14:01:04 #startmeeting sahara 14:01:12 Meeting started Thu Sep 20 14:01:04 2018 UTC and is due to finish in 60 minutes. The chair is tellesnobrega. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:13 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:01:16 The meeting name has been set to 'sahara' 14:01:23 this time I'm not late \o/ 14:01:26 o/ 14:01:29 o/ 14:01:45 * jeremyfreudberg needs to leave at 40 minutes past the hour 14:01:59 no problem, luigi is 30 past the hour 14:02:02 so lets rush 14:02:07 #topic News/Updates 14:02:16 technically not anymore (last minute change), but I'm fine closing earlier anyway :) 14:02:30 oh 14:02:31 ok 14:02:59 Not much changed since last week, I'm working on split plugins but now I'm down to 0 failing tests on the sahara-plugins side 14:03:50 I have a question for you both on the subject, but I can save it for later 14:04:06 PTG was productive; in order to help with the plugin split process, I'm refactoring the documentation 14:04:07 nothing new on my end 14:04:13 one small patch is already out 14:04:58 #link https://review.openstack.org/#/c/602458/ 14:05:03 this one ^ 14:05:04 the big one is almost done, it's touching ~16 files 14:05:05 right? 14:05:34 yep 14:06:14 ok, lets move on 14:06:20 #topic Plugins Split 14:06:30 Let me jump to my question 14:07:19 if you look here https://github.com/openstack/sahara/blob/master/sahara/plugins/cdh/plugin_utils.py#L240 14:07:27 you see a reference to res.Resource 14:07:37 that comes from sahara/conductor/resource 14:08:33 so, I will need refactor this bit so we don't import from sahara.conductor directly 14:08:45 buut, this checks if isinstance 14:09:10 so I actually need to create a class that resource will inherit from so the check works 14:09:20 does that make sense to everyone? 14:09:59 or expose a function which does the check 14:10:03 I personally don't really like this, because we will have a top class under sahara/plugins and sahara/conductor/resource will inherit and they are not really related, but I'm not seeing a better way to fix this 14:10:41 you can expose a function, like tosky says, or you can use hasattr 14:11:01 tosky, you mean add a isinstance on sahara/plugins that does this check instead of doing it on plugins side itself? 14:11:36 not called isinstance, but is_resource, or something like that, BUT you need to return a res.Resource too, so it won't work 14:12:17 unless I do that under sahara/plugins/something and call from the sahara_plugins side 14:12:27 but you are right 14:12:33 it would work for the check 14:12:35 but not the return 14:12:57 I wonder if it really returns to return a res.Resource 14:13:04 the only place where it's used seem to use a dict 14:13:35 in sahara/plugins/cdh/cloudera_utils.py 14:14:09 yeah, and i think it eventually ends up here, no? https://github.com/openstack/sahara/blob/4b4effa9c8062c26fafe2b9c20146bf67b44b2b5/sahara/utils/configs.py#L17 14:14:32 that's what it looks like 14:15:31 so you are saying we can return something else? instead of a resource? 14:15:42 maybe just a dict 14:16:07 yes, i believe just a dict should work 14:16:47 the output of convert_process_configs is only used in that one place in cloudera_utils.py, and then it's immediately passed to merge_configs(), which accepts anything dict-like 14:17:00 ok, I will give it a try 14:17:32 I *think* that is the only remaining direct link 14:17:33 it looks like the fact that Resource is a special dictionary (immutable object) is not much used 14:17:53 but in case please add more unit tests around the code which is touched, before and after the split 14:17:57 I will double check and if need I will come up with more questions 14:17:58 so that should say something 14:18:26 tosky, sure, I will give it a try 14:19:38 thanks you both 14:19:45 any specific topics from you? 14:20:26 there are also two general patches about the build system: one fix grenade, the other remove the cover job in post (never used) 14:21:34 respectively https://review.openstack.org/#/c/602506/ and https://review.openstack.org/#/c/600686/ (the latter depends on the former) 14:21:54 lets go into open discussion so we can cover all remaining topics 14:21:58 #topic Open Discussion 14:22:15 tosky, both are ready to merge? 14:22:31 the second more when the first lands for sure 14:23:03 https://review.openstack.org/#/c/600686/ can't land without https://review.openstack.org/#/c/602506/, there is a dependency 14:23:03 thanks jeremyfreudberg for the quick +2+A 14:23:56 sure 14:25:00 it all looked fine to me 14:25:38 cool 14:25:40 anything else 14:26:45 nothing from me 14:26:49 nothing else so far; expect the big doc patch $soon :) 14:27:31 tosky, good to hear 14:27:54 I guess we can close early and have 33 minutes back to work 14:27:57 thanks all 14:28:02 thanks 14:28:14 #endmeeting