| *** markvoelker has joined #refstack | 00:19 | |
| *** tosky has quit IRC | 00:19 | |
| *** markvoelker has quit IRC | 03:20 | |
| *** markvoelker has joined #refstack | 04:21 | |
| *** markvoelker has quit IRC | 04:55 | |
| *** markvoelker has joined #refstack | 05:52 | |
| *** markvoelker has quit IRC | 06:25 | |
| *** kopecmartin|off is now known as kopecmartin | 07:19 | |
| *** markvoelker has joined #refstack | 07:22 | |
| openstackgerrit | Martin Kopec proposed openstack/python-tempestconf master: Enable manila plugin in devstack https://review.openstack.org/625191 | 07:44 |
|---|---|---|
| *** markvoelker has quit IRC | 07:55 | |
| *** tosky has joined #refstack | 08:26 | |
| *** markvoelker has joined #refstack | 08:52 | |
| *** markvoelker has quit IRC | 09:25 | |
| *** markvoelker has joined #refstack | 10:17 | |
| openstackgerrit | Martin Kopec proposed openstack/python-tempestconf master: Enable manila plugin in devstack https://review.openstack.org/625191 | 12:23 |
| *** markvoelker has quit IRC | 13:17 | |
| *** markvoelker has joined #refstack | 13:18 | |
| *** markvoelker has quit IRC | 13:22 | |
| kopecmartin | tosky, hi, can be task 22143 here https://storyboard.openstack.org/#!/story/2002574 considered done by this https://github.com/openstack/python-tempestconf/commit/7faab49110cd121d65f84c1bd3cf605a63d1049c | 13:36 |
| tosky | kopecmartin: it looks like it's the case | 13:40 |
| kopecmartin | tosky, cool, I'm cleaning the storyboard a little | 13:40 |
| tosky | oh, good :) | 13:41 |
| kopecmartin | tosky, just quick question, do you remember why we hardcoded this? https://github.com/openstack/python-tempestconf/blob/master/config_tempest/main.py#L98 | 14:09 |
| tosky | kopecmartin: I think that a value for that key is needed, but I can't be sure 100% | 14:11 |
| kopecmartin | tosky, i think so too, the problem is, that on devstack the role is not created during deployment and then if it's set in tempest.conf some tests are failing because they can't find the role | 14:14 |
| kopecmartin | and now I'm not sure how to fix that, 1. option is to check if the role exists and if not don't write it to tempest.conf | 14:14 |
| kopecmartin | 2. option is to check if exists and if don't then create it | 14:14 |
| tosky | use member | 14:15 |
| kopecmartin | hmm, ok, that could work , I'm gonna try | 14:15 |
| tosky | it's the standard at least from rocky, but then I'm not sure whether it works on devstack/queens and devstack/pike | 14:15 |
| *** markvoelker has joined #refstack | 14:18 | |
| kopecmartin | tosky, manila is working \o/ , can we merge this? https://review.openstack.org/#/c/625191/ | 14:52 |
| *** markvoelker has quit IRC | 14:53 | |
| kopecmartin | actually we don't have to rush, because it depends on another review | 14:53 |
| tosky | kopecmartin: oh, at least there was a real bug | 14:53 |
| kopecmartin | tosky, yeah | 14:53 |
| *** markvoelker has joined #refstack | 15:50 | |
| *** tosky has quit IRC | 15:53 | |
| *** tosky has joined #refstack | 15:54 | |
| kopecmartin | tosky, regarding the cinder review | 15:58 |
| tosky | kopecmartin: yes? | 15:58 |
| kopecmartin | tosky, is there a chance, that there will be older versions of the service deployed except the newest one? | 15:59 |
| kopecmartin | for example volume, volumev2 but no volumev3? | 15:59 |
| tosky | kopecmartin: uh, I really don't know; would that be a problem? | 16:00 |
| kopecmartin | i'm afraid yes :/ | 16:00 |
| tosky | volumev3 is pretty "old" at this point | 16:01 |
| tosky | which are the problematic combinations? | 16:01 |
| kopecmartin | tosky, this loop https://github.com/openstack/python-tempestconf/blob/master/config_tempest/services/services.py#L231 | 16:02 |
| kopecmartin | if there is v2, the service will be true , however if there is no v3, it will be set back to false | 16:02 |
| kopecmartin | oh, but it can be solved by changing the order | 16:02 |
| kopecmartin | never mind | 16:03 |
| kopecmartin | it's a dict | 16:03 |
| kopecmartin | hm | 16:03 |
| tosky | dict are without order, at least until python 3.6; unless you switch to OrderedDict | 16:03 |
| tosky | but | 16:03 |
| tosky | I think it's rare, but what if volumev3 is enabled but volumev2 is not? | 16:03 |
| kopecmartin | no idea , if one of them is enabled the whole service is, isn't it? | 16:04 |
| tosky | kopecmartin: uhm, I think that the patch is incorrect | 16:05 |
| tosky | don't we just need to override get_service_names? | 16:05 |
| kopecmartin | tosky, yeah :/ .. i don't know, i need to rethink it | 16:06 |
| tosky | kopecmartin: or better, get_service_names already return the correct information | 16:06 |
| tosky | that's how it should work | 16:06 |
| kopecmartin | the method was implemented for that purpose .. it seems we didn't finish its usage | 16:08 |
| tosky | yay | 16:08 |
| tosky | did I already mention that the migration to a pluggable configuration is not finished? :) | 16:08 |
| tosky | there should be a story for it iirc | 16:09 |
| kopecmartin | :D | 16:09 |
| kopecmartin | i don't see any story | 16:09 |
| tosky | 2002787 I think | 16:10 |
| tosky | but, uhm, it's finished | 16:11 |
| kopecmartin | can you pass me the link? i really can't see it | 16:13 |
| kopecmartin | :o | 16:13 |
| tosky | but why we still have that list of services in constants.py? I may have overlooked it | 16:13 |
| tosky | https://storyboard.openstack.org/#!/story/2002787 | 16:13 |
| tosky | we shouldn't need to use SERVICE_NAMES anymore - but that's unrelated with the current issue | 16:13 |
| kopecmartin | oh, i was looking for an active story | 16:14 |
| kopecmartin | tosky, i thought so too, but then i noticed, that the SERVICE_NAMES because of the translation to codename of the service | 16:15 |
| kopecmartin | *that we have | 16:15 |
| tosky | I think it was postponed because arxcruz had other duties | 16:16 |
| tosky | the entire code of set_service_availability should be rewritten, removing the service-specific cases | 16:17 |
| kopecmartin | yes , that sounds familiar :D | 16:18 |
| arxcruz | i don't remember | 16:19 |
| arxcruz | i think i left because of some compatibility | 16:20 |
| arxcruz | sorry guys, i'm not in my best today... | 16:20 |
| kopecmartin | arxcruz, don't worry, I hope you'll feel better tomorrow :) | 16:21 |
| *** markvoelker has quit IRC | 16:23 | |
| tosky | kopecmartin: would you mind adding another task to that story? | 16:26 |
| kopecmartin | tosky, sure, I'll add one | 16:27 |
| tosky | thanks | 16:28 |
| tosky | regardless of that, cinderv2/v3 should be discovered even now | 16:28 |
| kopecmartin | its' not, because "volumev2/v3" doesn't match with "volume" in the dict | 16:28 |
| tosky | ok, so the migration should be brought forward \o/ | 16:31 |
| kopecmartin | or let's call it refactoring :D that's even more hated word :D | 16:31 |
| kopecmartin | anyway, i need to run, see you tomorrow , bye | 16:32 |
| *** kopecmartin is now known as kopecmartin|off | 16:33 | |
| *** markvoelker has joined #refstack | 17:20 | |
| *** markvoelker has quit IRC | 17:53 | |
| *** markvoelker has joined #refstack | 18:50 | |
| *** markvoelker has quit IRC | 19:23 | |
| *** markvoelker has joined #refstack | 20:21 | |
| *** markvoelker has quit IRC | 20:55 | |
| *** markvoelker has joined #refstack | 21:53 | |
| *** markvoelker has quit IRC | 22:25 | |
| *** tosky has quit IRC | 23:11 | |
| *** markvoelker has joined #refstack | 23:22 | |
| *** markvoelker has quit IRC | 23:55 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!