| *** richm has quit IRC | 00:00 | |
| *** dfisher has quit IRC | 00:01 | |
| *** zhangjn has joined #puppet-openstack | 00:02 | |
| EmilienM | patrickeast: have you tried https://wiki.openstack.org/wiki/Puppet/Deploy ? | 00:11 |
|---|---|---|
| EmilienM | oh yeah, just saw your link | 00:11 |
| EmilienM | nibalizer: ok np | 00:11 |
| *** zhangjn has quit IRC | 00:11 | |
| *** markvoelker has quit IRC | 00:11 | |
| patrickeast | EmilienM: yep, thats basically what i'm using with a few tweaks for ips/passwords to get started | 00:12 |
| patrickeast | EmilienM: if i remove that check thats trying to do this token issue call here https://github.com/openstack/puppet-keystone/blob/master/lib/puppet/provider/keystone_user/openstack.rb#L118 then i can apply the manifest more than once and it is happy | 00:13 |
| patrickeast | not sure if its a bug or just something i did wrong in my setup | 00:13 |
| _ody | patrickeast: are you setting down and openrc file on your first run? | 00:18 |
| patrickeast | _ody: i'm creating a openrc file, do i need to configure one for use somewhere? | 00:21 |
| patrickeast | _ody: and i'm not sure what you mean by 'down' so probably not | 00:22 |
| _ody | The one from openstack_extras causing the exact problem you are having, delete it and I bet your run works without error. | 00:22 |
| patrickeast | _ody: where would it be? i'm only seeing the file i am explicity creating in my manifest like so http://paste.openstack.org/show/477920/ | 00:25 |
| *** dgurtner has quit IRC | 00:26 | |
| _ody | patrickeast: /root/openrc usually. Another file location might make you safe; unless you've sourced the file and the variables are set in the same shell you are running puppet from. | 00:28 |
| patrickeast | _ody: ahhh ok i see | 00:32 |
| patrickeast | yea i had sourced my 'keystonerc_admin' file and it was picking that up with doing the puppet apply | 00:32 |
| patrickeast | clearing the exports makes it go away | 00:33 |
| patrickeast | _ody: thanks! | 00:33 |
| *** aimon has quit IRC | 00:37 | |
| *** ntpttr has joined #puppet-openstack | 00:41 | |
| *** jerrygb has joined #puppet-openstack | 00:44 | |
| *** bapalm has quit IRC | 00:46 | |
| *** bapalm has joined #puppet-openstack | 00:48 | |
| *** jerrygb has quit IRC | 00:48 | |
| *** zhangjn has joined #puppet-openstack | 00:58 | |
| *** aimon has joined #puppet-openstack | 01:01 | |
| *** dtalton has quit IRC | 01:03 | |
| *** gildub has joined #puppet-openstack | 01:08 | |
| *** ntpttr has quit IRC | 01:10 | |
| *** markvoelker has joined #puppet-openstack | 01:12 | |
| *** markvoelker has quit IRC | 01:17 | |
| *** rahulait has quit IRC | 01:19 | |
| *** angdraug has quit IRC | 01:25 | |
| *** nlahouti1 has quit IRC | 01:27 | |
| *** rahulait has joined #puppet-openstack | 01:55 | |
| openstackgerrit | Adam Vinsh proposed openstack/puppet-swift: Manage swift with swiftinit service provider https://review.openstack.org/203220 | 01:57 |
| *** jerrygb has joined #puppet-openstack | 02:00 | |
| *** ducttape_ has joined #puppet-openstack | 02:16 | |
| *** rahulait has quit IRC | 02:22 | |
| *** michchap has quit IRC | 02:24 | |
| *** michchap has joined #puppet-openstack | 02:25 | |
| *** cwolferh has quit IRC | 02:26 | |
| *** markvoelker has joined #puppet-openstack | 02:43 | |
| *** ilbot3 has quit IRC | 02:44 | |
| *** markvoelker has quit IRC | 02:48 | |
| *** ilbot3 has joined #puppet-openstack | 02:48 | |
| *** ntpttr has joined #puppet-openstack | 02:53 | |
| *** ntpttr has left #puppet-openstack | 02:53 | |
| *** ken_barber has quit IRC | 03:04 | |
| *** ken_barber has joined #puppet-openstack | 03:10 | |
| *** ducttape_ has quit IRC | 03:15 | |
| *** ducttape_ has joined #puppet-openstack | 03:15 | |
| *** ducttape_ has quit IRC | 03:19 | |
| *** jmccrory has quit IRC | 03:23 | |
| *** rahulait has joined #puppet-openstack | 03:30 | |
| *** nlahouti has joined #puppet-openstack | 03:49 | |
| *** nlahouti1 has joined #puppet-openstack | 03:51 | |
| *** nlahouti has quit IRC | 03:53 | |
| *** ducttape_ has joined #puppet-openstack | 04:01 | |
| *** severion has joined #puppet-openstack | 04:06 | |
| *** aglarendil__ has joined #puppet-openstack | 04:08 | |
| *** rhefner_ has joined #puppet-openstack | 04:08 | |
| *** simonmcc_ has joined #puppet-openstack | 04:08 | |
| *** bapalm_ has joined #puppet-openstack | 04:10 | |
| *** v1k0d3n has quit IRC | 04:11 | |
| *** _morgan has quit IRC | 04:11 | |
| *** aglarendil has quit IRC | 04:11 | |
| *** rhefner has quit IRC | 04:11 | |
| *** simonmcc has quit IRC | 04:11 | |
| *** mjb has quit IRC | 04:11 | |
| *** bapalm has quit IRC | 04:11 | |
| *** aglarendil__ is now known as aglarendil | 04:11 | |
| *** rhefner_ is now known as rhefner | 04:11 | |
| *** _morgan has joined #puppet-openstack | 04:12 | |
| *** simonmcc_ is now known as simonmcc | 04:13 | |
| *** mjb has joined #puppet-openstack | 04:15 | |
| *** zhangjn has quit IRC | 04:30 | |
| *** zhangjn has joined #puppet-openstack | 04:31 | |
| *** zhangjn has quit IRC | 04:31 | |
| *** zhangjn has joined #puppet-openstack | 04:32 | |
| *** ducttape_ has quit IRC | 04:36 | |
| *** jmccrory has joined #puppet-openstack | 04:44 | |
| *** markvoelker has joined #puppet-openstack | 04:44 | |
| *** markvoelker has quit IRC | 04:49 | |
| *** fedexo has joined #puppet-openstack | 04:53 | |
| openstackgerrit | Doug Knight proposed openstack/puppet-neutron: require nova account in nova_admin_tenant_id_setter https://review.openstack.org/241466 | 04:55 |
| *** rahulait has quit IRC | 05:05 | |
| openstackgerrit | Gilles Dubreuil proposed openstack/puppet-openstack-integration: Bump UCA and RDO to Liberty GA https://review.openstack.org/240636 | 05:06 |
| openstackgerrit | Gilles Dubreuil proposed openstack/puppet-neutron: beaker: bump UCA and RDO to Liberty GA repository https://review.openstack.org/239258 | 05:07 |
| openstackgerrit | Doug Knight proposed openstack/puppet-glance: Stop collecting glance services by name https://review.openstack.org/240630 | 05:08 |
| *** nlahouti has joined #puppet-openstack | 05:11 | |
| *** nlahouti1 has quit IRC | 05:11 | |
| openstackgerrit | Doug Knight proposed openstack/puppet-glance: Stop collecting glance services by name https://review.openstack.org/240630 | 05:13 |
| *** dh64 has joined #puppet-openstack | 05:41 | |
| *** jerrygb has quit IRC | 06:01 | |
| *** rahulait has joined #puppet-openstack | 06:10 | |
| *** nlahouti has quit IRC | 06:13 | |
| *** markvoelker has joined #puppet-openstack | 06:45 | |
| *** markvoelker has quit IRC | 06:49 | |
| *** gildub has quit IRC | 06:58 | |
| *** markvoelker has joined #puppet-openstack | 07:00 | |
| *** markvoelker has quit IRC | 07:05 | |
| *** asilenkov has quit IRC | 07:25 | |
| *** bogdando has quit IRC | 07:26 | |
| *** bogdando has joined #puppet-openstack | 07:26 | |
| *** asilenkov has joined #puppet-openstack | 07:28 | |
| *** jerrygb has joined #puppet-openstack | 07:47 | |
| *** dgurtner has joined #puppet-openstack | 07:50 | |
| *** jerrygb has quit IRC | 07:52 | |
| *** samuelBartel has joined #puppet-openstack | 08:12 | |
| *** samuelBartel has quit IRC | 08:17 | |
| *** rahulait has quit IRC | 08:17 | |
| openstackgerrit | Mykyta Karpin proposed openstack/puppet-keystone: add policy driver option for keystone https://review.openstack.org/240946 | 08:28 |
| *** jtomasek has quit IRC | 08:37 | |
| *** gfidente|afk has quit IRC | 08:38 | |
| *** jtomasek has joined #puppet-openstack | 08:38 | |
| *** jpena has joined #puppet-openstack | 08:40 | |
| bkero | EmilienM: Did you run into Unable to resolve dependencies: beaker requires fog-google (~> 0.0.9); fog requires fog-google (>= 0.1.1)? | 08:51 |
| *** samuelBartel has joined #puppet-openstack | 08:53 | |
| *** dh64 has quit IRC | 08:55 | |
| *** dh64 has joined #puppet-openstack | 08:55 | |
| *** gfidente has joined #puppet-openstack | 08:59 | |
| *** markvoelker has joined #puppet-openstack | 09:01 | |
| *** dgurtner has quit IRC | 09:04 | |
| *** markvoelker has quit IRC | 09:06 | |
| *** jistr has joined #puppet-openstack | 09:08 | |
| openstackgerrit | Mykyta Karpin proposed openstack/puppet-keystone: add policy driver option for keystone https://review.openstack.org/240946 | 09:11 |
| *** karimb has joined #puppet-openstack | 09:13 | |
| zigo | Can someone help me with adding a patch to puppet-openstacklib ? | 09:33 |
| zigo | This is still related to the Debian style packages on top of Ubuntu... | 09:34 |
| *** aimon has quit IRC | 09:40 | |
| *** gfidente has quit IRC | 09:44 | |
| *** fwdit has joined #puppet-openstack | 09:47 | |
| *** gildub has joined #puppet-openstack | 09:49 | |
| zigo | EmilienM: Could you ping me when you're up please? | 09:49 |
| *** gfidente has joined #puppet-openstack | 09:50 | |
| *** dgurtner has joined #puppet-openstack | 09:55 | |
| *** derekh has joined #puppet-openstack | 09:55 | |
| *** paramite has joined #puppet-openstack | 10:13 | |
| *** fedexo has quit IRC | 10:19 | |
| *** dgurtner has quit IRC | 10:23 | |
| *** gildub has quit IRC | 10:37 | |
| *** aimon has joined #puppet-openstack | 10:45 | |
| openstackgerrit | Serg Melikyan proposed openstack/puppet-murano: Application resource implementation https://review.openstack.org/220238 | 10:46 |
| *** fvollero_ has joined #puppet-openstack | 10:48 | |
| *** jistr_ has joined #puppet-openstack | 10:48 | |
| *** paramite_ has joined #puppet-openstack | 10:48 | |
| *** jistr has quit IRC | 10:48 | |
| *** fvollero has quit IRC | 10:49 | |
| *** paramite has quit IRC | 10:49 | |
| *** fvollero_ has quit IRC | 10:52 | |
| *** paramite_ has quit IRC | 10:52 | |
| *** fvollero has joined #puppet-openstack | 10:53 | |
| *** jistr_ has quit IRC | 10:53 | |
| *** chem has quit IRC | 10:54 | |
| *** chem` has joined #puppet-openstack | 10:54 | |
| *** aimon has quit IRC | 10:54 | |
| *** khomkrit has joined #puppet-openstack | 10:57 | |
| *** chem` has quit IRC | 10:59 | |
| *** chem has joined #puppet-openstack | 11:00 | |
| *** markvoelker has joined #puppet-openstack | 11:02 | |
| *** aimon_ has joined #puppet-openstack | 11:04 | |
| *** paramite_ has joined #puppet-openstack | 11:05 | |
| *** jistr_ has joined #puppet-openstack | 11:05 | |
| *** markvoelker has quit IRC | 11:06 | |
| openstackgerrit | Athlan-Guyot sofer proposed openstack/puppet-keystone: Support for services different only by type. https://review.openstack.org/238954 | 11:30 |
| *** fwdit has quit IRC | 11:38 | |
| karimb | so EmilienM getting back at you, i m instantiating neutron::agents::ml2::ovs but i dont see the values reflected in the /etc/neutron/plugins/openvswitch/ovs_neutron_plugin.ini | 11:39 |
| karimb | am i hitin some bug or something ? | 11:39 |
| *** zhangjn has quit IRC | 12:02 | |
| *** ferest_ has joined #puppet-openstack | 12:03 | |
| *** ferest_ has quit IRC | 12:07 | |
| *** fwdit has joined #puppet-openstack | 12:08 | |
| *** cdelatte has joined #puppet-openstack | 12:15 | |
| *** paramite_ is now known as paramite | 12:17 | |
| *** markvoelker has joined #puppet-openstack | 12:17 | |
| *** markvoelker has quit IRC | 12:22 | |
| *** dgurtner has joined #puppet-openstack | 12:24 | |
| *** bogdando has quit IRC | 12:25 | |
| *** bogdando has joined #puppet-openstack | 12:25 | |
| *** dgurtner has quit IRC | 12:34 | |
| *** jistr_ is now known as jistr | 12:44 | |
| *** ferest has joined #puppet-openstack | 12:50 | |
| *** dprince has joined #puppet-openstack | 13:09 | |
| *** jayg|g0n3 is now known as jayg | 13:16 | |
| *** ferest has quit IRC | 13:17 | |
| EmilienM | bkero: yes, beaker was updated to pin the right version | 13:17 |
| EmilienM | zigo: in 10m | 13:18 |
| *** ferest has joined #puppet-openstack | 13:20 | |
| *** markvoelker has joined #puppet-openstack | 13:23 | |
| iurygregory | morning people o/ | 13:34 |
| zigo | EmilienM: ping? | 13:37 |
| EmilienM | zigo: hey, sorry I was eating my pancakes with mapple sirup :) | 13:38 |
| zigo | EmilienM: :) | 13:39 |
| zigo | You definitively need to take your time for that... | 13:39 |
| zigo | EmilienM: I'd like to know what's best place in puppet-openstacklib to do some "lsb_release -is" checks and define os_package_type depending on this. | 13:40 |
| EmilienM | mwhahaha: good morning, please look my reply on https://trello.com/c/XLJJJBF0/71-move-modules-to-the-os-service-default-pattern | 13:40 |
| EmilienM | zigo: there is a fact already in Puppet | 13:41 |
| EmilienM | a sec | 13:41 |
| zigo | (if that variable isn't defined yet through os_package_type.txt as discussed in the summit) | 13:41 |
| EmilienM | zigo: in PM I sent you the result of facts on my machine | 13:41 |
| *** xarses has joined #puppet-openstack | 13:41 | |
| zigo | So, it'd be: if ! $::os_package_type { if `lsb_release -is` = 'Debian' { ... } else{ ... } } | 13:42 |
| EmilienM | you can use it in Puppet manifests with $::operatingsystemmajrelease for example | 13:42 |
| EmilienM | if $::operatingsystemmajrelease == '7' then distro = 'wheezy' | 13:42 |
| EmilienM | well, then is not Puppetish :) | 13:43 |
| zigo | Yup, sorry... | 13:43 |
| zigo | So, it'd be: if ! $::os_package_type { if $::operatingsystem = 'Debian' { ... } else{ ... } } | 13:43 |
| EmilienM | I suggest you to read https://puppetlabs.com/facter | 13:43 |
| EmilienM | yeah, something like this | 13:43 |
| zigo | I did... | 13:43 |
| zigo | So, where to place that code in puppet-openstacklib, so that it's available for all project-specific puppet manifests? (like puppet-horizon and puppet-nova) | 13:44 |
| EmilienM | zigo: for $::os_package_type ? | 13:44 |
| zigo | EmilienM: ^ | 13:44 |
| EmilienM | so IIUC at the Summit we said : | 13:45 |
| zigo | $::os_package_type would be either debian (because $::operatingsystem == 'Debian' or if forced through a /etc fact...) or ubuntu. | 13:45 |
| zigo | Then we can reuse that in puppet-horizon and puppet-nova. | 13:45 |
| EmilienM | "as a Puppet deployer, I'll install a Debian package that will create /etc/facts.d/<custom-fact>.txt and consume the fact directly in manifests | 13:45 |
| zigo | Well, it can be a Debian package, or it can just be manually prepared on the base os image. | 13:46 |
| zigo | Probably MOS people will just drop a .txt file manually when preparing the base image. | 13:46 |
| EmilienM | I think there is nothing to do in Puppet | 13:46 |
| zigo | (easiest path...) | 13:46 |
| zigo | Well, there is. | 13:47 |
| zigo | This fact file shouldn't be mandatory. | 13:47 |
| bkero | EmilienM: I know, I updated it. :) | 13:47 |
| zigo | And if it's not there, I'd like to have $::os_package_type defined properly, without having to do a complex check. | 13:47 |
| bkero | EmilienM: I'm wondering if it was worked around in the tests since that hasnt made it to a release yet | 13:48 |
| zigo | See this: https://review.openstack.org/gitweb?p=openstack/puppet-horizon.git;a=commitdiff;h=8cf1db18198e4b66b2489e9775544ba5417e8338 | 13:48 |
| zigo | It could be simplified if we had $::os_package_type properly defined. | 13:48 |
| *** morazi has joined #puppet-openstack | 13:49 | |
| EmilienM | zigo: a sec | 13:49 |
| zigo | EmilienM: Is lib/facter/os_service_default.rb (in puppet-openstacklib) always executed before all individual project specific manifest/init.pp ? | 13:50 |
| *** jpena is now known as jpena|lunch | 13:51 | |
| *** bklei has joined #puppet-openstack | 13:51 | |
| EmilienM | zigo: yes | 13:51 |
| zigo | EmilienM: Ok, then that's a good place to put the checks, no? | 13:52 |
| EmilienM | zigo: I thought we already found the solution in Tokyo, why are we retalking about it? | 13:52 |
| EmilienM | we need one fact, which is created by a package. That's all about it | 13:52 |
| zigo | EmilienM: We did, I'm just discussing the implementation ! :) | 13:52 |
| zigo | EmilienM: Quite not all. Once that package is in use (or not), we still need to check for it. I just want to factor that check (existance of the fact) and the system in use, so that $::os_package_type is always defined correctly, no mater what. | 13:53 |
| zigo | (ie: always contains a value, even if the fact file isn't present) | 13:54 |
| *** sergmelikyan has joined #puppet-openstack | 13:56 | |
| karimb | EmilienM shouldnt this declaration properly set my ovs agent? http://pastebin.com/H1rS2zEL | 13:57 |
| EmilienM | zigo: I have an idea | 13:59 |
| EmilienM | zigo: a sec | 13:59 |
| *** fwdit has quit IRC | 13:59 | |
| EmilienM | zigo: https://github.com/openstack/puppet-openstack_extras/tree/master/manifests/repo/debian | 14:00 |
| EmilienM | I think we can manage this package in there | 14:00 |
| EmilienM | with a new param | 14:00 |
| EmilienM | so in your manifests we would use the fact provided by your package | 14:01 |
| EmilienM | if the package is not installed, that's because you don't want to use your packaging, so the fact will be empty | 14:01 |
| *** richm has joined #puppet-openstack | 14:02 | |
| EmilienM | but if the package is installed, you'll hit your condition and you can manage your own service /package name | 14:02 |
| zigo | EmilienM: Well, we don't want to *require* the pacakage. Running in plain vanilia Debian we don't need it ... | 14:02 |
| EmilienM | zigo: we won't require | 14:02 |
| EmilienM | but we need to provide a way to our users to install it with Puppet | 14:02 |
| EmilienM | and https://github.com/openstack/puppet-openstack_extras/tree/master/manifests/repo/debian is the right place | 14:02 |
| zigo | Just let me finish my patch to puppet-openstacklib, I believe it will be doing what we need. | 14:02 |
| zigo | Well, not really. | 14:02 |
| EmilienM | zigo: have you sent something already? | 14:02 |
| zigo | EmilienM: For the case of MOS, we wouldn't be using the Debian repositories. | 14:02 |
| zigo | Give me a few seconds, I'm writting the comments and then I do "git review" | 14:03 |
| *** samuelBartel has quit IRC | 14:10 | |
| openstackgerrit | Mykyta Karpin proposed openstack/puppet-nova: Update parameters for nova https://review.openstack.org/241614 | 14:12 |
| EmilienM | zigo: take time by the way | 14:12 |
| EmilienM | zigo: this won't be for liberty anyway | 14:13 |
| *** ferest has quit IRC | 14:13 | |
| openstackgerrit | Thomas Goirand proposed openstack/puppet-openstacklib: Prepare $::os_package_type https://review.openstack.org/241615 | 14:13 |
| zigo | EmilienM: We do need this for Liberty. | 14:13 |
| zigo | There you go, please check my patch. | 14:13 |
| EmilienM | zigo: you'll have to backport it | 14:13 |
| zigo | There's a bunch of things which need to be fixed for Liberty, which aren't in puppet-horizon yet. | 14:14 |
| zigo | Sure, no problem! :) | 14:14 |
| EmilienM | we're cutting stable/liberty soon | 14:14 |
| zigo | It's good that you're nearly in sync with the upstream release. | 14:14 |
| openstackgerrit | Athlan-Guyot sofer proposed openstack/puppet-keystone: Support for services different only by type. https://review.openstack.org/238954 | 14:14 |
| zigo | You *can't* be in sync just yet, as packages are always lagging behind though, but that's soon enough. | 14:14 |
| EmilienM | zigo: your patch is not good, you should use ruby code | 14:14 |
| zigo | EmilienM: I'm a complete newbie... What's that mean? :) | 14:15 |
| zigo | if end ? | 14:15 |
| zigo | if <foo> do <bar> end ? | 14:15 |
| openstackgerrit | Athlan-Guyot sofer proposed openstack/puppet-keystone: Keystone_endpoint match service by name/type. https://review.openstack.org/239434 | 14:15 |
| EmilienM | first I'm trying to understand what you're trying to do | 14:15 |
| zigo | EmilienM: Let me push something else. | 14:16 |
| EmilienM | again zigo... it looks super complicated | 14:17 |
| EmilienM | during the summit I proposed you a simple solution | 14:17 |
| EmilienM | I'm not sure you got it | 14:17 |
| zigo | I don't think it's super complicated at all. | 14:17 |
| zigo | And I believe my patch attempt is just implementing what we've talked about. | 14:18 |
| EmilienM | my proposal does not need any patch in puppet-openstacklib | 14:18 |
| openstackgerrit | Thomas Goirand proposed openstack/puppet-openstacklib: Prepare $::os_package_type https://review.openstack.org/241615 | 14:18 |
| EmilienM | zigo: let me show you something | 14:18 |
| EmilienM | zigo: I'm doing a patch for puppet-nova so you'll see what I'm thinking | 14:18 |
| zigo | EmilienM: https://review.openstack.org/gitweb?p=openstack/puppet-horizon.git;a=commitdiff;h=8cf1db18198e4b66b2489e9775544ba5417e8338 | 14:18 |
| zigo | See my patch there ... | 14:18 |
| zigo | It's too complicated, when it shouldn't. | 14:19 |
| zigo | I'm trying to factor the checks into puppet-openstacklib, so we don't need to write if $::operatingsystem != Debian or ! $::os_package_type or $::os_package_type != 'debian' { | 14:19 |
| zigo | (which is horrible to read...) | 14:19 |
| zigo | EmilienM: Is this syntaxically correct now? https://review.openstack.org/gitweb?p=openstack/puppet-openstacklib.git;a=commitdiff;h=7919501657e9388c42841b46dc6a1439867961e1 | 14:20 |
| vinsh | jpena|lunch: Good to see you in tokyo. I have factored in your expirer code to: https://review.openstack.org/#/c/203220/ | 14:21 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-nova: Do not merge - PoC for zigo https://review.openstack.org/241620 | 14:21 |
| EmilienM | zigo: ^ | 14:21 |
| zigo | EmilienM: Well, what happens if $::i_use_zigo_packaging isn't defined at all, and we're running in vanilia Debian? | 14:23 |
| EmilienM | zigo: if the fact is not defined, it will return False | 14:23 |
| zigo | Which will then lead to the wrong decision... | 14:23 |
| EmilienM | we need this package installed | 14:23 |
| zigo | So your test really should be: if $::i_use_zigo_packaging or $::operatingsystem == 'Debian' { | 14:24 |
| EmilienM | that was my proposal in Tokyo | 14:24 |
| zigo | I don't want to *require* this package to be installed if running Debian. I don't want to upload this package to Debian either. | 14:24 |
| EmilienM | zigo: if you see my code, we're already are in $::operatingsystem == 'Debian' conditionnal | 14:24 |
| zigo | That's a hack to run on Ubuntu. | 14:24 |
| zigo | Well, then your code is wrong, because we want to handle the ubuntu case ! :) | 14:25 |
| *** francois has quit IRC | 14:25 | |
| *** francois has joined #puppet-openstack | 14:25 | |
| EmilienM | zigo: not a problem, we can move the conditionnal | 14:26 |
| EmilienM | zigo: anyway, I don't think we need to patch openstacklib | 14:26 |
| zigo | EmilienM: What I would like to be able to do is replace the case $::operatingsystem by a case $::os_package_type | 14:27 |
| EmilienM | go ahead then | 14:27 |
| zigo | I see no other way to do that but to have the variable set in openstacklib. | 14:27 |
| EmilienM | I think all what you need to patch is params.pp of our modules. | 14:28 |
| EmilienM | Nothing else | 14:28 |
| zigo | Well, then: https://review.openstack.org/#/c/241615/ (patch set 2) | 14:28 |
| zigo | Is this correct? | 14:28 |
| *** jerrygb has joined #puppet-openstack | 14:28 | |
| EmilienM | no | 14:29 |
| zigo | I'm not sure about the leading :: | 14:29 |
| EmilienM | I told you, we don't need to do that i think | 14:29 |
| EmilienM | moreover, why are you patching os_service_default | 14:29 |
| EmilienM | os_service_default is a very different thing | 14:29 |
| EmilienM | your code is currently noop | 14:30 |
| zigo | Any other file which is executed before param.pp of the puppet-FOO stuff would be a good fit... | 14:30 |
| zigo | What do you mean? | 14:31 |
| zigo | EmilienM: What do you mean? | 14:33 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-openstack-integration: Bump UCA and RDO to Liberty GA https://review.openstack.org/240636 | 14:36 |
| openstackgerrit | Thomas Goirand proposed openstack/puppet-openstacklib: Prepare $::os_package_type https://review.openstack.org/241615 | 14:39 |
| mwhahaha | zigo: you should create a new file called lib/facter/os_package_type.rb and put that code in there | 14:41 |
| zigo | mwhahaha: Thanks, will do that immediately. Will it be executed before all manifests/{init,param}.pp automatically ? | 14:42 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-openstack-integration: Bump UCA and RDO to Liberty GA https://review.openstack.org/240636 | 14:42 |
| mwhahaha | facts get executed before puppet runs | 14:42 |
| mwhahaha | so yes | 14:42 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-neutron: beaker: bump UCA and RDO to Liberty GA repository https://review.openstack.org/239258 | 14:42 |
| *** severion has quit IRC | 14:42 | |
| *** v1k0d3n has joined #puppet-openstack | 14:43 | |
| openstackgerrit | Thomas Goirand proposed openstack/puppet-openstacklib: Prepare $::os_package_type https://review.openstack.org/241615 | 14:43 |
| mwhahaha | zigo: also that doesn't actually set the fact | 14:43 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-ceilometer: Bump UCA and RDO to Liberty GA https://review.openstack.org/241623 | 14:44 |
| mwhahaha | let me help you out with this one :D | 14:44 |
| zigo | mwhahaha: Sorry, I'm just discovering puppet here ... :( | 14:44 |
| mwhahaha | it's all good | 14:44 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-ceilometer: Bump UCA and RDO to Liberty GA https://review.openstack.org/241623 | 14:45 |
| zigo | mwhahaha: As its been a reoccuring (and blocking) issue in MOS for using what I've been doing in Debian, and blocking lots of other stuff, it'd be really a good progress if we could have this one issue tackled. | 14:45 |
| mwhahaha | sure, gimme a sec to fix it up | 14:45 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-cinder: Bump UCA and RDO to Liberty GA https://review.openstack.org/241624 | 14:45 |
| zigo | Thanks. | 14:45 |
| EmilienM | mwhahaha: that's not what we decided during the summit | 14:45 |
| EmilienM | we said we would not manage the fact in puppet but by a package in debian | 14:46 |
| zigo | EmilienM: That's exactly what we are doing, really... | 14:46 |
| EmilienM | mwhahaha: can you look https://trello.com/c/XLJJJBF0/71-move-modules-to-the-os-service-default-pattern please ? Yanis replied and I did so, are you agree? | 14:46 |
| mwhahaha | EmilienM: so we can drop the static fact via package | 14:46 |
| EmilienM | zigo: no you're trying to do some fact in openstacklib | 14:46 |
| mwhahaha | so this would just pick on on it | 14:46 |
| *** openstackgerrit has quit IRC | 14:47 | |
| zigo | EmilienM: Yes, because that package should *only* be installed in the case of using Debian style packages over on top of Ubuntu, and we need to handle all other cases. | 14:47 |
| zigo | So that patch on puppet-openstacklib does the work for it. | 14:47 |
| *** openstackgerrit has joined #puppet-openstack | 14:47 | |
| EmilienM | I think I need to see it in action to understand | 14:48 |
| zigo | EmilienM: Once I have that patch added, I'll start fixing stuff in Horizon and Nova... | 14:48 |
| mwhahaha | EmilienM: I agree with the trello comments, not sure i can comment but i will look in a minute | 14:48 |
| *** trown is now known as trown|brb | 14:49 | |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-designate: Bump UCA and RDO to Liberty GA https://review.openstack.org/241625 | 14:49 |
| *** dfisher has joined #puppet-openstack | 14:50 | |
| EmilienM | mwhahaha: let's go ahead except for logging, so we keep backward compat | 14:52 |
| mwhahaha | EmilienM: i can't comment on that board do i have to get added? | 14:52 |
| mwhahaha | sure i'll fix the cinder logging as well | 14:52 |
| EmilienM | mwhahaha: do you have a trello account? | 14:52 |
| mwhahaha | yea | 14:52 |
| mwhahaha | user is mwhahaha | 14:52 |
| EmilienM | ok | 14:52 |
| *** trown|brb is now known as trown | 14:53 | |
| EmilienM | mwhahaha: done, you should be part of the board now | 14:53 |
| mwhahaha | yup thanks | 14:53 |
| *** v1k0d3n has quit IRC | 14:54 | |
| *** v1k0d3n has joined #puppet-openstack | 14:54 | |
| *** jpena|lunch is now known as jpena | 14:57 | |
| jpena | vinsh: cool, I'll give it another look | 14:57 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-openstack_extras: DNM - CI test https://review.openstack.org/241628 | 14:58 |
| mwhahaha | zigo: where are you assuming os_package_type is being set for your fact? | 14:59 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-openstack_extras: CI test in Kilo https://review.openstack.org/241629 | 14:59 |
| mwhahaha | right now i don't see a way to actually set os_package_type prior to calculating that fact | 14:59 |
| zigo | mwhahaha: I'm assuming that if someone is installing Debian style packages (ie: MOS for example, or a rebuild of Debian packages for Ubuntu), then one will as well have the os_package_type already defined in /etc/facter/facts.d/os_pacage_type.txt | 15:00 |
| zigo | Sorry, I do it again... | 15:00 |
| zigo | mwhahaha: I'm assuming that if someone is installing Debian style packages over on top of Ubuntu (ie: MOS for example, or a rebuild of Debian packages for Ubuntu), then one will as well have the os_package_type already defined in /etc/facter/facts.d/os_pacage_type.txt | 15:01 |
| zigo | The code is there to handle all other cases. | 15:01 |
| EmilienM | :b clayton | 15:02 |
| mwhahaha | k i'm not sure the ordering of static facts (facts in /etc/facter/facts.d/) vs dynamic facts (this .rb file) | 15:02 |
| mwhahaha | i'll have to test that out | 15:02 |
| EmilienM | clayton: I agree with your comment | 15:02 |
| EmilienM | (on the Trello card) | 15:02 |
| EmilienM | clayton: do you think we should go ahead with other modules? | 15:03 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-glance: Bump UCA and RDO to Liberty GA https://review.openstack.org/241631 | 15:04 |
| zigo | mwhahaha: Ah ok, then please let me know. | 15:05 |
| openstackgerrit | Martin Mágr proposed openstack/puppet-openstacklib: OpenStackConfig provider https://review.openstack.org/234727 | 15:06 |
| zigo | mwhahaha: If it's executed *after*, then my logic is correct. If it's *before*, then probably the static fact file from /etc/facter/facts.d will override what's done dynamically, in which case it's going to be even more simple to write. | 15:07 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-aodh: Bump UCA and RDO to Liberty GA https://review.openstack.org/241632 | 15:07 |
| mwhahaha | yea basically i think this .rb might just be the default logic and the file in facts.d would override this one | 15:08 |
| mwhahaha | which would be simpler but i'll need to test | 15:08 |
| *** paramite is now known as paramite|afk | 15:09 | |
| *** dhollinger has joined #puppet-openstack | 15:09 | |
| *** dgurtner has joined #puppet-openstack | 15:09 | |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-heat: Bump UCA and RDO to Liberty GA https://review.openstack.org/241636 | 15:10 |
| openstackgerrit | Athlan-Guyot sofer proposed openstack/puppet-keystone: Support for services different only by type. https://review.openstack.org/238954 | 15:10 |
| openstackgerrit | Alex Schultz proposed openstack/puppet-openstacklib: Prepare $::os_package_type https://review.openstack.org/241615 | 15:10 |
| *** ducttape_ has joined #puppet-openstack | 15:11 | |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-horizon: Bump UCA and RDO to Liberty GA https://review.openstack.org/241637 | 15:11 |
| *** paramite|afk is now known as paramite | 15:11 | |
| openstackgerrit | Athlan-Guyot sofer proposed openstack/puppet-keystone: Keystone_endpoint match service by name/type. https://review.openstack.org/239434 | 15:11 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-ironic: Bump UCA and RDO to Liberty GA https://review.openstack.org/241638 | 15:12 |
| openstackgerrit | Mykyta Karpin proposed openstack/puppet-nova: Update parameters for nova https://review.openstack.org/241614 | 15:12 |
| openstackgerrit | Paul Belanger proposed openstack/puppet-openstack_extras: [WIP] DNM https://review.openstack.org/241640 | 15:14 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-keystone: Bump UCA and RDO to Liberty GA https://review.openstack.org/241641 | 15:14 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-manila: Bump UCA and RDO to Liberty GA https://review.openstack.org/241642 | 15:16 |
| *** delattec has joined #puppet-openstack | 15:17 | |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-nova: Bump UCA and RDO to Liberty GA https://review.openstack.org/241643 | 15:18 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-sahara: Bump UCA and RDO to Liberty GA https://review.openstack.org/241644 | 15:20 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-swift: Bump UCA and RDO to Liberty GA https://review.openstack.org/241645 | 15:20 |
| *** cdelatte has quit IRC | 15:21 | |
| *** pabelanger has joined #puppet-openstack | 15:21 | |
| *** samuelBartel has joined #puppet-openstack | 15:21 | |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-trove: Bump UCA and RDO to Liberty GA https://review.openstack.org/241648 | 15:21 |
| *** dhollinger has quit IRC | 15:21 | |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-vswitch: Bump RDO to Liberty GA https://review.openstack.org/241649 | 15:23 |
| openstackgerrit | Merged openstack/puppet-monasca: Change default monasca api port to 8070 to avoid murano conflict https://review.openstack.org/241313 | 15:26 |
| *** dhollinger has joined #puppet-openstack | 15:27 | |
| *** rahulait has joined #puppet-openstack | 15:28 | |
| *** dhollinger has quit IRC | 15:28 | |
| *** dhollinger has joined #puppet-openstack | 15:29 | |
| *** zhangjn has joined #puppet-openstack | 15:32 | |
| openstackgerrit | Merged openstack/puppet-nova: do not manage python-greenlet anymore https://review.openstack.org/241373 | 15:33 |
| openstackgerrit | Alex Schultz proposed openstack/puppet-openstacklib: Prepare $::os_package_type https://review.openstack.org/241615 | 15:35 |
| mwhahaha | zigo: ok so i checked, a static fact in facts.d overrides that .rb file | 15:36 |
| mwhahaha | so it should be good now | 15:36 |
| EmilienM | mwhahaha: so we're creating the fact with oslib? | 15:37 |
| mwhahaha | no | 15:37 |
| EmilienM | what about the fact provided by packaging? | 15:37 |
| mwhahaha | so that's the default | 15:37 |
| mwhahaha | so the packaging would provide a facts.d file | 15:37 |
| mwhahaha | which would override what is set in oopenstacklib | 15:38 |
| mwhahaha | the openstacklib is the definition of the default (debian/ubuntu/rpm) based on osfamily | 15:38 |
| dfisher | (solaris::pkg) | 15:38 |
| mwhahaha | yea slowaris | 15:38 |
| dfisher | boo | 15:38 |
| mwhahaha | what's the osfamily package for that? | 15:38 |
| mwhahaha | i was going to add that in | 15:38 |
| dfisher | Solaris | 15:38 |
| dfisher | $ facter osfamily | 15:39 |
| dfisher | Solaris | 15:39 |
| mwhahaha | k | 15:39 |
| dfisher | and, thanks for adding that :) | 15:39 |
| dfisher | you get auto +1s from me forever. | 15:39 |
| EmilienM | crinkle: mgagne also suggested to add our Rakefile in a Gem on https://github.com/openstack/puppet-openstack_spec_helper | 15:42 |
| EmilienM | crinkle: beside what you wanted to Gemify with beaker | 15:42 |
| EmilienM | It makes sense to me, do you have some progress on this repo? Can we help? | 15:43 |
| *** paramite is now known as paramite|afk | 15:49 | |
| *** ntpttr has joined #puppet-openstack | 15:50 | |
| *** paramite|afk is now known as paramite | 15:51 | |
| openstackgerrit | Alex Schultz proposed openstack/puppet-openstacklib: Prepare $::os_package_type https://review.openstack.org/241615 | 15:52 |
| *** dgurtner has quit IRC | 15:53 | |
| openstackgerrit | Gael Chamoulaud proposed openstack/puppet-trove: Introduce trove::db class https://review.openstack.org/241666 | 15:56 |
| karimb | hello folks, the neutron::agents::ovs class seems to modify /etc/neutron/plugins/ml2/openvswitch_agent.ini and not /etc/neutron/plugins/openswitch/openvswitch_agent.ini | 16:01 |
| karimb | is that a change starting from kilo or something ? | 16:01 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-nova: Try to use zuul-cloner to prepare fixtures https://review.openstack.org/238623 | 16:01 |
| EmilienM | mgagne: ^ | 16:02 |
| openstackgerrit | Merged openstack/puppet-cinder: Try to use zuul-cloner to prepare fixtures https://review.openstack.org/238629 | 16:03 |
| openstackgerrit | Merged openstack/puppet-manila: Try to use zuul-cloner to prepare fixtures https://review.openstack.org/238625 | 16:03 |
| chem | EmilienM: when you have time could you have another look at https://review.openstack.org/#/c/226919/ | 16:05 |
| EmilienM | chem: that's big, I'll look this afternoon probably | 16:06 |
| chem | EmilienM: thanks a lot | 16:06 |
| *** zhangjn has quit IRC | 16:12 | |
| *** mdorman has joined #puppet-openstack | 16:13 | |
| *** trown is now known as trown|brb | 16:21 | |
| *** trown|brb is now known as trown | 16:24 | |
| *** dprince has quit IRC | 16:30 | |
| mfisch | anyone here used the policy.pp stuff before? | 16:32 |
| mfisch | in any module | 16:32 |
| EmilienM | mfisch: we did yes | 16:33 |
| mfisch | will the code overwrite existing stuff? | 16:33 |
| mfisch | for exmaple I want to change | 16:33 |
| mfisch | "deny_stack_user": "not role:heat_stack_user", | 16:33 |
| EmilienM | no I don't think so | 16:33 |
| EmilienM | lemme check | 16:33 |
| EmilienM | https://github.com/openstack/puppet-openstacklib/blob/master/manifests/policy/base.pp | 16:34 |
| EmilienM | it's using augeas, and looks for an existing value | 16:34 |
| EmilienM | so it's safe, and should not override anything | 16:34 |
| EmilienM | it adds the entry if not present, and then add the key/value in the entry | 16:35 |
| mfisch | I want it to overwrite | 16:39 |
| mfisch | I'll see what happens | 16:39 |
| mfisch | glad to see its augeas though so it should work | 16:39 |
| EmilienM | mfisch: it will work, we had the same use case | 16:39 |
| mfisch | I wonder if we should fix this issue upstream | 16:39 |
| *** ntpttr has quit IRC | 16:39 | |
| mfisch | you can change "heat_stack_user" but it does not affect policy.json | 16:39 |
| mfisch | I dont know if any modules really mess with policy.json | 16:40 |
| EmilienM | mfisch: first thing you can do is file a bug in heat | 16:40 |
| EmilienM | they'll tell you if it's intentional | 16:40 |
| EmilienM | mgagne: what do you suggest with https://review.openstack.org/#/c/238597/ ? | 16:40 |
| *** paramite has quit IRC | 16:40 | |
| mgagne | EmilienM: can't the section be moved outside like in master? | 16:41 |
| mgagne | it's currently within group :development, :test do | 16:41 |
| mgagne | which creates a nested group | 16:41 |
| EmilienM | mgagne: I can take the Gemfile from master | 16:41 |
| EmilienM | if that's what you want | 16:42 |
| mgagne | EmilienM: I didn't ask for that | 16:42 |
| mfisch | EmilienM: not sure why I'd file a bug in heat? | 16:42 |
| EmilienM | mfisch: for default value in policy.json | 16:42 |
| mgagne | EmilienM: https://github.com/openstack/puppet-manila/blob/stable/kilo/Gemfile#L24-L26 | 16:42 |
| mgagne | EmilienM: see how system_tests is not nested within :development, :test groups | 16:42 |
| EmilienM | oh ok | 16:42 |
| mgagne | EmilienM: you only need to move it 13 lines below | 16:43 |
| EmilienM | mgagne: so I should take the manila gemfile and copy it in the patch you -2'ed? | 16:43 |
| EmilienM | ok I got it | 16:43 |
| mgagne | EmilienM: you don,tneed to copy anything | 16:43 |
| EmilienM | sending a patch | 16:43 |
| EmilienM | kk | 16:43 |
| mfisch | EmilienM: the policy.json default is fine, the issue is when someone changing the setting with puppet, it changes heat.conf but not policy.json. But I dont know of any modules that change policy.json off hand | 16:45 |
| EmilienM | mgagne: actually, I'm not sure I get what you want. Can you push over my patch please? | 16:47 |
| EmilienM | oh no I get it I think | 16:47 |
| mgagne | o_O | 16:47 |
| EmilienM | I need to move out the group system_tests | 16:47 |
| mgagne | yea | 16:48 |
| EmilienM | to be outside the dev/test | 16:48 |
| EmilienM | I'm dumb. | 16:48 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-trove: Try to use zuul-cloner to prepare fixtures https://review.openstack.org/238597 | 16:49 |
| EmilienM | should be good now, mgagne ^ | 16:49 |
| *** dprince has joined #puppet-openstack | 16:50 | |
| EmilienM | mfisch: next week I'm in holidays, somewhere where Internet might not exist. | 16:51 |
| EmilienM | mfisch: are you willing to lead the weekly meeting? | 16:51 |
| mfisch | sure | 16:52 |
| EmilienM | mfisch: thanks | 16:52 |
| EmilienM | mfisch: I'll be here: https://goo.gl/maps/eNcpssXUZS92 - not sure I can have Wifi though | 16:53 |
| mfisch | nice | 16:53 |
| *** cwolferh has joined #puppet-openstack | 16:57 | |
| openstackgerrit | Merged openstack/puppet-openstack-integration: run_tests: install puppet from puppetlabs.com repo https://review.openstack.org/241340 | 16:59 |
| mfisch | EmilienM: are you using hiera to set any policies? | 17:00 |
| *** sergmelikyan has quit IRC | 17:00 | |
| EmilienM | mfisch: yes | 17:00 |
| openstackgerrit | Alex Schultz proposed openstack/puppet-cinder: Fix cinder default log_dir https://review.openstack.org/241691 | 17:00 |
| EmilienM | I can point you an example, a sec | 17:00 |
| mfisch | thanks, its not working for me | 17:01 |
| *** sergmelikyan has joined #puppet-openstack | 17:01 | |
| *** nlahouti has joined #puppet-openstack | 17:02 | |
| *** nlahouti has quit IRC | 17:02 | |
| *** nlahouti has joined #puppet-openstack | 17:03 | |
| *** dgurtner has joined #puppet-openstack | 17:03 | |
| *** dgurtner has quit IRC | 17:03 | |
| *** dgurtner has joined #puppet-openstack | 17:03 | |
| EmilienM | mfisch: i don't have it | 17:04 |
| EmilienM | but I can find an example in puppet- modules | 17:04 |
| *** nlahouti has quit IRC | 17:04 | |
| EmilienM | https://github.com/openstack/puppet-heat/blob/master/manifests/policy.pp#L11-L15 | 17:04 |
| mfisch | the code has examples but not hiera, I thnk I need to learn how to hiera ;) | 17:04 |
| mfisch | yep I have that open right now | 17:05 |
| _ody | Is there a chance that no all resources provided to nodepool are are configured 100% the same? | 17:05 |
| _ody | EmilienM: Some beaker tests seem to be failing on python-greenlet and others aren't. | 17:05 |
| EmilienM | _ody: I fixed the bug, it's should be fixed everywhere now | 17:06 |
| EmilienM | do recheck | 17:06 |
| *** sergmelikyan has quit IRC | 17:06 | |
| EmilienM | the patch landed this morning | 17:06 |
| _ody | EmilienM: What time? | 17:06 |
| _ody | Ah. Ok. | 17:06 |
| _ody | I did rechecks last night. | 17:06 |
| openstackgerrit | Athlan-Guyot sofer proposed openstack/puppet-keystone: Support for services different only by type. https://review.openstack.org/238954 | 17:06 |
| _ody | Thought you had already gotten it in. | 17:06 |
| EmilienM | _ody: unfortunatly not, I had to ping people to merge it this morning | 17:06 |
| EmilienM | mfisch: let me write it for you | 17:07 |
| mfisch | hahah | 17:07 |
| mfisch | its a hash of hashes which is probably my issue | 17:07 |
| openstackgerrit | Athlan-Guyot sofer proposed openstack/puppet-keystone: Keystone_endpoint match service by name/type. https://review.openstack.org/239434 | 17:07 |
| *** rahulait has quit IRC | 17:07 | |
| mfisch | we know hash here so I'll get it | 17:08 |
| *** jistr has quit IRC | 17:08 | |
| *** nlahouti has joined #puppet-openstack | 17:08 | |
| EmilienM | mfisch: http://paste.openstack.org/show/siLoTr5cFfGrQ2KVSqLs/ | 17:09 |
| EmilienM | it should work ^ | 17:09 |
| mfisch | thats similar to what I have | 17:10 |
| mfisch | there may be a bug in this, give me a minute to confirm | 17:10 |
| EmilienM | I would be surprised, we are using this code for quite a while | 17:11 |
| *** rahulait has joined #puppet-openstack | 17:12 | |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-nova: Try to use zuul-cloner to prepare fixtures https://review.openstack.org/238623 | 17:13 |
| EmilienM | mfisch: can you +A https://review.openstack.org/#/c/238597/ ? | 17:14 |
| mfisch | done | 17:14 |
| mfisch | Error: Could not retrieve catalog from remote server: Error 400 on SERVER: Invalid parameter context_is_admin on Openstacklib::Policy::Base[heat-context_is_admin] on node dev01-control-001.os.cloud.twc.net | 17:14 |
| mfisch | same issue all morning | 17:14 |
| EmilienM | weird | 17:14 |
| mfisch | heat::policy::policies: | 17:14 |
| mfisch | heat-context_is_admin: | 17:14 |
| mfisch | context_is_admin: 'true' | 17:14 |
| EmilienM | there is something else | 17:15 |
| EmilienM | oh I remember | 17:15 |
| EmilienM | you need to use key / value | 17:15 |
| mfisch | what do you mean? | 17:16 |
| EmilienM | mfisch: http://paste.openstack.org/show/477996/ | 17:16 |
| mfisch | ah | 17:17 |
| mfisch | literally "key" "value" | 17:17 |
| EmilienM | yeah | 17:17 |
| EmilienM | I think | 17:17 |
| EmilienM | optionaly, you can specify the policy_path in the hash | 17:18 |
| *** derekh has quit IRC | 17:18 | |
| EmilienM | but default should be /etc/heat/policy.json in your case | 17:18 |
| mfisch | that was it | 17:19 |
| mfisch | is the example puppet code right then? | 17:20 |
| EmilienM | mfisch: yes | 17:20 |
| EmilienM | mfisch: works for you? | 17:20 |
| mfisch | those map to "key" & "value" | 17:20 |
| mfisch | well catalog is compiling now, wiating on run | 17:20 |
| EmilienM | you should see the file and the service notify | 17:20 |
| EmilienM | api service should be restarted iirc | 17:21 |
| mfisch | I saw some notifies but no file change | 17:21 |
| mfisch | saw the restarts, let me look at logd | 17:21 |
| mfisch | logs | 17:21 |
| openstackgerrit | Merged openstack/puppet-trove: Try to use zuul-cloner to prepare fixtures https://review.openstack.org/238597 | 17:23 |
| *** Hunner has quit IRC | 17:25 | |
| *** _morgan has quit IRC | 17:26 | |
| mfisch | EmilienM: its notifying but not updating the file, let me try a new k/v pair | 17:27 |
| *** karimb has quit IRC | 17:29 | |
| *** _morgan has joined #puppet-openstack | 17:29 | |
| *** Hunner has joined #puppet-openstack | 17:30 | |
| *** Hunner has quit IRC | 17:30 | |
| *** Hunner has joined #puppet-openstack | 17:30 | |
| *** rahulait has quit IRC | 17:30 | |
| *** xarses has quit IRC | 17:32 | |
| *** xarses has joined #puppet-openstack | 17:32 | |
| *** xarses has quit IRC | 17:33 | |
| *** xarses has joined #puppet-openstack | 17:33 | |
| mfisch | EmilienM: its our own docker stuff that's breaking us | 17:36 |
| mfisch | we overwrite the file :( | 17:36 |
| *** dgurtner has quit IRC | 17:40 | |
| *** dgurtner has joined #puppet-openstack | 17:41 | |
| *** fedexo has joined #puppet-openstack | 17:46 | |
| openstackgerrit | Athlan-Guyot sofer proposed openstack/puppet-keystone: Keystone_endpoint match service by name/type. https://review.openstack.org/239434 | 17:47 |
| EmilienM | mfisch: :( | 17:49 |
| *** trown is now known as trown|lunch | 17:50 | |
| mfisch | EmilienM: do we rely on the aodh packages to make the aodh user? | 17:57 |
| EmilienM | mfisch: no, we try to stop managing POSIX users in our modules | 18:01 |
| EmilienM | mfisch: if you look at my patch the ordering makes sure we have the aodh-common package before creating the DB | 18:01 |
| EmilienM | if someone wants to change the user or manage it somewhere else, it's possible | 18:01 |
| crinkle | EmilienM: I haven't had time to work on it the gem, if you want to then go ahead, but i'm not sure that we really need it that critically | 18:08 |
| *** ddmitriev has quit IRC | 18:10 | |
| *** nlahouti has quit IRC | 18:17 | |
| *** xarses_ has joined #puppet-openstack | 18:19 | |
| *** xarses has quit IRC | 18:19 | |
| *** jerrygb has quit IRC | 18:21 | |
| *** jpena has quit IRC | 18:22 | |
| *** trown|lunch is now known as trown | 18:25 | |
| *** xarses has joined #puppet-openstack | 18:26 | |
| *** xarses_ has quit IRC | 18:26 | |
| *** dgurtner has quit IRC | 18:27 | |
| *** dgurtner has joined #puppet-openstack | 18:28 | |
| *** dgurtner has quit IRC | 18:28 | |
| *** dgurtner has joined #puppet-openstack | 18:28 | |
| EmilienM | crinkle: no problem, I'll look at it, though I never did :) | 18:28 |
| *** nlahouti has joined #puppet-openstack | 18:38 | |
| openstackgerrit | Adam Vinsh proposed openstack/puppet-swift: WIP multi port swift config https://review.openstack.org/241726 | 18:42 |
| EmilienM | mgagne, mfisch: the last one in the serie: https://review.openstack.org/#/c/238623/ - thanks :) | 18:43 |
| EmilienM | crinkle: you're right, that's not critical, but I think it would help to keep our repos consistent | 18:45 |
| *** gfidente is now known as gfidente|afk | 19:02 | |
| *** samuelBartel has quit IRC | 19:03 | |
| *** orionx has joined #puppet-openstack | 19:05 | |
| orionx | when i deploy a compute node, it gets the neutron-server service. that doesn't seem right to me. can someone confirm or deny if this is correct? i don't see where the compute module is pulling this in... | 19:07 |
| orionx | oh... maybe my compute node thinks it's a controller and `$::openstack::profile::base::is_controller` is evaling to true... | 19:11 |
| openstackgerrit | Merged openstack/puppet-nova: Try to use zuul-cloner to prepare fixtures https://review.openstack.org/238623 | 19:12 |
| *** dhollinger has quit IRC | 19:13 | |
| *** xavpaice has joined #puppet-openstack | 19:14 | |
| *** jerrygb has joined #puppet-openstack | 19:20 | |
| zigo | mwhahaha: Thanks a lot! lgtm then. | 19:20 |
| mwhahaha | no problem | 19:20 |
| zigo | EmilienM: Can you review the patch? | 19:21 |
| EmilienM | zigo: I will | 19:23 |
| EmilienM | I think I got it | 19:25 |
| EmilienM | zigo: I'm reviewing but it lgtm | 19:25 |
| *** xavpaice has left #puppet-openstack | 19:33 | |
| EmilienM | zigo, mwhahaha: could you patch puppet-nova with an example (and Depend-On) so we can see a use case? | 19:37 |
| mwhahaha | sure | 19:37 |
| mwhahaha | horizon is probably a better canidate | 19:37 |
| mwhahaha | but i'll take a stab at it here shortly | 19:38 |
| EmilienM | yeah horizon ok | 19:39 |
| mwhahaha | or maybe nova, i'll find something | 19:41 |
| *** ntpttr has joined #puppet-openstack | 19:41 | |
| mwhahaha | we should also look at https://github.com/mcanevet/rspec-puppet-facts for our operating system facts | 19:47 |
| EmilienM | indeed, that's nice | 19:52 |
| openstackgerrit | Alex Schultz proposed openstack/puppet-openstacklib: Prepare $::os_package_type https://review.openstack.org/241615 | 19:53 |
| mwhahaha | oh we can use this fact for the fedora and RH6/7 split too | 19:57 |
| mwhahaha | or not, maybe too feature creepy | 19:58 |
| openstackgerrit | Alex Schultz proposed openstack/puppet-nova: Switch nova to leverage os_package_type fact https://review.openstack.org/241741 | 20:00 |
| mwhahaha | EmilienM: -^ | 20:00 |
| EmilienM | mwhahaha: looking | 20:01 |
| openstackgerrit | Iury Gregory Melo Ferreira proposed openstack/puppet-keystone: Support for Keystone as Service Provider https://review.openstack.org/216821 | 20:01 |
| EmilienM | mwhahaha: well I'm not sure you'll be able to run Debian packaging on Ubuntu with the patches ^ | 20:02 |
| mwhahaha | which patches? | 20:03 |
| mwhahaha | or rather why not? | 20:03 |
| mwhahaha | so we do it today for fuel since our packages are debian based so we're doing this type of thing via other methods | 20:04 |
| mwhahaha | at least as I understand it | 20:04 |
| EmilienM | mwhahaha: when running debian packages on ubuntu, what is os_package_type ? 'ubuntu' ? | 20:04 |
| mwhahaha | we would use a package to define a fact to override it as 'debian' | 20:05 |
| EmilienM | ah ok | 20:05 |
| EmilienM | then, ok | 20:05 |
| mwhahaha | by default if no such fact is applied, you would get the ubuntu packages on ubuntu because os_package_type would be ubuntu | 20:05 |
| mwhahaha | but this allows us to define that we want *debian* stuff even if operatingsystem == 'Ubuntu' | 20:06 |
| EmilienM | I see ok | 20:06 |
| mwhahaha | and it's targeted to just openstack stuff, rather than trying to override operatingsystem as that'll probably jack with a whole bunch of other stuff | 20:06 |
| EmilienM | I would be interested to know if people would like to switch from UCA to debian packaging on ubuntu jobs | 20:07 |
| EmilienM | I personnaly +1 | 20:07 |
| EmilienM | I know fuel folks would +1 too, but I'm curious about others | 20:07 |
| zigo | EmilienM: mwhahaha: I can do the patch. What's the way to do the depends on? | 20:07 |
| mwhahaha | zigo: i got it | 20:07 |
| mwhahaha | zigo: https://review.openstack.org/241741 | 20:08 |
| zigo | mwhahaha: There's a few things that also needs update there. | 20:08 |
| mwhahaha | sure, you can take a look at that and adjust if you know of other items | 20:08 |
| zigo | Like log group is now adm. | 20:08 |
| mwhahaha | for debian? so that matches ubuntu now? | 20:09 |
| zigo | mwhahaha: Yup. | 20:13 |
| *** dhollinger has joined #puppet-openstack | 20:13 | |
| zigo | mwhahaha: This was fixed within openstack-pkg-tools, and now absolutely all packages are using adm for /var/log/PROJECT | 20:13 |
| mwhahaha | ah ok, that might just need to be a separate change but we can just move that out of that case block | 20:14 |
| zigo | mwhahaha: I don't understand why you're doing: | 20:14 |
| zigo | context 'on Ubuntu system' do | 20:14 |
| zigo | [...] | 20:14 |
| zigo | :os_package_type => 'ubuntu' | 20:14 |
| zigo | Isn't this overriding what we've done? | 20:14 |
| mwhahaha | so that's for the tests which we have to mock out | 20:15 |
| zigo | Ah... | 20:15 |
| mwhahaha | tests don't automagically get facts applied so we have to specify them | 20:15 |
| zigo | That's beyond my knowledge then. | 20:15 |
| mwhahaha | yup which is why i'm helping :) | 20:15 |
| zigo | Oh, I see it in the rest of the patch. It makes sense now... | 20:16 |
| zigo | I'll ignore the $nova_log_group for now, and I'll fix it in another patch (after that one is approved, to avoid merge conflicts). | 20:16 |
| EmilienM | can I troll? | 20:21 |
| iurygregory | lol | 20:21 |
| EmilienM | zigo doing puppet is like http://goo.gl/d7O8YV | 20:21 |
| iurygregory | hasuahsuahsuashaush omg | 20:22 |
| zigo | :) | 20:23 |
| *** angdraug has joined #puppet-openstack | 20:23 | |
| openstackgerrit | Thomas Goirand proposed openstack/puppet-horizon: Do not run collectstatic or compress in Debian https://review.openstack.org/241218 | 20:23 |
| iurygregory | hey EmilienM , can you give me a hand on https://review.openstack.org/#/c/216821/ ? | 20:24 |
| EmilienM | iurygregory: sure, what's up? | 20:25 |
| iurygregory | trying to make the concat work =/ | 20:25 |
| openstackgerrit | Merged openstack/puppet-aodh: Implement Evaluator service https://review.openstack.org/237662 | 20:26 |
| openstackgerrit | Merged openstack/puppet-aodh: Implement Client class https://review.openstack.org/237706 | 20:26 |
| *** khomkrit has quit IRC | 20:26 | |
| openstackgerrit | Merged openstack/puppet-aodh: Implement Notifier service https://review.openstack.org/237708 | 20:26 |
| iurygregory | i've got 1;31mError: Duplicate declaration: File[/etc/apache2/sites-available] is already declared in file /etc/puppet/modules/apache/manifests/init.pp:214; cannot redeclare at /etc/puppet/modules/concat/manifests/init.pp:195 on node devstack-trusty-rax-iad-5634336[0m | 20:26 |
| iurygregory | and now Duplicate declaration: Class[Keystone::Wsgi::Apache] is already declared in file =/ | 20:26 |
| EmilienM | you should not declare class {'::keystone::wsgi::apache': } in the manifest | 20:27 |
| EmilienM | it's done by our users already | 20:27 |
| EmilienM | though you should make sure somehow the custom_fragment is used | 20:27 |
| iurygregory | yeah, i'm lost in how to make sure the file is changed =/ | 20:28 |
| *** rahulait has joined #puppet-openstack | 20:34 | |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-aodh: Implement Listener service https://review.openstack.org/237709 | 20:37 |
| *** dgurtner has quit IRC | 20:37 | |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-aodh: Tag all Aodh packages https://review.openstack.org/239290 | 20:38 |
| *** dgurtner has joined #puppet-openstack | 20:39 | |
| *** dgurtner has quit IRC | 20:39 | |
| *** dgurtner has joined #puppet-openstack | 20:39 | |
| EmilienM | iurygregory: so if you need the users pass a specific parameter, I think the best thing to do is documentation | 20:40 |
| *** nlahouti has quit IRC | 20:40 | |
| EmilienM | you can also testing the resource in the manifest | 20:40 |
| iurygregory | oh ok ;) i'll try to use the custom fragment in the manifest | 20:41 |
| *** trown is now known as trown|hairscut | 20:48 | |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-gnocchi: Implement API service & dependencies https://review.openstack.org/238811 | 20:48 |
| *** bklei_ has joined #puppet-openstack | 20:52 | |
| *** bklei has quit IRC | 20:55 | |
| *** karimb has joined #puppet-openstack | 21:02 | |
| openstackgerrit | Thomas Goirand proposed openstack/puppet-horizon: Do not run collectstatic or compress in Debian https://review.openstack.org/241218 | 21:02 |
| openstackgerrit | Merged openstack/puppet-aodh: Implement Listener service https://review.openstack.org/237709 | 21:04 |
| *** dgurtner has quit IRC | 21:09 | |
| *** dgurtner has joined #puppet-openstack | 21:11 | |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-gnocchi: Implement Gnocchi Statsd support https://review.openstack.org/240663 | 21:18 |
| karmix | EmilienM: If you have a minute to respond to https://review.openstack.org/#/c/241466 , I need to know how to proceed. Thanks. | 21:18 |
| EmilienM | karmix: looking | 21:18 |
| EmilienM | karmix: you should abandon your patch and submit it against master | 21:19 |
| EmilienM | but I don't think your patch is really good | 21:19 |
| EmilienM | I mean, the require | 21:19 |
| openstackgerrit | Merged openstack/puppet-keystone: keystone_endpoint provider for Keystone v3 api https://review.openstack.org/178456 | 21:20 |
| karmix | What whould be better? | 21:20 |
| EmilienM | because in some cases, neutron-server does not run on the same node as keystone | 21:20 |
| EmilienM | so the Keystone_user_role resource won't be in the catalog | 21:20 |
| karmix | I could use a collector for that. | 21:20 |
| EmilienM | yeah | 21:21 |
| karmix | Is the keystone user role applied on the keystone server, and not the neutron server, then? | 21:21 |
| EmilienM | karmix: why don't we have the issue in our CI? | 21:21 |
| EmilienM | karmix: we're using the class in our CI and we don't have the bug | 21:21 |
| openstackgerrit | Colleen Murphy proposed openstack/puppet-openstack_spec_helper: Initial commit - lay down basic structure https://review.openstack.org/241772 | 21:21 |
| karmix | It's the same as the other ones, just lucky. I don't think CI is using 4.2, either, and it looks like the ordering algorithm may have changed recently to match the manifest ordering. | 21:22 |
| EmilienM | we're using 3.8 for beaker & integration jobs | 21:22 |
| EmilienM | Hunner and _ody might know about that kind of change | 21:22 |
| EmilienM | I don't think we're lucky, it never failed because of this until now. | 21:23 |
| karmix | I know that in 3.8, the order is very different. Give me a sec, and I'll show you the links I found. | 21:23 |
| karmix | I say lucky because in 3.8, the docs state that puppet chooses whatever order it thinks is best given the environment. It worked in our CI environment, because puppet didn't happen to choose a bad order in that configuration. | 21:24 |
| *** jayg is now known as jayg|g0n3 | 21:24 | |
| karmix | It's in the first paragraph of the docs. Here's the change. | 21:25 |
| karmix | For 3.8: https://docs.puppetlabs.com/puppet/3.8/reference/lang_relationships.html | 21:25 |
| karmix | For 4.2 (which I'm using): https://docs.puppetlabs.com/puppet/3.8/reference/lang_relationships.html | 21:26 |
| *** bklei_ has quit IRC | 21:26 | |
| karmix | ack. 4.2 was: https://docs.puppetlabs.com/puppet/4.2/reference/lang_relationships.html | 21:26 |
| *** bklei has joined #puppet-openstack | 21:27 | |
| openstackgerrit | Merged openstack/puppet-keystone: Add composite namevar for tenant, user, user_role. https://review.openstack.org/226919 | 21:28 |
| EmilienM | karmix: good point, then your patch is valid. So I suggest you submit it against master and then use a collector | 21:28 |
| EmilienM | richm: the namevar patch is merged, I'm testing the CI to see if we did not break anything | 21:28 |
| openstackgerrit | Emilien Macchi proposed openstack/puppet-neutron: DNM - CI test https://review.openstack.org/241774 | 21:28 |
| karmix | EmilienM: Since the code was removed from master, are you asking me to patch pre-919a3a93 then merge the changes back out? | 21:29 |
| EmilienM | karmix: removed from master ? where | 21:29 |
| karmix | in https://github.com/openstack/puppet-neutron/commit/919a3a932eccff507a8f974512cb8118ec6609e9 | 21:30 |
| EmilienM | oh I get it now | 21:30 |
| EmilienM | can we backport the patch? | 21:30 |
| EmilienM | ie: does kilo has nova_admin_tenant_name ? | 21:31 |
| karmix | Like I said in the review, it would change the module interface, which we probably don't want to do on a stable branch. | 21:31 |
| EmilienM | no it does not change the interface | 21:31 |
| karmix | Yes, the kilo docs say it's there (though it wasn't in the config file I initially looked at). | 21:31 |
| EmilienM | there is no param change | 21:31 |
| richm | EmilienM: re: namevar - ack | 21:32 |
| *** trown|hairscut is now known as trown | 21:32 | |
| karmix | EmilienM: But if anybody was using the type in their manifests, it would make their manifests invalid. | 21:32 |
| karmix | Looking up a tennant id seems kinda useful, so I wouldn't put it past people to use it directly. And I don't see anything declaring it as for internal use only. | 21:33 |
| *** nlahouti has joined #puppet-openstack | 21:33 | |
| openstackgerrit | Iury Gregory Melo Ferreira proposed openstack/puppet-keystone: Support for Keystone as Service Provider https://review.openstack.org/216821 | 21:36 |
| *** jerrygb has quit IRC | 21:36 | |
| *** jerrygb has joined #puppet-openstack | 21:38 | |
| *** gfidente|afk has quit IRC | 21:38 | |
| openstackgerrit | Merged openstack/puppet-aodh: Tag all Aodh packages https://review.openstack.org/239290 | 21:40 |
| karmix | EmilienM: You are right, this is probably safe. In this case, it would be very difficult to use the custom type for anything useful outside the class. | 21:40 |
| *** jtomasek has quit IRC | 21:41 | |
| karmix | I was thinking it was more generic, where you gave it the config setting and credentials and it would lookup the tenant id then set that value. | 21:42 |
| *** fedexo has quit IRC | 21:44 | |
| EmilienM | karmix: I guess we can backport the patch then | 21:44 |
| karmix | EmilienM: I think so. I'm looking into that, now. | 21:46 |
| *** xarses has quit IRC | 21:47 | |
| *** gildub has joined #puppet-openstack | 22:04 | |
| karmix | EmilienM: BTW, the resource ordering in 4.2 means that people will have dependency issues that CI didn't encounter simply because the resources appear in their manifest in a different order than CI used. This is going to make it more important to test dependencies during acceptance testing. | 22:09 |
| *** rahulait has quit IRC | 22:19 | |
| *** dgurtner has quit IRC | 22:26 | |
| *** bklei_ has joined #puppet-openstack | 22:26 | |
| *** bklei has quit IRC | 22:29 | |
| *** bklei_ has quit IRC | 22:51 | |
| *** bklei has joined #puppet-openstack | 22:52 | |
| *** ntpttr has left #puppet-openstack | 22:53 | |
| *** dprince has quit IRC | 22:58 | |
| *** orionx has quit IRC | 23:09 | |
| *** trown is now known as trown|outttypeww | 23:14 | |
| *** mjblack has quit IRC | 23:23 | |
| *** mjblack has joined #puppet-openstack | 23:27 | |
| *** ducttape_ has quit IRC | 23:30 | |
| *** bklei has quit IRC | 23:31 | |
| Hunner | crinkle: beaker is out | 23:37 |
| crinkle | Hunner: woot | 23:38 |
| karmix | ?? | 23:38 |
| karmix | I'm pretty sure that has been out for a while. :) | 23:38 |
| *** zhangjn has joined #puppet-openstack | 23:39 | |
| *** nlahouti has quit IRC | 23:49 | |
| *** nlahouti has joined #puppet-openstack | 23:51 | |
| *** jerrygb has quit IRC | 23:56 | |
| *** zhangjn has quit IRC | 23:57 | |
| Hunner | karmix: I mean latest release is out; she was waiting for it | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!