14:11:45 #startmeeting Rally 14:11:45 Meeting started Mon Aug 31 14:11:45 2015 UTC and is due to finish in 60 minutes. The chair is boris-42. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:11:46 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:11:48 The meeting name has been set to 'rally' 14:12:02 hi 14:12:06 e0ne: hi 14:12:13 hi 14:12:23 ikhudoshyn: ping 14:12:26 hi all 14:12:30 andreykurilin: ping 14:12:32 0/ 14:12:33 oanufriev: hi 14:13:27 hi 14:13:31 yfried: hi hi 14:14:17 hi all 14:15:38 albertw1: hi 14:15:44 so let's start 14:15:51 #topic Meeting Agenda 14:16:38 o/ 14:16:43 #link https://wiki.openstack.org/wiki/Meetings/Rally#Agenda 14:16:47 janonymous: hi 14:16:52 hi 14:17:02 #topic [amaretskiy] Status for patches related to HTML reports and new charts classes 14:17:38 amaretskiy1: ^ 14:17:46 status for reports-related patches is quite short: all these 4 patches looks working and ready for review, so please review them. eom 14:18:35 amaretskiy1: great status but i believe that many of those who is sitting here don't know what are you patches 14:18:39 amaretskiy1: and what they do 14:19:12 there are significant changes in HTML report processing 14:19:56 amaretskiy1: the new html look nice, the code is really complicated (for me) to review 14:20:05 first patch https://review.openstack.org/#/c/214695/ adds a method that re-formats task results data so they have some aggregated info 14:20:22 next patch https://review.openstack.org/#/c/146814/ is the most important 14:21:00 it uses method from first patch and also uses already merged new classes for report generation 14:21:40 so now we can process tasks data in single cycle (streaming processing) 14:21:59 and we have OOP-code for that and get rid of mess in plot.py 14:22:22 and now new charts can be easily added, based on base class Chart 14:22:29 there ar ea lot of benefits 14:23:02 another 2 patches simply give improvemnents 14:23:08 like LoadProfile chart 14:23:32 and showing task UUID of task results file name in report HTML 14:23:41 eom 14:23:52 amaretskiy1: ^ thanks for more detailed update 14:23:58 and showing task UUID `OR' task results file name in report HTML 14:24:01 does anybody have any questions? 14:24:54 looks like nobody has 14:25:00 okay let's move on 14:25:12 #topic [ikhudoshyn] Object-oriented design of task configuration 14:25:31 https://review.openstack.org/#/c/211160/ here is the patch 14:25:39 it represents ongoing work 14:25:54 on moving to a new task format 14:26:19 the patch is not ready -- lots of tests age red 14:26:30 but there are two points 14:27:10 1) I'd like you guys and especially you, boris-42, to look at it and tell me if it is ok functionality-wise 14:27:24 2) There was a comment from P.Boldin.. 14:28:18 as for extracting a common part of "scenario" scheme and using "allOf" json scheme mechanism to extend it properly 14:28:43 unfortunately, "allOf" does not work with "additionalProperties": false 14:28:51 ikhudoshyn: yep I know that 14:29:02 ikhudoshyn: I faced the same issue before with jsonschema 14:29:19 ikhudoshyn: so I belive that allOf is cosmetic change 14:29:22 so two options arise: 1.1) update python dict explicitly -- does not look great for me 14:29:29 ikhudoshyn: and addtitionalProperties is very important 14:29:31 1.2) leave it 14:29:42 boris-42: I agree 14:30:25 ikhudoshyn: so what we need is working code with some small note that points bug in jsonshema (or explains it) 14:30:38 if there's no other objections, I'd like once again to ask you to review it as for functionality provided 14:30:45 meanwhile I'll do tests 14:30:46 eom 14:32:05 ikhudoshyn: I was just looking at it there are some naming issues 14:32:51 I see. But we already have 'task' variable across the code and it references to DB model. So I'd like to separate those 14:33:12 what do you think? 14:33:29 ikhudoshyn: it's better to start doing right* 14:33:44 ok, will do 14:34:47 okay 14:34:52 any questions guys? 14:35:50 seems like no questions 14:36:29 okay let's move to the next topic 14:36:53 #topic [boris-42] Rally Plugin reference 14:36:59 Okay guys it's interesting stuff 14:37:20 I spend friday to get familiar 14:37:29 with docutils 14:37:42 and find the way to generate single page that contains all information aobut all plugins 14:37:53 yfried: ^ i remember you was looking for this 14:39:59 boris-42: I wass 14:40:12 yfried: so here it will be https://review.openstack.org/#/c/218874/ 14:40:15 #link https://review.openstack.org/#/c/218874/ 14:41:31 so any questions? 14:41:43 amaretskiy1: ^ andreykurilin ^ 14:41:51 oanufriev: ^ 14:42:42 boris-42: no questions for now, will take a look at this patch 14:43:11 #topic open discussion 14:43:18 do we have soemthign to discuss? 14:43:21 boris-42: can this be applied to "rally info" command as well? 14:43:33 yfried: so I am going to remove rally info command 14:43:38 we have an issue reported regarding installing rally as murano app 14:43:43 yfried: there will be rally plugin find (command) 14:43:52 yfried: actually show/list 14:44:02 yfried: that will show all plugins in the same way 14:44:05 boris-42: and it will replace info and encompass all plugins? 14:44:11 boris-42: greate 14:44:11 yfried: it already does 14:44:24 yfried: here it is https://review.openstack.org/#/c/217005/3 14:44:32 yfried: I need just to finish part of unittests 14:44:37 yfried: and it will be ready 14:44:57 boris-42: I assume they use the same code for pulling out of docstrings... 14:45:18 boris-42, maybe we can remove rally info without deprecation? 14:45:46 yfried: yep 14:45:56 yfried: and the fact that now all plugins are working in the same way 14:46:01 boris-42: great. I'll review 14:46:16 yfried: so we can have 1 mechanism for all of them 14:46:29 e.g. info, validation, types, .... 14:47:57 amaretskiy1: did you have a cahnce to take a look at bug 14:48:07 amaretskiy1: with murano? 14:48:25 boris-42: only short look 14:49:17 boris-42: https://bugs.launchpad.net/murano-apps/+bug/1490586 14:49:18 Launchpad bug 1490586 in murano-apps "unable to deploy Rally application" [Undecided,New] 14:50:16 amaretskiy1: seems the problem is in install_rally.sh 14:50:28 amaretskiy1: we need to do system wide installation with sudo 14:50:28 boris-42: fatal: Not a git repository (or any of the parent directories): .git 14:50:42 amaretskiy1: nope take a look later message 14:50:47 boris-42: yes, murano app should work from the box 14:50:53 install_rally.sh: ERROR: missing software prerequisites 14:50:57 amaretskiy1: ^ 14:51:11 amaretskiy1: this means that someobdy run install_rally as a not system wide 14:51:24 amaretskiy1: so could you take a look at murano application installations? 14:51:38 boris-42: yes, I will do that 14:52:19 boris-42: I've assigned this bug on me 14:52:35 amaretskiy1: https://github.com/openstack/murano-apps/blob/master/Rally/package/Resources/scripts/installRally.sh#L15-L19 14:52:45 amaretskiy1: I found the issue you se there is no sudo here 14:52:54 okay so we can actually finish meeting 14:52:58 and go to rally chat 14:53:00 #endmeeting