Perl 6 - the future is here, just unevenly distributed

IRC log for #puppet-openstack, 2014-11-18

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

All times shown according to UTC.

Time Nick Message
00:11 sputnik13 joined #puppet-openstack
00:49 sputnik13 joined #puppet-openstack
01:05 sputnik13 joined #puppet-openstack
01:09 gildub joined #puppet-openstack
01:23 topshare joined #puppet-openstack
01:30 EmilienM dmsimard_away: checking
01:31 openstackgerrit Emilien Macchi proposed stackforge/puppet-nova: Service Validation for Nova-API  https://review.openstack.org/126458
01:38 rmoe joined #puppet-openstack
01:38 openstackgerrit Emilien Macchi proposed stackforge/puppet-nova: Update [neutron] section in nova.conf for Juno  https://review.openstack.org/131566
01:39 hogepodge joined #puppet-openstack
01:44 sputnik13 joined #puppet-openstack
01:50 xingchao joined #puppet-openstack
01:54 topshare joined #puppet-openstack
02:03 xingchao_ joined #puppet-openstack
02:17 RichardRaseley joined #puppet-openstack
02:33 xingchao joined #puppet-openstack
02:39 sputnik13 joined #puppet-openstack
02:40 topshare joined #puppet-openstack
03:05 EmilienM dmsimard_away: I don't see any diff
03:06 EmilienM our puppet-nova is already ready for RBD in Juno
03:11 sputnik13 joined #puppet-openstack
03:21 openstackgerrit Richard Megginson proposed stackforge/puppet-keystone: support Keystone with read-only LDAP identity backend  https://review.openstack.org/133601
03:24 openstackgerrit Richard Megginson proposed stackforge/puppet-keystone: support the ldap user_enabled_invert parameter  https://review.openstack.org/129951
03:35 openstackgerrit Richard Megginson proposed stackforge/puppet-keystone: support Keystone with read-only LDAP identity backend  https://review.openstack.org/133601
03:44 xarses joined #puppet-openstack
04:14 subscope joined #puppet-openstack
04:48 openstackgerrit Mike Dorman proposed stackforge/puppet-neutron: Implement better nova_admin_tenant_id_setter exists? method  https://review.openstack.org/134667
04:58 xingchao joined #puppet-openstack
05:25 sputnik13 joined #puppet-openstack
05:31 openstackgerrit Merged stackforge/puppet-nova: Update [glance] section in nova.conf for Juno  https://review.openstack.org/131572
05:39 openstackgerrit Mike Dorman proposed stackforge/puppet-neutron: Implement better nova_admin_tenant_id_setter exists? method  https://review.openstack.org/134667
05:47 sputnik13 joined #puppet-openstack
05:50 mschmitt joined #puppet-openstack
05:52 sputnik13 joined #puppet-openstack
05:58 openstackgerrit Merged stackforge/puppet-openstacklib: Implement openstacklib::messaging::rabbitmq  https://review.openstack.org/134522
06:07 sputnik1_ joined #puppet-openstack
06:12 sputnik13 joined #puppet-openstack
06:19 topshare joined #puppet-openstack
06:25 topshare joined #puppet-openstack
06:57 openstackgerrit Gilles Dubreuil proposed stackforge/puppet-vswitch: Fixed issue with vlan on redhat provider  https://review.openstack.org/135180
07:10 sputnik13 joined #puppet-openstack
07:10 stamak joined #puppet-openstack
07:44 topshare joined #puppet-openstack
07:48 sputnik13 joined #puppet-openstack
08:20 openstackgerrit Malik Bougacha proposed stackforge/puppet-nova: supporting lxc cpu mode The default cpu_mode with the lxc libvirt type doesn't support the cpu_mode configuration. If one is defined, the nova-compute service will fail.  https://review.openstack.org/133987
08:22 tlbr_ joined #puppet-openstack
08:27 patcable joined #puppet-openstack
08:30 arnaud_orange joined #puppet-openstack
08:30 slashme joined #puppet-openstack
08:40 jawed joined #puppet-openstack
08:41 henditan joined #puppet-openstack
08:42 openstackgerrit Sebastien Badia proposed stackforge/puppet-heat: Remove deprecated api-cfn and api-cloudwatch classes  https://review.openstack.org/135191
08:44 xingchao joined #puppet-openstack
08:54 openstackgerrit Malik Bougacha proposed stackforge/puppet-nova: supporting lxc cpu mode The default cpu_mode with the lxc libvirt type doesn't support the cpu_mode configuration. If one is defined, the nova-compute service will fail.  https://review.openstack.org/133987
08:57 rcallawa joined #puppet-openstack
09:03 giorgiodinapoli joined #puppet-openstack
09:03 derekh joined #puppet-openstack
09:04 rcallawa joined #puppet-openstack
09:06 imcsk8 joined #puppet-openstack
09:20 dgurtner joined #puppet-openstack
09:21 topshare joined #puppet-openstack
09:44 ilbot3 joined #puppet-openstack
09:44 Topic for #puppet-openstack is now Place to collaborate on Puppet/OpenStack tools: logs at http://irclog.perlgeek.de/puppet-openstack/today
09:44 rcallawa joined #puppet-openstack
09:46 tlbr joined #puppet-openstack
09:49 todin Hi, why does the puppet manifiest install the ovs on the contoller node? I think there the ovs is not nessasary.
10:00 qba73 joined #puppet-openstack
10:04 rcallawa joined #puppet-openstack
10:12 topshare joined #puppet-openstack
10:15 admin0 joined #puppet-openstack
11:16 rcallawa joined #puppet-openstack
11:16 rcallawa joined #puppet-openstack
11:21 xingchao joined #puppet-openstack
11:34 aarefiev joined #puppet-openstack
11:37 aarefiev joined #puppet-openstack
11:42 stamak joined #puppet-openstack
11:59 EmilienM good morning!
12:11 igajsin joined #puppet-openstack
12:20 EmilienM todin: could you be more precise and explain which module/classes are you using?
12:20 EmilienM todin: because you don't have to install OVS on controller nodes if not needed, indeed
12:21 rcallawa joined #puppet-openstack
12:42 aarefiev joined #puppet-openstack
12:47 vb morning :)
13:00 asilenkov joined #puppet-openstack
13:07 jawed joined #puppet-openstack
13:08 rcallawa joined #puppet-openstack
13:18 rcallawa joined #puppet-openstack
13:31 markvoelker joined #puppet-openstack
13:35 morazi joined #puppet-openstack
13:44 topshare joined #puppet-openstack
14:14 mmagr joined #puppet-openstack
14:18 sputnik13 joined #puppet-openstack
14:24 rushiagr hello folks
14:24 rushiagr I'm having a problem with glance
14:25 rushiagr I'm deploying only keystone and glance via puppet, but it seems there is some problem with qpid
14:25 rushiagr AMQP server on <IP> is unreachable it says
14:27 rushiagr it seems it requires rabbit, but rabbit is not listed as dependency in glance in the Modulefile
14:27 rushiagr is this a bug? or am i doing something wrongly?
14:33 dmsimard EmilienM: You're right about rbd.. I'm tired T_T
14:36 dmsimard dachary: ping
14:38 dachary dmsimard: pong
14:38 trhoden joined #puppet-openstack
14:38 dmsimard dachary: Check line 11623 (amongst other things) http://paste.ubuntu.com/8976650/
14:48 richm1 joined #puppet-openstack
14:49 * dachary checking
14:50 dachary dmsimard: not familiar with this error Error: Cannot retrieve metalink for repository: ext-epel-6.8.
14:51 dachary dmsimard: do you still need the old machine or can it be torn to pieces ?
14:51 dmsimard dachary: Perhaps we can leave it up in case I need to refer to it. I'm going to spend some time on this today. Hoping to fix stuff.
14:52 dachary ok. Is it ok if we schedule it to be destroyed on friday ?
14:52 dachary dmsimard: ^
14:53 dmsimard dachary: Sure.
14:54 dachary cool
15:01 asilenkov joined #puppet-openstack
15:04 openstackgerrit joined #puppet-openstack
15:06 topshare joined #puppet-openstack
15:12 topshare joined #puppet-openstack
15:13 topshare joined #puppet-openstack
15:33 jawed joined #puppet-openstack
15:45 ericpeterson joined #puppet-openstack
15:47 ibravo joined #puppet-openstack
15:47 tfz joined #puppet-openstack
15:57 dmsimard dachary: Running tests manually right now (without gerrit). Going to see what happens.
15:57 dmsimard dachary: Interestingly enough I found https://bugs.launchpad.net/puppet-ceph/+bug/1264959
15:59 dmsimard dachary: There has also been two commits directly to github that haven't been done upstream (yet) about rhel stuff: https://github.com/pdrakeweb/puppet-ceph/commit/e24e2c298dcb0261789fa0006c22d4f21816a018 and https://github.com/stackforge/puppet-ceph/commit/7ebaf527b3968c716de4fa5989868bfff5cc4537
16:05 patcable joined #puppet-openstack
16:09 rcallawa joined #puppet-openstack
16:19 mdorman joined #puppet-openstack
16:27 rcallawa joined #puppet-openstack
16:29 dgurtner joined #puppet-openstack
16:29 dgurtner joined #puppet-openstack
16:33 dachary :-/
16:36 ericpeterson joined #puppet-openstack
16:38 dtalton joined #puppet-openstack
16:53 stupidnic joined #puppet-openstack
16:54 sputnik13 joined #puppet-openstack
16:56 stupidnic I am getting an error on keystone_user_role definitions. Unknown Attribute: auth_user_id. Looking at the command I can see that the --user-id switch is missing the user-id (it is blank). Can any one assist me in where I should be looking to correct this issue? I don't think it is a problem with my ECN, since some of the errors are with the services tenant
16:57 stupidnic s/ECN/ENC
17:06 vinsh joined #puppet-openstack
17:08 dmsimard dachary, xarses: I'm going to be pushing stuff in puppet-ceph to fix the integration tests. Would appreciate reviews asap.
17:11 openstackgerrit David Moreau Simard proposed stackforge/puppet-ceph: Remove templatedir directive from puppet.conf  https://review.openstack.org/135362
17:15 mschmitt joined #puppet-openstack
17:18 xarses dmsimard: will do
17:18 arnaud_orange left #puppet-openstack
17:21 rmoe joined #puppet-openstack
17:39 ddmitriev1 joined #puppet-openstack
17:41 xarses joined #puppet-openstack
17:45 xarses dmsimard: are you having problems geting 'bundle exec rake spec' to run?
17:46 todin EmilienM: I use the puppetlabs-openstack module in a multinode setup.
17:47 xarses dmsimard: http://paste.openstack.org/show/134463/
17:57 mmaglana joined #puppet-openstack
17:57 dmsimard xarses: You need to do bundle install with the "original" Gemfile. To run integration tests (spec:system) you need to move Gemfile-rspec-system to Gemfile and then install/use that bundle
17:58 dmsimard I haven't found a way to make spec:system work on OSX Yosemite yet.. can't install the gem due to libxml and other development dependency stuff so doing my tests under Ubuntu right now.
18:03 mgagne joined #puppet-openstack
18:04 xingchao joined #puppet-openstack
18:05 sputnik13 joined #puppet-openstack
18:08 xarses dmsimard: original gemfile?
18:08 dmsimard xarses: https://github.com/stackforge/puppet-ceph/blob/master/Gemfile
18:09 dmsimard xarses: https://github.com/stackforge/puppet-ceph/blob/master/Gemfile-rspec-system
18:09 xarses ah, ok
18:09 xarses so ya, I have that gem in place, and bundle install pretends to be happy
18:10 xarses and I get that hiera_config error
18:11 mgagne anyone else having issues with latest version of puppetlabs-mongodb? I'm getting Error: Could not find scope for mongodb::params
18:12 dmsimard xarses: Weird, this is what I get: http://paste.openstack.org/show/134471/
18:13 sputnik13 joined #puppet-openstack
18:17 openstackgerrit Bradley Klein proposed stackforge/puppet-monasca: Refresh stackforge repo  https://review.openstack.org/135384
18:17 dmsimard dachary: There's 48 integration tests broken due to a lack of 'config.yaml' file (Error: Could not run: config.yaml must exist). Does that ring you a bell ? http://paste.ubuntu.com/9075762/
18:18 dmsimard dachary: The only place I ever see it mentioned is here: https://github.com/stackforge/puppet-ceph/blob/master/spec/system/ceph_osds_spec.rb#L31
18:20 dmsimard Looks scenario_node_terminus related ..
18:24 dachary dmsimard: yes, it's most probably related to the *local* installation
18:25 dachary dmsimard: do you have a git clone of scenario_node_terminus in the right place ?
18:25 dmsimard dachary: Yeah. It's expecting scenario_node_terminus to be in "../"
18:25 dachary and it is
18:26 dachary right ?
18:26 dmsimard dachary: Yeah.
18:26 dachary something else then
18:26 dachary it's within the VM itself ?
18:26 dmsimard dachary: Well as far as I can understand, the integration tests use scenario_node_terminus which depends on a config.yaml file to determine which scenario to run.
18:27 sputnik13 joined #puppet-openstack
18:27 dachary if it's in the VM I don't know why it would fail. Since it's re-created from scratch on every run anyway.
18:28 dmsimard dachary: And that config.yaml file is only ever created/written in the osds spec here https://github.com/stackforge/puppet-ceph/blob/master/spec/system/ceph_osds_spec.rb#L63-L67
18:28 dachary scenario_node_terminus/data/config.yaml
18:28 dachary yes
18:29 dmsimard dachary: There's no data folder in the scenario_node_terminus repo. https://github.com/bodepd/scenario_node_terminus
18:29 dachary nevermind, it's in the fixture directory of puppet-ceph
18:29 dachary not relevant
18:29 dmsimard dachary: We create/delete the data folder in /etc/puppet
18:29 dmsimard and write the config.yaml file in it
18:29 dachary I bet it's a side effect of a previous error
18:30 dmsimard hmm
18:30 dmsimard I can flush and start again
18:30 dmsimard And see what happens ..
18:30 dachary flush ?
18:30 dachary destroy the VMs you mean ?
18:31 dmsimard well, I got rid of the repo, the fixtures, the bundles, the vms, everything :)
18:32 sputnik13 joined #puppet-openstack
18:33 mdorman looking for reviews on https://review.openstack.org/#/c/134667/ please.  â€œImplement better nova_admin_tenant_id_setter exists? method”
18:41 openstackgerrit Mike Dorman proposed stackforge/puppet-nova: Add Nova Aggregate support  https://review.openstack.org/70344
18:41 ericpeterson joined #puppet-openstack
18:49 xavpaice mdorman: I'll test it today, just wondering if you found a way to remove the rubygems requirement in that provider?
18:53 mdorman to be honest i didn’t even look at that
18:53 xarses mdorman: I'm wondering if there should be some more tests that show that the exists? methods work as expected
18:53 dmsimard sbadia: What's your general feeling with puppet-designate ? Is it usable ?
18:53 xarses dmsimard: that broken?
18:53 mdorman xarses:  yeah i could probably add some more tests
18:53 dmsimard xarses: What ?
18:54 ericpeterson joined #puppet-openstack
18:54 xarses dmsimard: sorry, confused a subject
18:58 f10d4 joined #puppet-openstack
18:58 f10d4 ohai
19:07 vinsh joined #puppet-openstack
19:13 dmsimard dachary: It looks like all tests are passing.. then the scenario_node_terminus tests run and all the tests after that fail
19:14 openstackgerrit Mike Dorman proposed stackforge/puppet-nova: Add Nova Aggregate support  https://review.openstack.org/70344
19:15 dmsimard dachary: Going to check live as tests are run.. maybe the config isn't being cleaned properly after the scenario tests
19:17 photocyte joined #puppet-openstack
19:23 sputnik13 joined #puppet-openstack
19:36 tfz joined #puppet-openstack
19:37 dachary hum
19:37 dachary dmsimard: what I find odd is that it previously worked
19:38 dachary but
19:38 dmsimard dachary: I confirm the issue
19:38 dachary removing /etc/puppet/data is not exactly a good idea
19:38 dmsimard dachary: it deletes /etc/puppet/data but doesn't remove the scenario setting from /etc/puppet/puppet.conf
19:39 dachary how could it every succeed in the past ? that eludes me.
19:39 dmsimard I kind of hurried up and removed the scenario setting from puppet.conf as the tests ran and the tests started passing again
19:39 dachary https://github.com/stackforge/puppet-ceph/blob/master/spec/system/ceph_osds_spec.rb#L110 is guilty for sure
19:40 dmsimard dachary: yeah but the setting is expected to be removed here: https://github.com/stackforge/puppet-ceph/blob/master/spec/system/ceph_osds_spec.rb#L85-L94
19:40 dmsimard dachary: but it doesn't look like that works
19:40 dachary oh !
19:42 dachary dmsimard: it does not remove the file and shows nothing in the output ?
19:43 dmsimard dachary: /etc/puppet/data is deleted successfully. What fails is the removal of the config line "node_terminus = scenario" from puppet.conf
19:43 dmsimard since "node_terminus = scenario" remains, it keeps trying to find the config.yaml file
19:44 dmsimard I'll run another test with a dirty hack in that'll make sure that line is absent and check what happens
19:44 dachary and there is no trace of the attempt to remove in the logs ?
19:44 dmsimard dachary: hard to tell, there's 10k lines to search in
19:44 dmsimard dachary: http://paste.ubuntu.com/9075762/
19:45 dmsimard dachary: Around line 5556 it seems to do it successfully..
19:45 xingchao joined #puppet-openstack
19:50 dachary that does not make sense to me
19:51 dachary 5566 it looks like it applies what's in 5556 but the output that follows is about removing the mons...
19:52 dachary I forgot how difficult it is to make sense of the puppet debug outout
19:53 dachary line 5597 it fails for the first time but that's because it removes data before applying the script
19:54 dachary which does not make sense because it's the other way around in https://github.com/stackforge/puppet-ceph/blob/master/spec/system/ceph_osds_spec.rb#L96
19:58 sputnik13 joined #puppet-openstack
20:00 dachary dmsimard: there must be something different on the local machine that triggers this problem. It is on the local machine that the https://github.com/stackforge/puppet-ceph/blob/master/spec/system/ceph_osds_spec.rb#L96 is interpretted and run. It then drives the virtual machines. I would check the ruby version, the installed gems and try to match the version numbers exactly.
20:00 dachary it looks like two things happen in parallel although we expect them to happen sequentially
20:01 dachary namely
20:02 dachary line 5585 shows an empty puppet file is applied which is found at https://github.com/stackforge/puppet-ceph/blob/master/spec/system/ceph_osds_spec.rb#L129
20:02 dmsimard dachary: Running the tests with my hack right now, just to confirm that this is the issue
20:03 dachary hum
20:03 dachary it's not paralle
20:03 dachary l
20:03 dachary it's the order that changes
20:04 dachary https://github.com/stackforge/puppet-ceph/blob/master/spec/system/ceph_osds_spec.rb#L116
20:04 dachary runs after
20:04 dachary https://github.com/stackforge/puppet-ceph/blob/master/spec/system/ceph_osds_spec.rb#L85
20:04 dachary which removes
20:04 dachary https://github.com/stackforge/puppet-ceph/blob/master/spec/system/ceph_osds_spec.rb#L110
20:04 dachary and kaboom
20:04 dachary I suspect the version installed on the former machine did the other way around and that was fine
20:06 dachary moving https://github.com/stackforge/puppet-ceph/blob/master/spec/system/ceph_osds_spec.rb#L110  after line 134 will probably fix the problem
20:12 openstackgerrit Merged stackforge/puppet-ceph: Remove templatedir directive from puppet.conf  https://review.openstack.org/135362
20:19 dmsimard dachary: I'll try that then, let you know how it goes
20:28 dmsimard dachary: Lots of clutter with the "facter -yaml" commands in the output.. I'd probably output it only if the exit is 1 or something
20:32 dmsimard dachary: Nope. That didn't work.
20:32 dachary what's the output this time ?
20:33 dachary dmsimard: ^
20:33 dmsimard dachary: Same thing, complaining about config.yaml all over.. It's really because the node_terminus parameter from puppet.conf isn't deleted properly
20:34 dachary could you share the paste link ?
20:36 vinsh joined #puppet-openstack
20:36 dachary it would be real strange if we get the exact same output  dmsimard
20:36 dachary hum you're right
20:37 dmsimard dachary: butchered the output, another test is already running with my hack in it
20:37 dachary now, I don't quite understand why this ini_setting is a noop
20:37 dmsimard mgagne: Have you ever heard of something like this ? http://lists.ceph.com/pipermail/ceph-users-ceph.com/2014-November/044742.html
20:38 dachary never, but that's why I don't write Loïc but Loic in files ;-)
20:39 dmsimard Ok, fair enough... I'll fix that
20:40 dmsimard sorry fc__ :)
20:40 dachary dmsimard: which ruby version are you running ?
20:41 dachary puppetlabs/inifile is pinned to 1.0.3 it's not that
20:41 dmsimard dachary: I'll tell you after the test ends running but this is on trusty
20:44 openstackgerrit David Moreau Simard proposed stackforge/puppet-ceph: Remove unnecessary non-ascii characters  https://review.openstack.org/135421
20:44 fc__ dmsimard: heh, yes, I'm regularly hitting this issue, we have a workaround about this in our puppet setup
20:44 dmsimard fc__: Really.. first time I see that.
20:45 fc__ dmsimard: https://github.com/enovance/edeploy-roles/blob/master/puppet-master.install#L49-L51
20:46 fc__ dmsimard: but I also frequently remove these utf-8 characters :)
20:46 dachary note that the problem happens on an ubuntu precise 12.04.2 LTS (line 5403 of http://paste.ubuntu.com/9075762/). It has 512MB of RAM (that's not much but there is no indication it's the problem)
20:46 dachary fc__: maybe we get to have our names spelled correctly in 2050 ;-)
20:47 fc__ dachary: yes … maybe
20:48 fc__ dachary: we first have to convince teachers to let the kids put accents on uppercase letters … my second son's name being Émile, I will insist on this
20:49 dachary My children don't have accents in their names, you're a brave man ;-)
20:49 fc__ dachary: pfff
20:49 dmsimard dachary: ruby 1.9.1
20:50 * fc__ back to housework
20:50 dmsimard dachary: Looks like your box is running 1.9.3
20:51 dachary dmsimard: in any case you're right, sorry for misleading you. The root of the problem is that the ensure absent is a noop in the virtual machine.
20:51 dmsimard dachary: Wait, no. 1.9.3.
20:51 dachary The ordering of the statements is not the problem.
20:51 dachary 12.04 has ruby1.8 it seems
20:52 dmsimard fc__: lol @ your comment in the review.. one for the books
20:52 dachary line 5301 of http://paste.ubuntu.com/9075762/ is rubysitedir: /usr/local/lib/site_ruby/1.8
20:53 dachary dmsimard: I must go, please let me know what you find, that's quite a mystery :-(
20:54 dmsimard dachary: Yeah, I know what the issue is.. just tricky to fix
20:54 dmsimard dachary: I'll send you an e-mail and push a fix hopefully
20:54 dachary note that it runs 12.04.2 but the latest is 12.04.5 http://releases.ubuntu.com/precise/
20:54 dachary the VM that fails I mean
20:55 openstackgerrit Merged stackforge/puppet-ceph: Remove unnecessary non-ascii characters  https://review.openstack.org/135421
20:55 dachary dmsimard: you could try to run the test on the former machine and see what it gives
20:59 xarses joined #puppet-openstack
21:01 dmsimard dachary: gone already ?
21:01 gildub joined #puppet-openstack
21:25 sputnik13 joined #puppet-openstack
21:25 stamak joined #puppet-openstack
21:44 patcable where is the firewall module getting the idea that glance is running on port 9292?
21:44 patcable sorry, wrong question
21:44 patcable where is the firewall module getting the idea that glance is running on port *89292*
21:44 patcable because its running on 9292. but it makes rules for 89292. which doesnt make sense at all
21:51 ibravo joined #puppet-openstack
21:53 dgurtner joined #puppet-openstack
21:54 openstackgerrit Merged stackforge/puppet-nova: Update [neutron] section in nova.conf for Juno  https://review.openstack.org/131566
21:59 EmilienM mgagne: thanks for reviews, and also for dropping your -2
22:03 xingchao joined #puppet-openstack
22:04 EmilienM mgagne: what do you think about my last patchset on https://review.openstack.org/#/c/126458/ ?
22:05 rcallawa joined #puppet-openstack
22:06 openstackgerrit Merged stackforge/puppet-nova: Tag all nova packages  https://review.openstack.org/134291
22:07 rcallawa joined #puppet-openstack
22:10 openstackgerrit Merged stackforge/puppet-keystone: Fix config options deprecated in Juno  https://review.openstack.org/133751
22:10 dmsimard dgurtner: OH, hi.
22:10 dmsimard dgurtner: I was sending you an e-mail...
22:10 dachary dmsimard: back
22:11 dachary dmsimard: but only back shortly ;-)
22:11 dachary it's getting late here
22:11 dmsimard dachary: It's ok, I'll check if dgurtner has an idea on how to fix.. I can't figure it out
22:12 dachary dmsimard: did you try running it on the old machine ?
22:14 dmsimard dachary: trying now
22:14 RichardRaseley joined #puppet-openstack
22:15 EmilienM crinkle: is your patch in openstacklib still WIP (swap aviator)
22:15 EmilienM (it's a question)
22:15 dachary it's good to have a baseline when things go crazy ;-)
22:16 crinkle EmilienM: yes, I'll mark it so
22:16 crinkle i want to do some more experimentation and consensus hasn't been reached on whether it's a good idea
22:16 openstackgerrit Malik Bougacha proposed stackforge/puppet-nova: supporting lxc cpu mode The default cpu_mode with the lxc libvirt type doesn't support the cpu_mode configuration. If one is defined, the nova-compute service will fail.  https://review.openstack.org/133987
22:16 EmilienM crinkle: if you don't mark it and Jenkins pass, I use to review the patches
22:17 crinkle EmilienM: I would still appreciate feedback on it :)
22:18 RichardRaseley What is the current status of support for CentOS 7 in the stackforge/puppet-* modules? I haven't really had any issues up until trying to use ::nova::compute.
22:18 RichardRaseley It is choking on finding the init scripts for 'messagebus' and 'libvirtd'.
22:19 EmilienM crinkle: good to know. I'll take some time to make it asap.
22:19 mgagne RichardRaseley: until someone uses CentOS 7 and want support for it, no one will start working on it I guess
22:19 mgagne RichardRaseley: scratch your own itch driven development
22:19 crinkle i've been using it on centos 7 without issue
22:19 EmilienM (me too)
22:19 crinkle for juno anyways
22:19 xarses RichardRaseley: ya, there are likely many gotcha's around the init scripts since it's moving from the rhel family being systemv init, to RHEL7 family being systemd
22:20 nibalizer centos 6 was upstart actually
22:20 patcable was it?
22:20 patcable i thought it was sysv
22:20 nibalizer nerp
22:20 xarses its sysv
22:20 nibalizer oh doi have this wrong
22:20 xarses only ubunut used upstart
22:20 nibalizer my bad
22:20 patcable /etc/init.d on a centos 6 machine im using has a ton of init scripts in it..
22:21 nibalizer ya they might have done the ubuntu thing where you can mix and match with old init scripts
22:22 nibalizer but some googling shows that centos 6 definitely has upstart
22:23 EmilienM crinkle: do we have stable/juno now?
22:23 xarses this is the first that I've seen any upstart stuff for centos
22:24 crinkle EmilienM: no
22:24 xarses I'm guessing nearly none of the scripts where migrated then, because I never ran into one on centos
22:24 EmilienM crinkle: what are we waiting for?
22:24 crinkle I don't think we're done with the juno compatibility patches, are we?
22:24 xarses left #puppet-openstack
22:24 xarses joined #puppet-openstack
22:24 crinkle EmilienM: also didn't you want to wait for integration tests before juno?
22:24 EmilienM crinkle: we miss that one: https://review.openstack.org/#/c/131576/
22:25 EmilienM crinkle: I don't see any reason why we should wait for integration tests.
22:26 EmilienM also https://review.openstack.org/#/c/128066/
22:26 EmilienM crinkle: and I think that's it
22:26 crinkle that one has a merge conflict
22:26 EmilienM mfisch: ^
22:27 crinkle EmilienM: should we make a 5.0.0 release when we make that branch? or wait a bit?
22:28 nibalizer ya dont wait on integration tests
22:28 EmilienM crinkle: yes
22:28 EmilienM nibalizer: lol
22:28 nibalizer its in progress, not blocked or anything, but still a ways off
22:28 EmilienM no problem, thanks for taking care of this
22:29 EmilienM crinkle: except if anyone has good arguments, we should make a Juno release this week
22:29 crinkle EmilienM: okay
22:30 patcable what's the proper repo to follow? puppetlabs/puppetlabs-openstack or stackforge/puppet-openstack ?
22:31 EmilienM patcable: are you looking for a composition layer?
22:32 patcable er..
22:32 patcable yes
22:32 patcable I think so
22:32 patcable or at least: we've been using that module to bring everything together, yes
22:33 patcable making implementation changes to the openstack module instead of the sub modules makes merging from upstream a lot easier
22:33 RichardRaseley crinkle: I am using it with RDO Icehouse on CentOS 7.
22:34 sputnik13 joined #puppet-openstack
22:39 mmaglana joined #puppet-openstack
22:43 badiane_ka joined #puppet-openstack
22:43 crinkle RichardRaseley: I'm not sure icehouse is supported on centos 7, going by http://docs.openstack.org/icehouse/install-guide/install/yum/content/basics-packages.html
22:44 crinkle patcable: stackforge/puppet-openstack is deprecated, there are lots of composition modules but no official one yet afaik
22:47 sputnik13 joined #puppet-openstack
22:47 RichardRaseley crinkle: I'm not sure either, but gosh darn it if I am not going to try. =]
22:48 dgurtner joined #puppet-openstack
22:48 dgurtner joined #puppet-openstack
22:48 crinkle RichardRaseley: apparently we are making a juno release this week if you want to wait a bit :P
22:49 mdorman RichardRaseley: fyi i’ve been working on moving to cent7 the last couple weeks and we’re on icehouse.  aside from the init scripts/systemd unit files issue discussed above, for the most part everything is working for me.
22:50 RichardRaseley crinkle: I would *love* to move to Juno, but our SDN vendor hasn't released their Juno packages yet (soon).
22:50 RichardRaseley mdorman: Thanks for that input.
22:51 mdorman we build our own rpms using anvil, so i just handled the systemd stuff there
22:51 mdorman the puppet service resource type understands systemd, so the puppet stuff just works
22:52 RichardRaseley mdorman: Can you share how you worked around the issue? In looking at stackforge/puppet-nova (release 4.1.0) at file manifests/compute/libvirt.pp it appears as if the provider for 'service' should default to undef on centos 7, so I am not sure why it is complaining. Or maybe the error is originating somewhere else and that is a red herring.
22:52 RichardRaseley You already answered that first question, sorry.
22:54 dgurtner dmsimard: dachary hi!
22:54 dachary howdy
22:55 dgurtner dmsimard: yup, saw your email just now. I'll try to look into it, but currently stuck in a hotel with spotty network connection at best :/
22:55 dmsimard oi
22:55 dmsimard puppetlabs doesn't provide an image better than 12.04.2
22:55 dmsimard doesn't provide 14.04 either
22:56 dmsimard rspec-system will become a blocker soon..
22:56 dmsimard EmilienM: You guys are going to use beaker now for CI ? Or what ?
22:56 dmsimard EmilienM: I left kind of early in that session
22:56 dgurtner dmsimard: my issue atm is, that i'm unable to download virtualbox without network loss :/
22:56 mdorman RichardRaseley:  to be honest, i haven’t gotten to upgrading any compute nodes yet.  so it’s quite possible i’ll run into the same thing with libvirt
22:57 mdorman RichardRaseley:   where do you see the provider as undef for cent 7?
22:57 dmsimard dgurtner: Do you have access to the old integration box ? You can test from there
22:57 RichardRaseley It actually isnt, as 'operatingsystem' for centos os 'CentOS', not redhat
22:59 dmsimard dachary: It's from the new one. Do you want a paste from the old one ?
23:01 dachary just so we're clear, could you re-iterate which is the old and the new output ?
23:01 RichardRaseley crinkle: Quick question, how is this (https://github.com/stackforge/puppet-nova/blob/master/manifests/params.pp#L43-L57) block of code even a thing?
23:01 mdorman RichardRaseley:   hmm, ok, i’m confused then.  what exactly are the errors you’re getting?
23:01 RichardRaseley crinkle: insofar as, if it is redhat it will *never* hit that second block.
23:02 RichardRaseley mdorman: The issue is that the default service provider in params.pp is 'init', and there is no case for 'CentOS'.
23:02 dmsimard Hunner: Any idea why upstream is dead for https://github.com/puppetlabs/puppet-vagrant-boxes ?
23:02 crinkle RichardRaseley: I don't understand, if it is Redhat it will go to https://github.com/stackforge/puppet-nova/blob/master/manifests/params.pp#L47
23:02 mdorman RichardRaseley: line 47, right?  it does seem like line 51 ought to probably be systemd and not undef
23:02 RichardRaseley mdorman: The error is, therefore, Error: /Service[messagebus]: Could not evaluate: Could not find init script for 'messagebus' since it is trying to use the init.
23:03 crinkle I think provider => undef means it will use the default provider
23:03 crinkle rather than overriding it
23:03 crinkle so i think that's correct
23:03 mdorman can you actaully trace $special_service_provider to the service resources for messagebus and libvirtd?
23:03 mdorman ^ RichardRaseley
23:03 mdorman crinkle:  ah, yeah, that makes more sense
23:03 mfisch EmilienM: looking, I guess you guys merged Richard's before mine
23:04 RichardRaseley OK, so the confusion here is, in part, that I am on release 4.1.0 of puppet-nova, which looks like this (https://github.com/stackforge/puppet-nova/blob/d6827929ea4c49b168d4164a048dc6280b8b6e9d/manifests/params.pp#L47-L57).
23:04 RichardRaseley crinkle: We can't set the same variable twice, no?
23:04 crinkle no, and that was fixed in 4.2.0
23:05 RichardRaseley OK, therein lies the problem.
23:05 mdorman RichardRaseley: ah, ok, yeah that makes sense.
23:05 mdorman so go to 4.2.0?
23:05 RichardRaseley Let me change my ref to 4.2.0 from 4.1.0
23:05 RichardRaseley Yeah.
23:05 RichardRaseley I appreciate the hand holding.
23:05 crinkle I did the same "wat" when I saw that, which is why it is now fixed :)
23:05 RichardRaseley crinkle: Nice
23:08 RichardRaseley crinkle mdorman: Aww yeah, got me some of that sweet icehouse on centos 7 action going.
23:09 crinkle nice
23:09 RichardRaseley 4.2.0 fixed the issue
23:09 RichardRaseley Thanks
23:12 mdorman glad it worked out
23:14 xingchao joined #puppet-openstack
23:15 openstackgerrit Matt Fischer proposed stackforge/puppet-keystone: Fix deprecated LDAP config options  https://review.openstack.org/128066
23:15 openstackgerrit Ivan Chavero proposed stackforge/puppet-nova: Support allow_start and allow_end parameters  https://review.openstack.org/135123
23:17 mfisch EmilienM: fixed ^
23:24 mdorman mfisch:  you going to meetup tonight?
23:24 mfisch mdorman: nope, but the crew from here is
23:25 mdorman cool
23:25 mfisch is Kris going?
23:25 mdorman trying to con kris int coming down, he’s waivering.  think i have him convinced
23:26 mfisch I'm going to the airport to pick up family instead
23:28 Hunner dmsimard: https://github.com/puppetlabs/puppetlabs-packer
23:28 Hunner We use packer now instead of veewee
23:32 crinkle do we know what's going on with https://review.openstack.org/#/c/132482/ ?
23:33 crinkle it doesn't show up in http://status.openstack.org/zuul/
23:36 aimon joined #puppet-openstack
23:40 mmaglana joined #puppet-openstack
23:42 mfisch mdorman: hey on your control nodes, do you have more nova-api processes than the workers setting?
23:43 mfisch mdorman: osapi_compute_workers is the setting
23:44 mdorman hmm, let me look
23:45 mdorman mfisch:  yeah, looks like there’s about 16 more
23:45 mdorman setting is 40, actual child processes running 56
23:45 mfisch okay, we have 32 and 66
23:45 mdorman wonder if that setting is jsut how many start at first.  like a lower limit, like apache, or something
23:46 mfisch most likely
23:46 mfisch I'm going to do an experiment
23:49 dmsimard Hunner: ty
23:49 mfisch mdorman: its started by WSGI
23:50 mdorman mfisch:  gotta head out to meetup now.  curious to know what you find out

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