14:01:08 #startmeeting Rally 14:01:09 Meeting started Mon May 25 14:01:08 2015 UTC and is due to finish in 60 minutes. The chair is amaretskiy. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:10 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:01:13 The meeting name has been set to 'rally' 14:01:19 hi all 14:02:43 hi 14:02:53 hi all 14:03:08 let's start 14:03:29 #topic new HTML reports classes 14:04:08 I'm currently working on new improved classes for report charts 14:04:24 #link https://review.openstack.org/#/c/159458/ 14:04:39 #link https://review.openstack.org/#/c/146814/ 14:04:50 #link https://review.openstack.org/#/c/169828/ 14:04:59 these patches will be updated soon 14:05:12 so I believe tomorrow they will be ready for review 14:05:25 and we will have these patches in next release 14:05:51 #topic New script for gate jobs 14:06:46 redixin what status have https://review.openstack.org/#/c/175549/ 14:06:59 I believe it is complete 14:07:06 it have only on +1 from you ;) 14:07:20 yep 14:07:24 i think so 14:07:24 this patch needs reviews 14:07:43 colleagues, please review it 14:08:01 this patch is critical for next release 14:08:32 okay, waiting for reviews from others.... 14:08:44 #topic Fuel benchmarks 14:10:04 this topic is related to several patches, first patch is one that registers fuelclient and adds simple scenario - https://review.openstack.org/#/c/178731/ 14:10:23 there is a small nit with this patch - I will submit new patch set today 14:10:39 but we have real problem on gates 14:11:21 since fuelclient package is in optional-requirements.txt and we have ImportError on gates 14:11:21 so are we going to skip testing scenario samples? 14:11:38 this is a point for discussion 14:12:10 I like this way of solution 14:12:10 for my point of view, skip is a simple and good solution 14:12:52 but I do not remember that boris-42 has approved that :( 14:12:55 if can't solve it in normal way 14:13:14 redixin, what do you think, can we start implementing of this skip without boris-42 opinion, or we are blocked? 14:14:09 amaretskiy: redixin will the Fuel scenarios be tested on the Mirantis Gate? 14:14:23 yfried_, it already tested 14:14:41 fuel-community-6.0 SUCCESS (non voting) in 4m 12s 14:14:54 yfried_ the problem is regarding job "gate-rally-dsvm-cli" 14:15:11 amaretskiy, imo we should add patch with "try import except ImportError self.skip" 14:15:18 redixin: then make it voting and IMO it should be enough. if the main gait doesn't allow for it 14:15:59 yfried_, the problem is in another job. that job is trying to do "rally task validate" for each task sample 14:16:12 yfried_, and it is failing with ImportError fuelclient 14:16:30 okay, does anybody want to make this patch? 14:16:31 redixin: I get it. You want to skip (=igonre importError) on the main gate. 14:16:32 so we can just skip it if there is no fuelclient installed 14:17:11 redixin: IMO this is only valid of the Mirantis gate has it as voting. 14:17:14 redixin, we can even parse if `fuelclient' is in ImportError messahe 14:17:25 *message 14:17:29 amaretskiy, wait 14:17:38 redixin: amaretskiy: I don't understand why have the fuel scenario on main gate at all 14:17:40 Ok I will make patch that skip ImportError for Fuel samples 14:17:52 rvasilets: great 14:17:53 try import fuelclient except importError // should be enough 14:18:22 redixin: okay 14:18:39 yfried_, it is not fuel scenario. it is job which checks all task samples 14:19:09 so, it seems that we have decided what to do with this problem, lets continue... 14:19:39 hi 14:19:45 we have another patches related to Fuel scenarios 14:19:49 e0ne hi 14:19:52 yfried_, https://github.com/openstack/rally/blob/master/tests/functional/test_task_samples.py 14:22:21 amaretskiy, oh. we cant just add "import fuelclient" in try/except block. we should do something other. like here: https://github.com/openstack/rally/blob/master/tests/functional/test_task_samples.py#L49-L50 14:23:21 I'll solve this issue 14:23:27 good 14:23:33 rvasilets, great! 14:23:38 ok, let's proceed 14:23:51 oanufriev please give your status for your patches regarding Fuel scenarios 14:25:18 Well, there are two patches: https://review.openstack.org/182873 and https://review.openstack.org/181270 14:25:18 They are under development - no unit tests. 14:25:18 are these patches scheduled for next release? 14:25:42 But looks like they do the job 14:25:59 EOM 14:26:14 According to current release management doc - I guess no 14:26:53 ok. I will take a look anyway 14:26:56 Okay, these patches are important in any case, waiting for unit tests 14:27:05 ok 14:27:12 #topic Specs 14:27:57 we have a lot of important patches related to specs 14:28:42 yair_ please tell us about refactoring of scenario utils 14:28:53 I see interesting spec about that 14:29:30 yfried_* 14:29:36 yfried_ ^ 14:29:54 amaretskiy: the idea is to make the scenaio/*/utils.py not part of inheritance tree 14:30:41 we suffer for this problem for a long time, this spec is important 14:30:53 amaretskiy: it's blocked on 2 things 14:31:07 1. AtomicMixin 14:31:42 2. https://review.openstack.org/#/q/status:open+project:openstack/rally+branch:master+topic:refactor_a_bit_plugins,n,z 14:31:46 yes, atomic actions are not well-described in this spec (that I have mentioned in review) 14:32:41 amaretskiy: 1 - should remove atomic actions from the scenario tree and make it plugins generic. I guess I'm gonna have to re-visit https://review.openstack.org/#/c/148637/ 14:33:27 amaretskiy: problem is, having 2 merged would make 1 much simpler if we make it a part of the generic plugin base 14:33:46 in any case, this spec should explain how atomic actions will be implemented 14:34:23 amaretskiy: I'm waiting on boris to help me draft the spec and finish #2 14:34:38 I had discussion with boris-42 about atomic actions - their correct implementation is very important 14:34:51 okay, redixin oanufriev rvasilets please look at this spec 14:35:11 okay, let's proceed 14:35:13 amaretskiy: I'm a little foggy on the details 14:35:32 amaretskiy: so any suggestions re atomic actions are welcome 14:35:42 amaretskiy: maybe a different spec??? 14:36:17 yfried_ mayvbe just add some idea (even foggy) to the spec so we will have a point to start discussion? 14:36:25 amaretskiy: will do 14:36:31 yfried_ thanks! 14:36:45 #topic Free discussion 14:37:00 does anyone have comments? 14:38:36 nop 14:38:48 okay, so end meeting 14:38:53 thanks to all!!! 14:38:58 #endmeeting