13:00:03 <morgan_orange> #startmeeting functest weekly meeting 1/09 13:00:03 <collabot> Meeting started Tue Sep 1 13:00:03 2015 UTC. The chair is morgan_orange. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:00:03 <collabot> Useful Commands: #action #agreed #help #info #idea #link #topic. 13:00:03 <collabot> The meeting name has been set to 'functest_weekly_meeting_1_09' 13:00:09 <morgan_orange> #info Morgan Richomme 13:00:22 <jose_lausuch> #info Jose Lausuch 13:02:04 <morgan_orange> and welcome back, I can feel the Spanish sun on the functest repo (and feel better) 13:02:39 <jose_lausuch> :p 13:02:40 <jose_lausuch> thanks 13:02:59 <jose_lausuch> but I missed rain a bit 13:03:16 <morgan_orange> You have to visit our OPNFV lab in Brittany... 13:03:31 <morgan_orange> so let's start, hope other people will join... 13:03:45 <morgan_orange> #topic actions follow-up 13:04:00 <morgan_orange> #link http://ircbot.wl.linuxfoundation.org/meetings/opnfv-testperf/2015/opnfv-testperf.2015-08-25-16.01.html 13:04:07 <morgan_orange> meimei1: are you around? 13:04:34 <morgan_orange> peter not connected 13:04:40 <morgan_orange> guy_rodrigue: are you there? 13:04:47 <jose_lausuch> well, at least I get an overview of whats happening 13:05:05 <guy_rodrigue> #info Guy Rodrigue Koffi 13:05:12 <morgan_orange> ahhhh 13:05:17 <guy_rodrigue> yes morgan_orange 13:05:27 <morgan_orange> in action points we had "plan a demo for week 26" 13:05:30 <jose_lausuch> Hi Guy :) 13:05:32 <morgan_orange> I think it is a bit early 13:05:34 <guy_rodrigue> welcome back jose_lausuch ! 13:05:40 <jose_lausuch> thanks! 13:05:50 <morgan_orange> but do you think we could propose a demo for thursday the 10th 13:06:09 <guy_rodrigue> yes 13:06:16 <jose_lausuch> next week 13:06:21 <morgan_orange> OK I take the point to propose this demo to teh agenda 13:06:32 <jose_lausuch> demo about_ 13:06:33 <jose_lausuch> ? 13:06:33 <morgan_orange> we will detail the status later 13:06:51 <jose_lausuch> ah ok, API demo 13:06:53 <jose_lausuch> I see 13:07:15 <morgan_orange> yep 13:07:18 <jose_lausuch> week 36, not 26 13:07:32 <morgan_orange> #action morgan_orange plan demo API September the 10th 13:07:42 <morgan_orange> that is all for the action points 13:07:55 <morgan_orange> #topic SR1 13:08:36 <morgan_orange> #info SR1 planned for end of September 13:08:38 <jose_lausuch> when is it, morgan_orange? 13:08:42 <jose_lausuch> ok 13:08:58 <morgan_orange> #info Functest must be OK (like for Arno stable) 13:09:10 <morgan_orange> #info wiki page created to follow SR1 status 13:09:46 <jose_lausuch> can you link it and info it? 13:09:46 <morgan_orange> #info LF POD have been reconfigured 13:09:59 <morgan_orange> foreman and fuel are installed alternatively on POD2 13:10:15 <jose_lausuch> ok 13:10:17 <morgan_orange> #link https://wiki.opnfv.org/functest_release_1 13:10:25 <jose_lausuch> I saw some emails about the credentials script 13:10:47 <morgan_orange> at the moment there is an open bug on the creds retrieval. It is OK on fuel but not OK with foreman 13:11:01 <jose_lausuch> ok 13:11:09 <jose_lausuch> I expected something like that 13:11:22 <morgan_orange> #info wait for tim green light (committing a patch for foreman) 13:11:23 <jose_lausuch> that script is not super robust 13:11:33 <jose_lausuch> I'll try to have a look 13:11:49 <morgan_orange> there was a path issue I corrected (we were not sourcing the file we built) 13:11:51 <morgan_orange> it is now OK 13:12:01 <morgan_orange> but there is an issue on the value we retrieve from foreman 13:12:06 <jose_lausuch> ok 13:12:13 <morgan_orange> #info bugs declared on foreman creds retrieval 13:12:28 <morgan_orange> #info there is also an issue with ODL suite, JIRA created 13:12:40 <morgan_orange> #info var issue (a priori not too complex) 13:12:57 <morgan_orange> #info vPing was (hopefully) still working with both installer 13:13:06 <morgan_orange> #info Rally issues 13:13:19 <morgan_orange> #info Tempest still OK but Rally bench scenraio cannot be started 13:13:28 <jose_lausuch> mmm 13:13:32 <morgan_orange> #info I opened a bug on Rally 13:13:33 <jose_lausuch> do you know why? 13:13:43 <jose_lausuch> its a rally problem or ours scripts? 13:13:56 <morgan_orange> it is a rally issue or a misconfiguration of rally from our script 13:14:08 <morgan_orange> I saw on the web someone who had the same issue 13:14:15 <jose_lausuch> ok 13:14:26 <jose_lausuch> I'll check that out 13:14:29 <morgan_orange> the workaround was not working on the fuel based solution (add an endpoint in config) 13:15:11 <morgan_orange> #link https://bugs.launchpad.net/rally/+bug/1489746/ 13:15:27 <morgan_orange> I planned to re-test the workaround on the foreman based installer... 13:15:41 <morgan_orange> so some work ahead for SR1 13:15:56 <morgan_orange> correction of fetch os + fix errors in ODL and Rally 13:16:19 <morgan_orange> any question for SR1? 13:16:43 <jose_lausuch> nop 13:16:52 <morgan_orange> ok so let's move to R2 topic... 13:16:59 <morgan_orange> #topic R2 dashboard 13:17:17 <morgan_orange> #info first meeting organized with LF, Ana on the test dashboard 13:17:41 <morgan_orange> #info guy_rodrigue and myself proposed a scenario 13:18:18 <jose_lausuch> are there minutes from that meeting? 13:18:19 <morgan_orange> #link https://wiki.opnfv.org/opnfv_test_dashboard 13:18:29 <morgan_orange> it was a gotomeeting 13:18:38 <jose_lausuch> ok 13:18:48 <morgan_orange> #info during this meeting they suggested other possible scenario plugin Jenkins 13:18:56 <morgan_orange> #info apparently could lead to perfo issues 13:19:13 <morgan_orange> #info so we continue with guy_rodrigue and shall be able to make a first demo next week 13:19:27 <jose_lausuch> what about bitergia? 13:19:29 <morgan_orange> #info basicaly we store results then generate json taht could be used by a portal and a js lib 13:19:39 <jose_lausuch> any news about that option? 13:20:02 <morgan_orange> #info bitergia was an option but LF would like to use thei own resource so more portal/drupal + js lib 13:20:17 <jose_lausuch> ahmm 13:20:19 <jose_lausuch> ok 13:20:19 <morgan_orange> #info a meeting is planned this week on thursday 13:20:28 <jose_lausuch> to discuss that 13:20:30 <jose_lausuch> fair enough 13:20:51 <morgan_orange> Andrew (who is leading that for LF) and Aric will eb there 13:21:06 <morgan_orange> I will send a mail on our scenario as we have almost a fully fonctional demo 13:21:48 <morgan_orange> it looks relatively light, no constraint on CI, almost no constraint on dashboardin - each test projects generate the graph they want to show 13:22:08 <jose_lausuch> well, that removes dependencies of graphs/results 13:22:22 <morgan_orange> #action morgan_orange mail to the community to be more precise that the wiki page (but basically it is just an real implementation of the scenario described in the page) 13:22:24 <jose_lausuch> among projects 13:22:59 <morgan_orange> ok for the dashboard? 13:23:18 <morgan_orange> #topic test profile 13:23:32 <jose_lausuch> yep 13:23:41 <morgan_orange> #info bryan initiated a doc on functest gerrit and a wiki page 13:24:08 <morgan_orange> #info the idea is to define reference traffic model for the different test projects in order, if possible to compare apples with apples 13:24:37 <morgan_orange> #link https://git.opnfv.org/cgit/functest/tree/commons 13:24:51 <morgan_orange> #info he accepted to lead this task 13:25:05 <morgan_orange> #info topic API test collection 13:25:27 <morgan_orange> #info LF DB setup and reachable from anywhere 13:25:51 <morgan_orange> #info for our demo we used our own DB because we must install a DB and the python framework to manage the API 13:25:56 <morgan_orange> guy_rodrigue: you can complete 13:26:08 <jose_lausuch> morgan_orange: what is that link? 13:26:10 <jose_lausuch> ims, mobile? 13:26:41 <morgan_orange> oups that is the place where the traffic profile will be stored 13:27:00 <morgan_orange> as well as the doc to list/explain them 13:27:02 <guy_rodrigue> yes we use motor and tornado as framework 13:27:13 <morgan_orange> #info use of motor + tornado for API 13:27:32 <morgan_orange> #info you can test it if you want... 13:28:09 <morgan_orange> #link curl -X GET -H "Content-Type: application/json" http://213.77.62.197:80/results 13:28:28 <morgan_orange> I will give details in the mail and during the demo 13:28:34 <jose_lausuch> ok 13:28:45 <jose_lausuch> looking forward :) 13:29:00 <morgan_orange> I think it was OK for this week for the topics 13:29:09 <morgan_orange> I initially planned some troubleshooting 13:29:16 <morgan_orange> but SR1 is the priority... 13:29:24 <morgan_orange> so we will troubleshoot once SR1 will be fine :) 13:29:41 <jose_lausuch> sounds good 13:29:50 <morgan_orange> #topic AoB 13:29:55 <jose_lausuch> hopefully not too much work 13:30:13 <morgan_orange> question, Debra planned the release weekly meeting every tuesday 13:30:23 <morgan_orange> at 6 PM CEST 13:30:36 <morgan_orange> it was one of the 2 time slot on which we were used to alternate... 13:30:56 <morgan_orange> shall we keep it (assuming that only 1 person is needed for Debra's meeting 13:31:39 <morgan_orange> if we keep onlt the 3PM we will not have lots of californian contributors...but we do not have a lot anyway... 13:31:53 <anac1> i have one question regarding setup tools 13:32:15 <morgan_orange> I think it is important to keep a dedicated synchro meeting 13:32:27 <jose_lausuch> shall we prioritize Debras? 13:32:34 <jose_lausuch> is it only for SR1 or for R2 also? 13:32:40 <morgan_orange> I think we must attend Debra's meeting 13:32:49 <morgan_orange> at least one of us 13:32:58 <morgan_orange> it is mainly for R2 13:33:13 <morgan_orange> the first meeting is today...so we will see exactly the content... 13:33:39 <morgan_orange> For the moment we can keep as we are and see if we can IRC in // of debra's meeting 13:33:52 <morgan_orange> and if it is too complex we will change our time slot 13:34:09 <morgan_orange> other information 13:34:22 <jose_lausuch> ok 13:34:31 <jose_lausuch> anac1: shoot :) 13:34:41 <morgan_orange> #info R2 page completed and lots of Jira created (not as many as yardstick...) but reasonable number 13:35:14 <anac1> do you have any dependencies to setup tools? 13:35:32 <anac1> yardstick need 17.1 or later, any objections? 13:35:57 <morgan_orange> no, for R2, one of the task is to put functest into a container to avoir dependencies 13:36:07 <jose_lausuch> yes 13:36:14 <jose_lausuch> I take that AP 13:36:15 <morgan_orange> we used to have to dependency issues and the different testbeds we deployed functest 13:36:25 <anac1> ok, good. just checking so we don't break anything for you... 13:36:27 <morgan_orange> jose_lausuch: good it is already assigned to you :) 13:37:11 <morgan_orange> but I think most of the project should probably adopt such container view other wise it will be probably lead to some issues 13:37:27 <anac1> agree 13:37:38 <morgan_orange> at least use virtual env 13:38:24 <morgan_orange> ok any question for this week? 13:39:03 <morgan_orange> I have also to see with the contributors - I know that Uli in Octopus will start a cleaning operation... 13:39:11 <morgan_orange> I imagine I could also do that 13:39:24 <morgan_orange> We will see next week if we are more.. :) 13:39:29 <jose_lausuch> no questions 13:39:43 <jose_lausuch> I might have some when I really start working :D 13:40:01 <morgan_orange> yep 13:40:05 <morgan_orange> #endmeeting