Perl 6 - the future is here, just unevenly distributed

IRC log for #puppet-openstack, 2014-04-21

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

All times shown according to UTC.

Time Nick Message
00:05 kitp joined #puppet-openstack
00:32 thuc joined #puppet-openstack
00:33 thuc_ joined #puppet-openstack
00:47 marun joined #puppet-openstack
01:06 kitp joined #puppet-openstack
01:11 xarses joined #puppet-openstack
01:57 xingchao_ joined #puppet-openstack
02:07 kitp joined #puppet-openstack
02:38 starmer joined #puppet-openstack
03:08 kitp joined #puppet-openstack
03:28 starmer joined #puppet-openstack
04:05 starmer joined #puppet-openstack
04:08 kitp joined #puppet-openstack
05:09 kitp joined #puppet-openstack
05:17 sanjayu joined #puppet-openstack
06:10 kitp joined #puppet-openstack
07:11 kitp joined #puppet-openstack
07:20 xingchao joined #puppet-openstack
07:33 sanjayu joined #puppet-openstack
07:33 sanjayu_ joined #puppet-openstack
07:52 tfz joined #puppet-openstack
08:09 sanjayu_ joined #puppet-openstack
08:11 kitp joined #puppet-openstack
08:22 sanjayu joined #puppet-openstack
08:26 xingchao_ joined #puppet-openstack
08:34 sanjayu joined #puppet-openstack
09:12 kitp joined #puppet-openstack
10:13 kitp joined #puppet-openstack
10:45 dmsimard joined #puppet-openstack
11:08 tfz joined #puppet-openstack
11:14 kitp joined #puppet-openstack
11:35 csschwe joined #puppet-openstack
11:56 sanjayu joined #puppet-openstack
12:14 kitp joined #puppet-openstack
12:16 m0ikz joined #puppet-openstack
12:34 bcrochet left #puppet-openstack
12:49 m0ikz joined #puppet-openstack
13:06 daneyon joined #puppet-openstack
13:06 morazi joined #puppet-openstack
13:07 daneyon joined #puppet-openstack
13:15 kitp joined #puppet-openstack
13:28 kkirkpatrick joined #puppet-openstack
13:28 Pax joined #puppet-openstack
13:48 tfz joined #puppet-openstack
13:49 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: ceph::key : update copyright notice  https://review.openstack.org/89349
13:49 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: ceph::key : remove redundant Exec blocks  https://review.openstack.org/89350
13:49 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: ceph::mon : disable client.admin key generation  https://review.openstack.org/89351
13:49 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: ceph::key : remove the client.admin keyring for tests  https://review.openstack.org/89352
13:49 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: ceph::key : reduce literal duplication in tests  https://review.openstack.org/89353
13:49 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: ceph::key : proofread documentation strings  https://review.openstack.org/89354
13:49 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: ceph::key : tests that shows how to create client.admin  https://review.openstack.org/89355
13:49 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: update the first use case to reflect that it works  https://review.openstack.org/89356
14:03 rharrison joined #puppet-openstack
14:04 kitp joined #puppet-openstack
14:11 rwsu joined #puppet-openstack
14:32 andi- joined #puppet-openstack
14:32 andi- joined #puppet-openstack
14:55 kkirkpatrick joined #puppet-openstack
14:58 Pax joined #puppet-openstack
15:17 Rajini_Ram joined #puppet-openstack
15:17 xarses joined #puppet-openstack
15:28 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: update the first use case to reflect that it works  https://review.openstack.org/89356
15:28 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: ceph::key : proofread documentation strings  https://review.openstack.org/89354
15:28 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: ceph::key : remove the client.admin keyring for tests  https://review.openstack.org/89352
15:28 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: ceph::key : reduce literal duplication in tests  https://review.openstack.org/89353
15:28 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: ceph::mon : control client.admin key generation  https://review.openstack.org/89351
15:30 kitp joined #puppet-openstack
15:35 xarses dachary: you squishing most // all of these?
15:36 dachary xarses: I think they are ready for review. One of needed squashing because it failed integration tests.
15:36 xarses ok
15:39 dachary Hopefully each of the review is small enough and you can give it a look when you have a few minutes. That's my rationale behind splitting them anyways ;-)
15:39 dachary xarses: ^
15:39 dachary s/behind/for/
15:40 xarses ={
15:40 xarses erm
15:40 xarses =P
15:41 dachary :-)
15:42 openstackgerrit Lars Kellogg-Stedman proposed a change to stackforge/puppet-neutron: Neutron Nova Interactions Support  https://review.openstack.org/88410
15:45 markvoelker larsks: I came to the same conclusion over the weekend. =)
15:45 larsks markvoelker: Yeah.  I remember running into something like this before.  Darn puppet.  Anyway, this seems to make things work.
15:46 larsks Would love to know if it works for you, too.
15:46 markvoelker I'll test it over lunch and let you know
15:46 markvoelker looks more promising
15:46 larsks Awesome.
15:47 kitp joined #puppet-openstack
15:51 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: ENV['RELEASES'] overrides the [cuttlefish dumpling emperor]  https://review.openstack.org/89210
15:59 rmoe joined #puppet-openstack
16:11 openstackgerrit Lars Kellogg-Stedman proposed a change to stackforge/puppet-neutron: Neutron Nova Interactions Support  https://review.openstack.org/88410
16:12 marun joined #puppet-openstack
16:12 kaleng joined #puppet-openstack
16:14 Pax joined #puppet-openstack
16:27 kitp joined #puppet-openstack
16:28 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: ceph::key : remove the client.admin keyring for tests  https://review.openstack.org/89352
16:28 JonnyNomad joined #puppet-openstack
16:37 Aimon joined #puppet-openstack
16:39 kitp joined #puppet-openstack
16:53 starmer joined #puppet-openstack
16:58 rharrison joined #puppet-openstack
17:21 markvoelker larsks: was able to take that patch for a spin in a small (all in one) environment over lunch and it worked!  Will test in a bigger env and review the code later this afternoon.
17:21 larsks \o/
17:22 larsks markvoelker: Do you have any thoughts on making a keystone_user resource type available when running rspec?
17:22 larsks Not sure the best way of doing that.
17:23 markvoelker larsks: haven't looked at that bit just yet.  The issue is the cross-module dependency, or...?
17:23 larsks Yeah.  Inside manifests/server/notifications.pp is a dependency on Keystore_user, but when rspec runs that resource type isn't defined (because it comes from puppet-keystone).
17:23 larsks I guess I need to mock it out somehow.  Looking into that right now.
17:24 markvoelker Got it.  Yeah, I think mockin it out is the right thing to do.
17:24 larsks markvoelker: I think this has the answer: https://groups.google.com/forum/​#!topic/puppet-users/F21aarcJsWw
17:25 * markvoelker looks
17:45 larsks markvoelker: have a second?
17:46 markvoelker larsks: good timing. Just left a meeting.  =)
17:46 markvoelker What's up?
17:46 larsks I am confused by rspec.  As far as I can tell, the existing tests are not actually calling the provider.
17:47 larsks in spec/classes/neutron_server_notifications.rb
17:48 larsks ...which is making it hard to test things :/
17:48 Rajini_Ram joined #puppet-openstack
17:49 markvoelker hmm....
17:49 * markvoelker looks
17:51 markvoelker What happens if you explicitly set the admin tenant id to undef (say, in the default params list round line 45)?
17:53 * larsks looks.
17:54 larsks markvoelker: Nothing.
17:54 larsks Just to make sure, I've commented out all the stub_reqest() calls, and nothing is ever trying to perform any keystone lookups.
17:59 markvoelker larsks: let me set up an environment where I can play around with it a bit...
17:59 markvoelker Nothing's immediately jumping out
18:00 angdraug joined #puppet-openstack
18:00 larsks Hmm.
18:01 larsks Rspec is still a black box to me.
18:02 larsks Well, I will go acquire food and see if it makes more sense when I return.
18:05 xarses dachary: comments on https://review.openstack.org/#/c/89356/ and https://review.openstack.org/#/c/89268/
18:09 dachary xarses: thanks, I'll work on them :-)
18:10 xarses dachary: also, I'd like to get https://review.openstack.org/#/c/80710/ to work on CI, but it looks like it doesn't load the bundle version of puppet-lint. Can I fix this from our side or does the CI server need to be updated for this?
18:22 dachary GEM_HOME=~/.gems gerritexec --hostname review.openstack.org            --verbose            --username puppetceph            --script 'bash -c "mv Gemfile-rspec-system Gemfile ; bundle install ; bundle exec rake spec:system" > /tmp/out$$ 2>&1 ; r=$? ; pastebinit /tmp/out$$ ; rm -f /tmp/out$$ ; exit $r #'            --project stackforge/puppet-ceph
18:23 dachary xarses: and these http://paste.ubuntu.com/7301208/ are the ruby packages installed
18:28 sputnik1_ joined #puppet-openstack
18:30 dachary what are the requirements ?
18:31 xarses that it uses the checked out version of puppet-lint
18:40 xarses dachary: All, I can come up with is that gate-puppet-ceph-puppet-lint doesn't actually use builder
18:40 xarses http://logs.openstack.org/10/80710/2/check/gate-​puppet-ceph-puppet-lint/bf2253c/console.html.gz
18:45 starmer joined #puppet-openstack
18:45 dachary ah, right, I thought you were referring to the integration checks ;-)
18:46 xarses yep, sorry thats my bad. was poorly worded
18:46 dachary I've never tried to fix the gates
19:08 kkirkpatrick joined #puppet-openstack
19:12 larsks I am looking for some pointers w/r/t rspec tests and custom types/providers; any takers?
19:13 xarses hmm, we'd need something like https://review.openstack.org/#/c/82844
19:14 xarses larsks: most of the stackforge/puppet-* modules have tests for custom types/providers notably puppet-neutron and puppet-keystone
19:15 xarses I also wrote one for ini_file extensions used in puppet-ceph
19:15 larsks xarses: I am actually writing a test for puppet-neutron, and I appear to be unable to get the test to actually use the custom provider.  Even though a test for contains_custom_resource_type(...) succeeds.
19:16 larsks xarses: Would you mind glancing at a simple test?
19:17 xarses I can eye it, but I'm a bit stale on the subject
19:17 larsks Here's a fairly minimal test that demonstrates the problem: https://github.com/larsks/puppet-neutron/blob/work​/spec/classes/neutron_server_notifications_spec.rb
19:18 larsks There is a custom type "nova_admin_tenant_id_setter" that ultimately should result in the provider attempting to contact a keystone server, but the provider methods don't appear to be getting called.
19:18 larsks It works in practice, just not in the test environment.
19:20 starmer joined #puppet-openstack
19:21 angdraug joined #puppet-openstack
19:28 xarses where is "nova_admin_tenant_id_setter" coming from?
19:28 larsks xarses: lib/puppet/type/nova_admin_tenant_id_setter.rb in that module.
19:28 larsks https://github.com/larsks/puppet-neutron/blob/wor​k/lib/puppet/type/nova_admin_tenant_id_setter.rb
19:29 larsks And the provider is: https://github.com/larsks/puppet-neutr​on/blob/work/lib/puppet/provider/nova_​admin_tenant_id_setter/ini_setting.rb
19:46 xarses I would first test that nova_admin_tenant_id false/undef and go from there. I also don't see why you need that define, there appears to be no relationship
19:50 xarses also i'm not sure if it matters, but your provider doesn't need to be named ini_file.rb
19:50 larsks xarses: My new theory is that when testing classes with rspec, resources aren't actually realized.  That is, file{} resource don't actually create files, for example.
19:50 larsks xarses: That would completely explain this behavior,.
19:51 larsks xarses: Which define were you referring to?
19:51 xarses larsks: no files will not be created by rspec-puppet, you can force rspec or use rspec-system to do that
19:52 xarses https://github.com/larsks/puppet-ne​utron/blob/work/spec/classes/neutro​n_server_notifications_spec.rb#L7
19:52 larsks xarses: Right, but I think it is the same reason that the provider methods are not being called.  No?
19:52 larsks xarses: That define is required because of the following in manifests/server/notications.pp:
19:53 larsks Keystone_user <| title == 'nova' |> -> Class[neutron::server::notifications]
19:53 xarses larsks: it should still contain the refrence to the provider and shot its data
19:53 xarses larsks: thats just a collector, it should still work fine with out any objects to collect
19:54 larsks No, you get an "unknown resource type" error if it doesn't know about the resource type.
19:54 xarses hrm, ok
19:54 larsks "Resource type keystone_user doesn't exist on node ..."
19:54 xarses if you look at how we test the _config providers
19:54 xarses you can see that they are realized enough to see them working
19:55 larsks I'll take a closer look at those, I guess.  I have tried this with admin_tenant_id explicitly not defined. That "notify" resource in the manifest was there to verify which code path it was taking (with a "contains_notify('lookup')" condition).
19:56 xarses example of a custom provder https://github.com/stackforge/puppet-ceph/blob​/master/spec/classes/ceph_repo_spec.rb#L95-100
19:56 xarses example of ini_file custom provider testing https://github.com/stackforge/puppet-ceph/bl​ob/master/spec/classes/ceph_init_spec.rb#L45
19:57 larsks xarses: The thing is, I'm reasonably sure that those don't actually require the provider code to execute.
19:57 larsks I think that contain_<resource>() really only checks that the resource would be defined, but does not actually execute the provider.  That's my experience so far, at least.
19:58 xarses larsks: correct, if you want to exercise the provider you need to create unit tests for the provider and perform them in rspec, not rspec-puppet
19:59 larsks xarses: Okay.  Well, that would explain all of my questions, then.
19:59 xarses it rspec puppet will test if the item is in the manifest and what data was sent to it
19:59 larsks xarses: Okay.  So, separate tests to exercise provider code.
19:59 xarses see https://github.com/stackforge/pu​ppet-ceph/tree/master/spec/unit there are some examples of how i tested the actual ini_file provider
20:00 larsks xarses: Awesome, I will take a look.
20:00 xarses back shortly
20:02 Mr_D joined #puppet-openstack
20:02 xarses_ joined #puppet-openstack
20:04 xarses_ back
20:05 Mr_D hey guys, I'm working on a spec test for a new backend that I'm adding puppet-cinder, and am getting an error that indicates that it can't find the first new parameter in cinder.conf
20:05 Mr_D I'm modeling my spec test after cinder_backend_solidfire_spec.rb
20:06 Mr_D and one thing I don't understand is where the solidfire backend is actually being called into
20:07 Mr_D I'm guessing that my new spec test is not actually calling my new backend, but from all of the other backends, don't see anything that explicitly does this
20:07 Mr_D can anyone offer any words of wisdom?
20:17 xarses_ Mr_D: do you have a copy of the code to look over?
20:18 Mr_D 1 sec - will put it on pastebin...
20:19 Mr_D Here's the spec test: http://pastebin.com/DZqJ7rYK
20:20 Mr_D Here's the backend: http://pastebin.com/z4GuWbnf
20:21 Mr_D Here's the error: http://pastebin.com/rrGv0gGr
20:21 daneyon joined #puppet-openstack
20:23 Mr_D It must be something really obvious that I'm just overlooking...
20:38 social joined #puppet-openstack
20:50 xarses_ Mr_D: yep, got me confused to
20:50 xarses_ can't run it myself either
20:51 Mr_D well, good to know I'm not insane!
20:51 xarses_ I also simplified some of the contains
20:51 Mr_D I stuck a notify in the define, and don't see that being dumped out when I run the test
20:51 xarses_ no love, maybe eqlx isn't running
20:52 xarses_ notify wont show in rake spec
20:52 Mr_D yeah, that was my conclusion...
20:52 xarses_ you can should contain_notify ...
20:52 xarses_ but you wont see the message its self
20:52 Mr_D ah - thx
20:54 xarses_ maybe run the manifest its self somewhere to observe the result. I'd guess it isn't running
20:54 xarses_ but the spec should help us with that
20:56 andreww joined #puppet-openstack
21:00 Mr_D gack!  Well, running the manifest on my box answered the question: Warning: Could not retrieve fact fqdn
21:01 Mr_D was expecting the test framework to spit out any errors, but guess this is one case where it will not
21:01 Mr_D thx very much for your help xarses_!
21:02 xarses np
21:03 xarses hmm, thats odd, you dont ask for fqdn in your manifest
21:04 Mr_D no - i've been doing live testing on a different system that has the fqdn defined
21:08 prad_ joined #puppet-openstack
21:17 rharrison joined #puppet-openstack
22:06 openstackgerrit Lars Kellogg-Stedman proposed a change to stackforge/puppet-neutron: Neutron Nova Interactions Support  https://review.openstack.org/88410
22:35 JonnyNomad left #puppet-openstack
22:43 starmer joined #puppet-openstack
23:12 m0ikz joined #puppet-openstack
23:50 openstackgerrit joined #puppet-openstack

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