 mfa298 joined #puppet-openstack
16:53 Hunner michchap: I don't usually do core code (hiera, facter, puppet). That would probably best be run by finch
16:55 qba73 joined #puppet-openstack
16:57 aimon joined #puppet-openstack
17:04 rmoe joined #puppet-openstack
17:14 JayJ joined #puppet-openstack
17:19 JayJ I want to understand how puppet-modules are linked to Openstack modules. Issue is when I add a new compute node the puppet modules install the latest and the greatest form the stable branch while my other computes are running a few patches behind. Is there a way to control this? What's the best or suggested practice here?
17:23 openstackgerrit Richard Arends proposed a change to stackforge/puppet-ceph: Added hook 'ceph::profile::params::manage_repo'.  https://review.openstack.org/116465
17:24 Mosibi dmsimard: thx pointer the pointers in my change. In the last patch i forgot some files to include... my bad :(
17:24 Mosibi dmsimard: this one... https://review.openstack.org/116465
17:25 dmsimard Mosibi: Ah so you're Richard ? :)
17:25 Mosibi yes indeed :)
17:25 Mosibi blew my cover now ;)
17:27 dmsimard Yeah, looks good to me now. There's only one thing that wasn't addressed
17:27 dmsimard I added a comment on patch set 3 that the commit needed tests. You're touching something that is already spec/integration tested, see: https://github.com/stackforge/puppet-ceph/blob/master/spec/classes/ceph_profile_base_spec.rb / https://github.com/stackforge/puppet-ceph/blob/master/spec/system/ceph_profile_base_spec.rb
17:28 dmsimard We need to add a test that validates the behavior of manage_repo
17:28 Mosibi I understand...
17:28 Mosibi Can i get some help with that from someone, or a pointer to some documentation about test spec files?
17:29 dmsimard Sure, I can help you - let me know if you have any questions
17:29 dmsimard From reading the files I just linked, it should give you a rough idea of how the tests are laid out
17:30 Mosibi Okay. I will look into that..
17:31 dmsimard You basically need to test two behaviors, the default where manage_repo is true needs to validate that the class ceph::repo is in the catalog. If manage_repo is false, it needs to ensure that ceph::repo is not.
17:31 Mosibi That can't be hard to do i hope :)
17:31 dmsimard By default, the spec tests validates that ceph::repo is in the catalog: https://github.com/stackforge/puppet-ceph/blob/master/spec/classes/ceph_profile_base_spec.rb#L24
17:33 dmsimard So you need to do a test that if you set manage_repo to false, it doesn't contain ceph::repo
17:33 dmsimard You can set params this way: https://github.com/stackforge/puppet-ceph/blob/master/spec/classes/ceph_init_spec.rb#L23-L27
17:33 dmsimard Do you know how to run the tests locally on your machine ?
17:34 Mosibi nope. This is totally new to me.
17:34 cjellick joined #puppet-openstack
17:34 Mosibi ah....  describe "with custom params and specified fsid" do ...
17:34 Mosibi And there i can toggle with my var to 'false'
17:36 cjellick joined #puppet-openstack
17:36 tfz joined #puppet-openstack
17:40 shakamunyi joined #puppet-openstack
17:42 dmsimard Mosibi: What kind of system are you developing on ?
17:43 hogepodge joined #puppet-openstack
17:45 Mosibi dmsimard: I develop on my own (Fedora) laptop.
17:46 Mosibi I work with KVM to build test networks. My ceph hosts are virtual hosts, running Ubuntu 14.04
17:59 dmsimard Mosibi: Ok, you essentially need ruby and ruby's bundle and gem tools.
17:59 dmsimard Mosibi: After you've done what is here: https://github.com/stackforge/puppet-ceph#development
17:59 dmsimard You will be able to run "bundle exec rake lint" to run puppet-lint (syntax and standard checker)
18:00 dmsimard bundle exec rake spec for spec unit tests
18:00 dmsimard that was "bundle exec rake spec"
18:00 dmsimard for integration testing there are other dependencies listed here: https://github.com/stackforge/puppet-ceph#integration-tests
18:00 dmsimard The lint and spec tests are run by jenkins automatically when you push a commit
18:01 dmsimard The integration tests are run by an external machine, also automatically
18:02 cjellick_ joined #puppet-openstack
18:02 xarses joined #puppet-openstack
18:12 Mosibi dmsimard: i logged all your info and will look into it tommorow. Thanks for it!
18:12 dmsimard Mosibi: np :)
18:15 richm question about rspec-puppet and facts
18:16 richm If I have let :facts { :processorcount => 42 }
18:17 richm in an "it { should ... }" test, how do I refer to this value in a string?  e.g. "threads=#{facts[:processorcount]}"?
18:19 richm for example, I can refer to :params like this - "#{params[:foobar]}"
18:19 richm does :facts "work" the same way as :params?
18:27 rharrison joined #puppet-openstack
18:34 dmsimard richm: afaik, yes
18:36 dmsimard richm: Trying to find an example right now ..
18:55 richm dmsimard: ok - it is working
18:55 richm dmsimard: thanks
19:03 openstackgerrit Lars Kellogg-Stedman proposed a change to stackforge/puppet-nova: --vlan is not the same as --vlan_start  https://review.openstack.org/117592
19:06 openstackgerrit Lars Kellogg-Stedman proposed a change to stackforge/puppet-nova: --vlan is not the same as --vlan_start  https://review.openstack.org/117592
19:11 larsks That is a simple but annoying bug ^^^^
19:13 tdb joined #puppet-openstack
19:14 JayJ I want to understand how puppet-modules are linked to Openstack modules. Issue is when I add a new compute node the puppet modules install the latest and the greatest form the stable branch while my other computes are running a few patches behind. Is there a way to control this? What's the best or suggested practice here?
19:15 JayJ Anyone?
19:15 rharrison JayJ, sounds like you're just getting the latest packages from your distribution repositories. Its the default behavior of apt and yum.
19:17 JayJ rharrison: Yes, it gets the latest from the stable branch. How do I control what puppet scripts pull? Is there any literature on the topic?
19:18 larsks JayJ: I would ask, why are you installing the puppet modules on each system? If you're running a puppet master you only need them installed on the master, and if you're running them via "puppet apply" you maybe want to take care of deploying the modules yourself to ensure that all systems have the same version.
19:18 larsks JayJ: Or did I misunderstand your question?
19:18 larsks I think I did.
19:18 larsks Never mind.
19:19 rharrison JayJ, they'll pull the package that yum or apt sees as the latest. To control that you may want to mirror the repositories yourself and use the mirror instead of the upstream repos.  That way those repos will have the same version of the package until you update your mirror when you're ready to patch everything.
19:20 rharrison JayJ, puppet just calls the package manager to tell it to install package foo.  Your package manager is what's downloading and installing the packages.
19:20 JayJ larsks: I'm running from the puppet master. So Puppet master has the puppet modules. When I add a compute node, modules install openstack packages (for ex: nova-compute) which is pulled from the stable branch and happens to be latest from the branch.
19:20 larsks JayJ: Yeah, I got that on my second reading of the question :).
19:21 JayJ rharrison: Ah! I see. So I need to mirror the packages and that should take care of it. Thanks
19:22 rharrison JayJ, yeah and make sure you have your own puppet manifest that sets up the repository configs pointing at your mirror.
19:22 JayJ Makes sense. Thanks. I understand this at high level now. Implementing would be a different challenge :)
19:23 JayJ rharrison: Thanks.
19:24 rharrison JayJ, and keep in mind you'll have to be vigilant about keeping up with when your upstream repos are getting patches you know when you'll need to update your mirrors since they won't show up in 'yum check-update' any more.
19:33 openstackgerrit joined #puppet-openstack
20:01 qba73 joined #puppet-openstack
20:16 richm has anyone else noticed rspec test breakage since puppetlabs-apache converted vhost to use concat instead of file?
20:17 richm and, if you have fixed it, how did you change your with_content tests to work with concat fragments?
20:27 qba73 joined #puppet-openstack
20:32 JayJ joined #puppet-openstack
20:36 sputnik13 joined #puppet-openstack
20:40 mmaglana joined #puppet-openstack
21:06 cjellick joined #puppet-openstack
21:14 cjellick_ joined #puppet-openstack
21:43 ddmitriev joined #puppet-openstack
21:44 mmaglana_ joined #puppet-openstack
21:45 ashp left #puppet-openstack
22:32 daneyon joined #puppet-openstack
22:49 daneyon joined #puppet-openstack
22:53 xavpaice do people here have an opinion about the use of package_ensure?  I note in several places it's set to 'installed' but elsewhere it's set to $package_ensure - in several modules
22:54 * xavpaice is happy to submit changes to make them all honor a variable, but if that's not what people want I'll work around it in other ways
22:59 daneyon joined #puppet-openstack
23:05 aimon joined #puppet-openstack
23:24 openstackgerrit Ricardo Rocha proposed a change to stackforge/puppet-ceph: add ceph-extras and fast-cgi repositories.  https://review.openstack.org/96065
23:30 topshare joined #puppet-openstack
23:46 openstackgerrit Colleen Murphy proposed a change to stackforge/puppet-openstacklib: (WIP) Implement base aviator provider  https://review.openstack.org/116753
23:48 openstackgerrit joined #puppet-openstack
23:52 thumpba_ joined #puppet-openstack

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