Time |
Nick |
Message |
00:48 |
|
xingchao_ joined #puppet-openstack |
00:57 |
|
stevemar joined #puppet-openstack |
01:25 |
|
xingchao_ joined #puppet-openstack |
01:54 |
|
xarses joined #puppet-openstack |
02:03 |
|
khappone joined #puppet-openstack |
02:08 |
|
mgagne joined #puppet-openstack |
02:19 |
|
ericpeterson joined #puppet-openstack |
02:50 |
|
xingchao joined #puppet-openstack |
03:42 |
|
ericpeterson joined #puppet-openstack |
03:43 |
|
ericpeterson joined #puppet-openstack |
03:52 |
|
xingchao joined #puppet-openstack |
04:43 |
|
ericpeterson joined #puppet-openstack |
04:58 |
|
xingchao joined #puppet-openstack |
05:44 |
|
ddmitriev joined #puppet-openstack |
05:44 |
|
ericpeterson joined #puppet-openstack |
05:58 |
|
zerda joined #puppet-openstack |
06:03 |
|
xingchao joined #puppet-openstack |
06:04 |
openstackgerrit |
Gael Chamoulaud proposed stackforge/puppet-neutron: Configure OVS mechanism agent configs in its config file https://review.openstack.org/106144 |
06:52 |
|
saju_m joined #puppet-openstack |
07:08 |
|
xingchao joined #puppet-openstack |
07:11 |
|
mmagr joined #puppet-openstack |
07:11 |
|
stamak joined #puppet-openstack |
07:16 |
gchamoul |
morning |
07:17 |
|
tfz joined #puppet-openstack |
07:49 |
|
jpena joined #puppet-openstack |
07:56 |
|
tfz joined #puppet-openstack |
08:02 |
|
jistr joined #puppet-openstack |
08:09 |
|
xingchao joined #puppet-openstack |
08:12 |
|
IBerezovskiy joined #puppet-openstack |
08:16 |
|
holser_ joined #puppet-openstack |
08:20 |
|
tfz joined #puppet-openstack |
08:27 |
|
gfidente joined #puppet-openstack |
08:29 |
|
arnaud_orange joined #puppet-openstack |
08:48 |
|
stamak joined #puppet-openstack |
09:06 |
openstackgerrit |
Spencer Krum proposed stackforge/puppet-keystone: Beaker tests https://review.openstack.org/155873 |
09:22 |
|
ericpeterson joined #puppet-openstack |
09:36 |
|
andriikolesnikov joined #puppet-openstack |
09:50 |
|
saneax joined #puppet-openstack |
10:01 |
|
saneax joined #puppet-openstack |
10:06 |
|
Hunner joined #puppet-openstack |
10:22 |
|
xingchao joined #puppet-openstack |
10:25 |
|
ferest joined #puppet-openstack |
11:12 |
|
ericpeterson joined #puppet-openstack |
11:27 |
|
xingchao joined #puppet-openstack |
11:29 |
|
dprince joined #puppet-openstack |
11:32 |
|
ericpeterson joined #puppet-openstack |
11:38 |
|
sanjayu joined #puppet-openstack |
11:43 |
|
ericpeterson joined #puppet-openstack |
11:59 |
|
gfidente joined #puppet-openstack |
11:59 |
|
gfidente joined #puppet-openstack |
12:05 |
|
morazi joined #puppet-openstack |
12:32 |
|
ericpeterson joined #puppet-openstack |
12:32 |
|
xingchao joined #puppet-openstack |
12:49 |
gchamoul |
EmilienM: I updated the ovs patch as asked by Mathieu, but the tests are failing with puppet 3.6/3.7 on the gate ... but not in my laptop! |
12:52 |
|
ericpeterson joined #puppet-openstack |
12:53 |
EmilienM |
gchamoul: ok thanks, I'll look today |
12:54 |
|
jpena joined #puppet-openstack |
13:02 |
|
rcallawa joined #puppet-openstack |
13:17 |
|
sanjayu joined #puppet-openstack |
13:19 |
|
stevemar joined #puppet-openstack |
13:20 |
|
richm joined #puppet-openstack |
13:24 |
|
cdent joined #puppet-openstack |
13:25 |
|
jpena joined #puppet-openstack |
13:27 |
|
rcallawa_ joined #puppet-openstack |
13:28 |
|
stevemar joined #puppet-openstack |
13:38 |
|
xingchao joined #puppet-openstack |
13:54 |
|
dgurtner joined #puppet-openstack |
13:54 |
|
dgurtner joined #puppet-openstack |
13:54 |
|
jistr joined #puppet-openstack |
14:12 |
|
jistr joined #puppet-openstack |
14:15 |
|
ericpeterson joined #puppet-openstack |
14:24 |
|
tfz joined #puppet-openstack |
14:26 |
|
mdorman joined #puppet-openstack |
14:37 |
dmsimard |
what's that about me making weird noises :p |
14:38 |
|
ericpeterson joined #puppet-openstack |
14:43 |
|
mmagr joined #puppet-openstack |
14:43 |
|
xingchao joined #puppet-openstack |
14:44 |
|
vinsh joined #puppet-openstack |
14:51 |
|
jistr joined #puppet-openstack |
14:52 |
richm |
hello - puppet question for puppet gurus |
14:52 |
richm |
http://paste.openstack.org/show/200464/ |
14:52 |
richm |
Basically, at the puppet level, I need to get the id of the domain |
14:52 |
richm |
Which will be either in the output of openstack domain show $name |
14:53 |
richm |
or the value of the keystone_config "identity/default_domain_id" parameter |
14:53 |
|
ericpeterson joined #puppet-openstack |
14:53 |
richm |
I'm not sure if the above code will actually work, which is a gross misuse of inline_template |
14:54 |
richm |
running spec tests gives this error: |
14:54 |
richm |
Failed to parse inline template: No credentials provided. at /home/rmeggins/puppet-keystone/spec/fixtures/modules/keystone/manifests/init.pp:886 on node localhost.localdomain |
14:54 |
richm |
I guess it's trying to call openstack domain show with no credentials |
14:55 |
richm |
Line 26 in the paste is my line 886 |
14:56 |
richm |
So my question is: When is the inline_template evaulated? Is it evaluated before the domain is created e.g. at parse time? |
15:13 |
|
stevemar joined #puppet-openstack |
15:19 |
|
mschmitt joined #puppet-openstack |
15:20 |
|
nosleep77 joined #puppet-openstack |
15:30 |
|
ericpeterson joined #puppet-openstack |
15:30 |
|
dprince joined #puppet-openstack |
15:32 |
crinkle |
richm: you can't reach in to puppet internals from a puppet manifest |
15:33 |
richm |
crinkle: ok |
15:37 |
|
rwsu joined #puppet-openstack |
15:44 |
|
ozialien joined #puppet-openstack |
15:48 |
|
xingchao joined #puppet-openstack |
15:59 |
|
ozialien joined #puppet-openstack |
16:02 |
|
ericpeterson joined #puppet-openstack |
16:03 |
|
ericpeterson joined #puppet-openstack |
16:05 |
|
ozialien joined #puppet-openstack |
16:07 |
richm |
crinkle: Then my next approach is to create two "hybrid" resources |
16:08 |
richm |
crinkle: The first will be "keystone_default_domain" which will combine keystone_domain and keystone_config |
16:08 |
richm |
crinkle: The second will be keystone_domain_policy which will combine keystone_domain, keystone_config, and keystone::policy |
16:09 |
richm |
keystone_default_domain will mostly be keystone_domain, but will call keystone_config 'identity/default_domain_id' value => id of the default domain |
16:10 |
openstackgerrit |
Colleen Murphy proposed stackforge/puppet-keystone: Fix keystone unit tests https://review.openstack.org/171717 |
16:10 |
richm |
keystone_domain_policy will combine keystone_domain (for the admin domain), keystone_config (if no admin domain was specified, to get the id of the default domain from keystone_config 'identity/default_domain_id' value), and keystone::policy (to set the admin_domain_id in the policy) |
16:18 |
crinkle |
richm: what is the use case for making the default domain something other than 'default'? is that a critical feature? |
16:19 |
|
ozialien joined #puppet-openstack |
16:21 |
crinkle |
richm: for this part "keystone_config (if no admin domain was specified, to get the id of the default domain from keystone_config 'identity/default_domain_id' value)" that would be basically the same as getting the admin token from keystone.conf, so it doesn't need to be "hybrid", it just needs a method similar to get_admin_token |
16:24 |
crinkle |
richm: I'm looking at https://review.openstack.org/#/c/146926/20/lib/puppet/provider/keystone_domain/openstack.rb,cm again, it doesn't look like it actually puts the new domain_id in the policy file |
16:25 |
crinkle |
if there isn't actually any data that needs to be changed that's specific to the domain then i don't see why it needs to be done as part of a provider |
16:35 |
gfidente |
guys, I know it is terrible to poke for reviews but we're trying to use this from tripleo https://review.openstack.org/#/c/170407/ , so if anyone could help that would be nice :) |
16:36 |
gfidente |
*trying to use this from tripleo today |
16:45 |
crinkle |
gfidente: commented, did you test this in the case where the parameter is undef? |
16:45 |
|
rcallawa joined #puppet-openstack |
16:45 |
gfidente |
crinkle, yes, it takes the known route and it actually works when both assumptions it makes are true |
16:46 |
gfidente |
so when the ceph cluster is up already and the admin keyring is provisioned |
16:46 |
gfidente |
crinkle, thanks for looking into it |
16:47 |
gfidente |
crinkle, there are tests for the old route as well, passing as expected |
16:48 |
richm |
crinkle: The use case for making the default domain something other than 'default' is using an external LDAP server as the default identity store, and putting the service accounts into a 'services' domain |
16:48 |
gfidente |
I think it is expanding ${rbd_keyring} when defining the value of $libvirt_key, which is what I wanted |
16:49 |
crinkle |
gfidente: +a |
16:49 |
openstackgerrit |
Ryan Hefner proposed stackforge/puppet-manila: Support network plugin model https://review.openstack.org/171737 |
16:49 |
gfidente |
crinkle, but yes using single quotes there was probably more elegant |
16:49 |
gfidente |
thanks for helping, owe you beer |
16:51 |
|
xingchao joined #puppet-openstack |
16:56 |
openstackgerrit |
Merged stackforge/puppet-nova: Allow libvirt secret key setting from param https://review.openstack.org/170407 |
17:00 |
|
andriikolesnikov joined #puppet-openstack |
17:05 |
|
ericpeterson joined #puppet-openstack |
17:13 |
|
ericpeterson joined #puppet-openstack |
17:22 |
|
ozialien joined #puppet-openstack |
17:29 |
|
mgagne joined #puppet-openstack |
17:34 |
|
xarses joined #puppet-openstack |
17:35 |
openstackgerrit |
Spencer Krum proposed stackforge/puppet-keystone: Beaker tests https://review.openstack.org/155873 |
17:46 |
crinkle |
richm: so am I reading https://review.openstack.org/#/c/146926/20/lib/puppet/provider/keystone_domain/openstack.rb,cm wrong? based on yesterday's conversation i would expect domain_id:admin_domain_id to be changed, but it's only changing the is_admin:1 rule |
18:00 |
richm |
crinkle: it does both |
18:01 |
richm |
lines 40-41 add the is_admin:1 |
18:01 |
richm |
line 52 changes "admin_domain_id" to the real admin domain id |
18:01 |
richm |
crinkle: I wouldn't bother with reviewing that code for now - my next commit will be radically different |
18:02 |
crinkle |
okay |
18:13 |
openstackgerrit |
Ryan Hefner proposed stackforge/puppet-manila: Support network plugin model https://review.openstack.org/171737 |
18:14 |
|
ToMiles joined #puppet-openstack |
18:30 |
|
ericpeterson joined #puppet-openstack |
18:52 |
vinsh |
Anyone know if the stackforge puppet-swift has ever had multi-daemon config templates? |
18:52 |
vinsh |
I see the folks over at NeCTAR have a fork that has all of this. |
18:52 |
vinsh |
Wondered if it had been brought over yet.. or had not for some reason. |
19:04 |
openstackgerrit |
Ryan Hefner proposed stackforge/puppet-manila: Update share driver parameters https://review.openstack.org/171778 |
19:16 |
|
mmagr joined #puppet-openstack |
19:16 |
|
tfz joined #puppet-openstack |
19:33 |
openstackgerrit |
Matt Fischer proposed stackforge/puppet-glance: Fix API/Registry ensure for Ubuntu https://review.openstack.org/168082 |
19:35 |
|
rcallawa joined #puppet-openstack |
19:38 |
|
tfz joined #puppet-openstack |
20:03 |
|
khappone_ joined #puppet-openstack |
20:07 |
|
xingchao joined #puppet-openstack |
20:08 |
|
stevemar joined #puppet-openstack |
20:08 |
|
ericpeterson joined #puppet-openstack |
20:11 |
richm |
I'm trying to write a spec test for my new keystone_config_default_domain_id resource |
20:12 |
richm |
I get this error: http://paste.openstack.org/show/200671/ |
20:13 |
richm |
The error is here, in provider/keystone_config_default_domain_id/ini_setting.rb: |
20:13 |
richm |
:parent => Puppet::Type.type(:keystone_config).provider(:ini_setting) |
20:13 |
richm |
It is saying that Puppet::Type.type(:keystone_config) is nil |
20:13 |
richm |
Why can't puppet find the type? |
20:14 |
richm |
I even tried to add a spec test for keystone_config, thinking that perhaps I did something wrong |
20:14 |
EmilienM |
richm: do you have the code of the test somewhere? |
20:15 |
richm |
EmilienM: https://rmeggins.fedorapeople.org/ |
20:16 |
richm |
EmilienM: The new type is keystone_config_default_domain_id.rb - the provider is ini_setting.rb - the test is ini_setting_spec.rb |
20:16 |
richm |
EmilienM: If you want to try with the existing keystone_config class |
20:17 |
richm |
https://rmeggins.fedorapeople.org/ini_setting_spec.rb.keystone_config |
20:17 |
richm |
mkdir spec/unit/provider/keystone_config |
20:18 |
richm |
then copy that file to spec/unit/provider/keystone_config/ini_setting_spec.rb |
20:18 |
richm |
then GEM_HOME=vendor bundle exec rake spec SPEC=spec/unit/provider/keystone_config/ini_setting_spec.rb |
20:19 |
richm |
boom - explodes |
20:19 |
EmilienM |
indeed |
20:20 |
EmilienM |
richm: can you send the patch to gerrit so we can have it online? |
20:20 |
richm |
ok |
20:22 |
openstackgerrit |
Richard Megginson proposed stackforge/puppet-keystone: support for keystone v3 api https://review.openstack.org/146926 |
20:23 |
EmilienM |
what a patch \o/ |
20:23 |
richm |
ugh - please ignore the ~ file |
20:23 |
richm |
s |
20:23 |
EmilienM |
richm: you have emacs ? :P |
20:23 |
richm |
EmilienM: It's only going to ge uglier from here . . . |
20:24 |
richm |
EmilienM: yes, as there appears to be no decent IDE for ruby |
20:24 |
richm |
+ puppet |
20:24 |
EmilienM |
vi is okay |
20:24 |
EmilienM |
richm: so this is in https://review.openstack.org/#/c/146926/21/spec/unit/provider/keystone_config_default_domain_id/ini_setting_spec.rb,cm |
20:25 |
richm |
yes |
20:25 |
richm |
but |
20:25 |
richm |
it's not my code |
20:25 |
richm |
I can do something as simple as adding spec/unit/provider/keystone_config/ini_setting_spec.rb |
20:25 |
richm |
And I will get the same error |
20:27 |
richm |
Is this another instance of # LP#1408531/ |
20:27 |
richm |
? |
20:27 |
richm |
I tried adding the $LOAD_PATH hack to type/keystone_config.rb |
20:27 |
richm |
but that did not help |
20:31 |
richm |
are you able to reproduce the problem with just spec/unit/provider/keystone_config/ini_setting_spec.rb? |
20:34 |
|
stevemar joined #puppet-openstack |
20:40 |
|
tfz joined #puppet-openstack |
20:42 |
|
stevemar joined #puppet-openstack |
20:49 |
EmilienM |
richm: I'm looking now, sorry I was stuck on another problem |
20:50 |
EmilienM |
richm: I got the same |
20:50 |
richm |
what's going on???? |
20:52 |
EmilienM |
richm: I may have something |
20:52 |
EmilienM |
let me some time to debug |
20:53 |
richm |
EmilienM: merci |
20:53 |
EmilienM |
richm: BTW, value is never used |
20:53 |
EmilienM |
or maybe I missed it |
20:54 |
richm |
EmilienM: right - it is never used - I just did a copy/paste from keystone_config/ini_setting.rb - same problem there |
20:56 |
|
ericpeterson joined #puppet-openstack |
20:58 |
|
ericpeterson joined #puppet-openstack |
21:01 |
EmilienM |
richm: got it |
21:01 |
EmilienM |
can I send? |
21:02 |
openstackgerrit |
Emilien Macchi proposed stackforge/puppet-keystone: support for keystone v3 api https://review.openstack.org/146926 |
21:02 |
EmilienM |
ok I send |
21:03 |
EmilienM |
https://review.openstack.org/#/c/146926/21..22/spec/unit/provider/keystone_config_default_domain_id/ini_setting_spec.rb,cm |
21:03 |
|
dgurtner joined #puppet-openstack |
21:03 |
|
dgurtner joined #puppet-openstack |
21:05 |
|
rcallawa joined #puppet-openstack |
21:07 |
richm |
EmilienM: wow - ok - so why isn't that set by default? |
21:08 |
EmilienM |
I have no idea. |
21:08 |
EmilienM |
richm: if you need anything else, let me know. I would need some time to review your patch but AFIK crinkle already did most of the work of reviewing |
21:10 |
richm |
EmilienM: I'm working to address the review comments from crinkle, imcsk8, gilles |
21:12 |
|
rcallawa joined #puppet-openstack |
21:12 |
|
xingchao joined #puppet-openstack |
21:13 |
richm |
EmilienM: merci beaucoup for the fix |
21:14 |
EmilienM |
well, I'm not very proud of this.. at least it helps |
21:43 |
|
tfz joined #puppet-openstack |
21:45 |
|
cwolferh joined #puppet-openstack |
21:59 |
|
ddmitriev joined #puppet-openstack |
22:01 |
|
cwolferh joined #puppet-openstack |
22:04 |
|
xingchao joined #puppet-openstack |
22:20 |
|
gchamoul joined #puppet-openstack |
22:21 |
|
ericpeterson joined #puppet-openstack |
22:24 |
|
gchamoul- joined #puppet-openstack |
22:26 |
|
ericpeterson joined #puppet-openstack |
22:29 |
|
gchamoul joined #puppet-openstack |
22:58 |
|
cwolferh joined #puppet-openstack |
23:10 |
openstackgerrit |
Adam Vinsh proposed stackforge/puppet-swift: read_affinity requires affinity sorting_method https://review.openstack.org/171867 |
23:13 |
|
gildub joined #puppet-openstack |
23:28 |
openstackgerrit |
David Gurtner proposed stackforge/puppet-nova: Refactor libvirt migration to be configurable https://review.openstack.org/171877 |
23:35 |
openstackgerrit |
David Gurtner proposed stackforge/puppet-nova: Make libvirt migration security configurable https://review.openstack.org/171877 |
23:36 |
dgurtner |
mgagne: ping re: https://review.openstack.org/#/c/171877/1/manifests/compute/libvirt.pp |
23:36 |
mgagne |
sup |
23:36 |
dgurtner |
not entirely sure I understand your proposal |
23:37 |
mgagne |
dgurtner: only add parameters to nova::migration::libvirt. |
23:37 |
mgagne |
dgurtner: and ask users to use Hiera instead |
23:38 |
dgurtner |
ok |
23:38 |
mgagne |
dgurtner: or define class { 'nova::migration::libvirt': use_tls => true } before including nova::compute::libvirt |
23:38 |
dgurtner |
mgagne: yup, my worry is exactly that "before" part |
23:39 |
mgagne |
dgurtner: afaik, there is no other way to do it if you don't want to use Hiera. |
23:39 |
mgagne |
dgurtner: our past experience showed that we should avoid nested classes as much as possible |
23:40 |
mgagne |
dgurtner: see deprecated puppet-openstack =) |
23:40 |
dgurtner |
mgagne: ok. I'll change it |
23:40 |
|
gchamoul- joined #puppet-openstack |
23:46 |
openstackgerrit |
David Gurtner proposed stackforge/puppet-nova: Make libvirt migration security configurable https://review.openstack.org/171877 |
23:50 |
mgagne |
dgurtner: I like it better. I will have to cross-check with our use of libvirt to see if the proposed implementation doesn't restrict libvirt configurability. |
23:55 |
dgurtner |
mgagne: sure, just let me know - I'd rather make it more flexible so things like DN checks or similar could be added. Also this currently requires the certs to be deployed into the default locations, and doesn't allow for a crl yet. |
23:57 |
mgagne |
dgurtner: true. I'm however concerned that nova::migration::libvirt could soon grow a bit too large and requires its own puppet module |
23:58 |
Hunner |
mgagne: Thanks for your comment on rspec-puppet |
23:59 |
mgagne |
yw |
23:59 |
Hunner |
I updated it to pass the node_obj environment name through... how are you setting the environment? |