08:00:16 #startmeeting Functest weekly meeting September 6th 08:00:16 Meeting started Tue Sep 6 08:00:16 2016 UTC. The chair is morgan_orange. Information about MeetBot at http://wiki.debian.org/MeetBot. 08:00:16 Useful Commands: #action #agreed #help #info #idea #link #topic. 08:00:16 The meeting name has been set to 'functest_weekly_meeting_september_6th' 08:00:22 #topic call role 08:00:28 #info Morgan Richomme 08:00:28 SerenaFeng: it worked here https://build.opnfv.org/ci/job/functest-fuel-baremetal-daily-master/469/ 08:00:34 #info Jose Lausuch 08:00:34 #info SerenaFeng 08:00:35 #info Juha Kosonen 08:00:38 #info Viktor Tikkanen 08:00:40 #info Cristina Pauna (Enea) 08:00:57 #info agenda https://wiki.opnfv.org/display/functest/Functest+Meeting 08:01:04 #info raghavendrachari 08:01:13 We noticed it since last week, she's working on it now 08:01:25 SerenaFeng: OK we will indiate it in teh section on scenario 08:01:33 #topic Action point follow up 08:01:41 #link http://ircbot.wl.linuxfoundation.org/meetings/opnfv-functest/2016/opnfv-functest.2016-08-30-08.00.html 08:01:50 #info AP1: morgan_orange SerenaFeng sync with Juraj for dashboard activity 08:02:00 #info serenaFeng committed first patches last week, discussion to get bitergia contacts 08:02:06 #info AP2: morgan_richomme re-clean based on David's list 08:02:14 #info done, last list was clearer, do not forget to indicate teh target version for correction 08:02:21 #info AP3: morgan_richomme create JIRAs if needed to reflect bug fix activity 08:02:27 #info done I do not see any missing JIRAs regarding our activities 08:02:36 if you see anything missing, do not be shy and create a JIRA... 08:02:42 #info AP4; morgan_orange check reporting links http://testresults.opnfv.org/reporting/functest/release/colorado/index-status-compass.html 08:02:48 #info done during last meeting 08:02:53 #info AP5: SerenaFeng see with parser team if parser must be run on HA mode only 08:02:59 #info done: noha scenario disabled for parser (virtual resources - often linked to non ha scenario) lacking ressources for parser case 08:03:11 #info shall we plan in the future also a constarint on the type of pod (as HA can be done in virtual.. the match ha/noha bm/virtual is not systematic)? 08:03:24 #info AP6: morgan_orange adapt lxd scenario declaration + release note accordingly 08:03:43 #info done healthcheck excluded, Rally black list mechanism implemented by juhak 08:03:51 merged on master today 08:03:56 wait for execution on master 08:04:25 then see with joid for the tests to be excluded (but no joid CI at the moment, see next section) 08:04:31 #link https://gerrit.opnfv.org/gerrit/#/c/20361/ 08:04:38 #info AP7: viktor_t see if for odl_l3 we should exclude test cases 08:05:01 #info viktor_t mechanism to exclude always failing tests implemented (forgive me to have changed my mind...final decision was to not exclude the test case as it corresponds to a real bugs that prevent Ip allocation...so not very useful scenario) 08:05:17 #info release note updated to reference odl_l3 bug 08:05:23 #info AP8: jose_lausuch fix issue of artifacts not pushed in cased of exit -1 (any case failed) 08:05:29 #info done we have now artifacts whatever the run results. Nice (frightning) error reports from security_scan. 08:05:35 #link http://artifacts.opnfv.org/logs/functest/lf-pod1/2016-07-16_05-12-54/master/security_scan/192.30.9.5_2016-07-16_05-12-43/report.html 08:05:40 #info AP9: viktor_t once artifact issue fixed, study odl_l3 tempest errors 08:05:45 #info see AP7 08:05:52 #info AP10: morgan_orange create 2 topics related to new VNF integration 08:05:59 #info mail sent with several proposals + Serena proposal on htmlization / pdfization, see next scetion for vote 08:06:04 #info AP11: jose_lausuch morgan_orange check Colorado docker consistancy and fix branch for OPNFV internal repos in Colorado(stable) docker 08:06:11 #info not done, shall we fix all feature project version to Colorado in Colorado branch only now? 08:06:14 this is not done! 08:06:20 we are always running master branch I think 08:06:23 this is crucial 08:06:24 important 08:06:28 we should run colorado branch code 08:06:31 we run stable... 08:06:35 on colorado 08:06:41 and master on master 08:06:44 do we have a docker for that? 08:06:59 that is what I can read in the trace 08:07:04 hmmmm 08:07:05 yes 08:07:14 but the dockerfile is cloning the master branch, right? 08:07:21 we dont have different dockerfiles for each branch 08:07:39 so opnfv/functest:colorado.0.X clones also master branch 08:08:01 ok so we should have at least 2 docker stable => clone colorado, latest => clone master 08:08:20 the problem here is: 08:08:25 so far, as we cherry pick almost systematically master = colorado... 08:08:34 if we do that, we need to maintain 2 files which will differ depending on the branch 08:09:00 I would suggest: we clone clorado branch in the dockerfile and push that patch to master branch+cherry pick to colorado 08:09:17 and after release date we change the cloning to master and push it only to master branch 08:09:44 ok it means that until the release master and colorado will clone colorado 08:09:58 so pre test on master to see if there is no impact will not be useful 08:10:20 well, you are right... 08:10:20 mmmm 08:10:26 then we need to maintain 2 dockerfiles 08:10:32 and push the changes on colorado branch only 08:10:38 or? 08:10:40 Could we specify build arg when building the docker images (--build-arg)? 08:11:06 ollivier: good point, I hope, that would be another way 08:11:14 ollivier: need to investigate how to do that with docker 08:11:24 ap to me :) 08:11:31 #link https://docs.docker.com/engine/reference/commandline/build/ 08:11:35 #action jose_lausuch ollivier find best way to coloradoize our docker and try to keep a master to be able to test changes without breaking colorado gate 08:11:47 #info ollivier 08:11:53 ok any comments/remarks on the action points? 08:12:13 #topic Colorado status 08:12:21 #info scenarios apex 08:12:27 #link http://testresults.opnfv.org/reporting/functest/release/colorado/index-status-apex.html 08:12:33 #info first blue balls in Jenkins! i.e. security_scan and copper OK (in HA or with patch as indicated by Tim? to be confirmed) 08:12:59 #link https://build.opnfv.org/ci/view/functest/job/functest-apex-apex-daily-colorado-daily-colorado/94/ 08:13:10 #info issues with Tempest/odl_l3 due to ODL bug (will nto be fixed in C), still issues with bgpvn and sfc scenario 08:14:02 ap me for bgpvpn 08:14:17 #action jose_lausuch sync for bgpvpn 08:14:35 #undo 08:14:35 Removing item from minutes: 08:14:47 #action jose_lausuch sync for apex/bgpvpn 08:14:51 #action morgan_orange sync for apex/sfc 08:15:01 you can action me also for sfc 08:15:06 ok... 08:15:07 I know the guy who is fixing it 08:15:11 #undo 08:15:11 Removing item from minutes: 08:15:18 I'm talking to him right now :) 08:15:22 #action jose_lausuch sync for apex/sfc 08:15:31 but for onos-sfc... 08:15:40 https://build.opnfv.org/ci/view/functest/job/functest-apex-apex-daily-colorado-daily-colorado/99/console 08:15:47 the heathcheck does not pass 08:16:19 and for onos-scenario without sfc, the vping_userdata does not pass either https://build.opnfv.org/ci/view/functest/job/functest-apex-apex-daily-colorado-daily-colorado/97/console 08:16:20 OK I do not see in the reporting page because it fails early 08:16:31 May-meimei: ? 08:17:08 #action May-meimei sync for apex/onos and apex/onos-sfc 08:17:49 it seems that Tempest and rally are also not fully OK in all the scenarios 08:18:31 I will action viktor_t and juhak for all the installers at the end :) 08:18:41 #info scenario compass 08:18:49 #link http://testresults.opnfv.org/reporting/functest/release/colorado/index-status-compass.html 08:18:56 #info compass/moon issue on ODL identified and documented, work on moon side for odl and moon tests 08:19:01 #undo 08:19:01 Removing item from minutes: 08:19:07 jose_lausuch: thanks for the good news ;-) 08:19:18 #info compass/moon issue on ODL identified and documented, work on moon side for tempest and moon tests 08:19:22 jose_lausuch: which patch failed to merge? can I help you? 08:19:33 #info compass/odl_l3 idem apex 08:19:35 cgoncalves: yes, let's take it later :) 08:19:39 #info compass onos/sfc 08:19:45 #link http://testresults.opnfv.org/test/api/v1/results?case=onos_sfc&last=10&installer=compass&scenario=os-onos-sfc-ha 08:19:50 sure. sorry, didnt notice you were having a meeting. 08:19:52 #info note same test ok on other scenarios 08:19:58 #link http://testresults.opnfv.org/test/api/v1/results?case=onos_sfc&last=10 08:20:23 who is in touch with onos-sfc people? 08:20:47 good question 08:20:51 we know them at least 08:21:03 I would say May-meimei 08:21:05 I can action May-meimei as she committed a patch around recently 08:21:11 yes 08:21:14 May-meimei: tell us if any problem 08:21:25 #action May-meimei sync compass/onos-sfc 08:21:32 #info fuel scenarios 08:21:38 #link http://testresults.opnfv.org/reporting/functest/release/colorado/index-status-fuel.html 08:21:44 #info no more blocking case (except Tempest odl_l3 like for all the installers) 08:21:58 #info all the scenarios still not fully OK due to some errors in doctor/parser/promise/Rally/bgpvpn but case worked once at least 08:22:05 #info last 10 parser tests are OK 08:22:50 #info warning on fuel/doctor as last 3 run seems FAIl 08:23:00 #link http://testresults.opnfv.org/test/api/v1/results?case=doctor-notification&last=10&installer=fuel 08:23:06 cgoncalves: yes, I noticed doctor fails sometimes 08:23:27 you can AP fuel doctor to me 08:23:39 problem seems to be related to fuel configs and not doctor itself 08:23:45 #action SerenaFeng sync fuel/doctor 08:23:54 I know the people who integrate doctor with fuel 08:24:05 #info joid scenarios 08:24:15 #info no run due to work on INTEL PODs...(no CI resources 15 days before release...) 08:24:30 #link http://testresults.opnfv.org/reporting/functest/release/colorado/index-status-joid.html 08:24:49 #action viktor_t review scenarios where tempest is not fully sunny 08:25:03 #action juhak review scenarios where rally is not fully sunny 08:25:27 anything you want to add about the scenario 08:25:42 does joid need to add black rally tests for lxd scenarios? 08:25:51 jose_lausuch: yes 08:26:00 Juhak just merged the patch for that 08:26:05 would be nice to test that 08:26:15 the mechanism has been merged on master 08:26:24 need to cherry pick and select the case to be excluded 08:26:37 but joid is lacking CI ressource at the moment for testing 08:26:48 I can probably test it on Orange POD1 (lxd scenario deployed here) 08:27:01 #action morgan_orange test rally black list on lxd scenario 08:27:44 so for the release meeting today, I would say that I do not see critical points (except CI resources for joid) , integration is making progress on all the scenarios 08:27:55 agree 08:28:01 we have some point of vigilences with bgpvpn, sfc, and fuel/doctor 08:28:11 but compared to previous release... 08:28:15 that is critical and there are people working on that 08:28:22 just submited a patch for bgpvpn 08:28:32 a question on odl-sfc 08:28:38 there is no status pushed in DB 08:28:52 morgan_orange: right, I can take care of that, its just adding the function at the end 08:28:53 shall I add one? it allows an automatic integration in the reporting 08:28:57 action me 08:29:00 I know 08:29:04 it would be handy 08:29:05 you have already lots of things to do 08:29:08 I will action me 08:29:13 ok, thanks :) 08:29:22 #action morgan_orange add push to DB in odl_sfc 08:29:25 but you are now part of the TSC :) 08:29:42 my salary has been immediately multiplied.... 08:29:43 by 1 08:29:58 better than x0 :D 08:30:34 any other comment on the scenarios/Colorado? 08:30:53 I created a patch for doc cross review with other test projects (agreed during weekly meeting) 08:31:05 looking for volunteer to review yardstick documentation... 08:31:46 any volunteer? 08:31:58 mmmm 08:32:07 I could volunteer but you won't allow me :p 08:32:24 no 08:32:45 it should be relatively short, Raghav would you be ok? 08:33:02 okey, you two are too busy, ap me 08:33:31 Ok but you are also busy with the dashoard activities...and reflexion on D... 08:33:37 ok 08:33:40 yeah .. 08:33:56 #action SerenaFeng Raghav review Yardstick documentation 08:34:17 not connected CristinaPauna I mentioned ARM in the Functest different documents, it would good to have a ENEA review on that 08:34:50 #topic Dashboard evolution 08:34:50 morgan_orange: sure, add me or Alex Avadanii as reviewer 08:35:32 #action morgan_orange add CristinaPauna and Alex Avadanii in Functest doc review 08:35:38 for the dashboard short status 08:35:45 we agrreed to move to ELK 08:36:05 SerenaFeng: provided new patch to clean the result export form mongo to elasticsearch 08:36:24 I'd like to divide the dashboard evolution to three part 08:36:52 feel free to do it.... 08:36:54 first refactor the fixed template such as _testcases 08:37:18 and do some structure adjustment 08:37:37 at last, because I am not a website designer 08:37:45 I contacted LF to get also a contact in bitergia to see if they could do something with our data but probably nothing for C so we need to do somethign by our own 08:38:05 hehe none of us is a website designer... 08:38:22 when it comes to website design, I will publish an website, so we all can have a look 08:38:24 but last time the webdesigner we worked with ...ok I will not say anything... 08:38:52 if something is unreasonable or not very good, you can make your suggestion, then I can fix it 08:38:55 if I have some time I planned to train myself a little bit on ELK... 08:38:58 do you think it is a good way? 08:39:11 there is no other way... :) 08:39:53 ok, I will follow the pace 08:39:59 ok 08:40:22 anyone interested to help SerenaFeng on the dashboard? post processing of the results to draw results ) f(time) 08:40:27 JuhaHaapa: Raghav ? 08:40:35 yes, i'll 08:41:05 ok thanks SerenaFeng maybe we should indicate our view in a wiki page 08:41:22 yeah, I think so 08:41:26 #topic internship program 08:41:32 I sent a mail last week 08:41:45 #info proposals are: 08:41:46 #info 1: open Source VNF catalog 08:41:51 #info 2: Functest unit tests 08:41:57 #info 3: vCDN VNF bootstrap 08:42:02 #info 4: vPing evolution 08:42:09 #info 5: VNF integration based on Juju orchestrator 08:42:14 #info 6: add token to testapi and pdf-ize or html-ize testapi 08:42:20 #info 7: Automatic reporting evolution 08:42:29 you want to comment/add another topic? 08:43:08 #startvote vote for your 2 favorite subjects? 1, 2, 3, 4, 5, 6, 7 08:43:08 Begin voting on: vote for your 2 favorite subjects? Valid vote options are 1, 2, 3, 4, 5, 6, 7. 08:43:08 Vote using '#vote OPTION'. Only your last vote counts. 08:43:28 #vote 3 08:43:42 #vote 6 08:43:45 #vote 1 08:44:02 #vote 6 08:44:04 #vote 6 08:44:17 #vote 6 08:44:18 apparently only the last vote count 08:44:18 #vote 2 08:44:26 so we need to make two rounds... 08:44:26 #vote 2 08:44:39 #endvote 08:44:39 Voted on "vote for your 2 favorite subjects?" Results are 08:44:39 1 (1): morgan_orange 08:44:39 2 (2): jose_lausuch, SerenaFeng 08:44:39 6 (1): Raghav 08:44:46 #vote 6 08:44:57 yes it is only the last so let's redo it 08:45:07 hehe 08:45:08 #vote 3 08:45:18 #startvote vote for your favorite topic? 1, 2, 3, 4, 5, 6, 7 08:45:18 Begin voting on: vote for your favorite topic? Valid vote options are 1, 2, 3, 4, 5, 6, 7. 08:45:18 Vote using '#vote OPTION'. Only your last vote counts. 08:45:21 so your most favorite first, then your second favorite 08:45:28 yes 08:45:34 #vote 6 08:45:35 #vote 3 08:45:38 #vote 6 08:45:45 #vote 2 08:46:03 #vote 6 08:46:16 #endvote 08:46:16 Voted on "vote for your favorite topic?" Results are 08:46:16 3 (1): Raghav 08:46:16 2 (1): juhak 08:46:16 6 (3): SerenaFeng, jose_lausuch, morgan_orange 08:46:27 ok so topic 6 is selected 08:46:30 now second topic 08:46:43 #startvote vote for your favorite topic? 1, 2, 3, 4, 5, 7 08:46:43 Begin voting on: vote for your favorite topic? Valid vote options are 1, 2, 3, 4, 5, 7. 08:46:43 Vote using '#vote OPTION'. Only your last vote counts. 08:46:47 #vote 2 08:46:49 #vote 2 08:46:49 #vote 2 08:46:54 #vote 2 08:47:00 juhak: viktor_t there? 08:47:11 yes 08:47:15 I will put the proposals in the wiki for next time 08:47:25 #vote 2 08:47:32 and if people are happy to start working on vCDN...should not wait an internship :) 08:47:36 #endvote 08:47:36 Voted on "vote for your favorite topic?" Results are 08:47:36 2 (5): juhak, SerenaFeng, jose_lausuch, morgan_orange, Raghav 08:47:55 #info subject 2 (unit tests) will be publushed in the internship program page 08:48:19 #action morgan_orange pushich topic 2 and 6 08:48:32 #topic Preparation D release 08:48:59 I refactored the page https://wiki.opnfv.org/display/functest/Functextnexttaks 08:49:09 list of JIRAs for D 08:49:17 and also a figure 08:49:33 #info page for D refactored 08:49:51 #link https://wiki.opnfv.org/display/functest/Functextnexttaks 08:50:02 #info figure created regarding the discussion on templating 08:50:05 #link https://wiki.opnfv.org/download/attachments/2925495/Functest-feature-project-interfaces.png?version=1&modificationDate=1473091940000&api=v2 08:50:24 I tried to capture the interactions of the feature test project 08:51:06 SerenaFeng: already made some comments (you can add them in the wiki page) 08:51:34 ok, I will modify it after the meeting 08:51:50 we should try to abstract and APIze in D 08:52:05 there are probably 2 different steps 08:52:14 the project declaration oen click to create all the resources 08:52:42 and an extended APi to push results, logs, conenct to nodes, select your cases,/manage black lists 08:53:14 we will discuss that next week and we may also invite people from OAI who are interested to include some of their dev into Functest (vEPC) 08:54:26 I also affected the JIRA on Snaps to Steven from Cable Labs...I do not know the smoothest way...probably commit SNAPS and then see how to use it from existing case... 08:55:14 #action everybody review and comment D wiki page 08:55:53 A first presentation on OAI (OpenAirInterface) should be done this week during the Test weekly meeting 08:56:03 we should also plan a sync with Intel team on 01.org 08:56:10 #topic AoB 08:56:16 any info/comment 08:56:56 any information about the functest Release D meetup? 08:57:21 several peopel will be in Barcelona, so I suggest to organize it during the Summit 08:57:26 who will be in barcelona? 08:57:56 I will be there 08:57:58 I will be there if nothing wrong with my visa apply 08:58:21 I assume jose_lausuch and myself too (maybe Valentin...but he is back in school...) 08:58:24 I Dont know if I'll be there uyet 08:58:42 ah 08:58:45 ok 08:58:49 I'll try 08:58:58 Barcelona is end of October 08:59:21 yep 08:59:24 I assume next version will be in March/April 08:59:31 there is also sdn world congress in the hague 08:59:31 maybe 08:59:36 october as well 08:59:46 your boss was there last year in Dusseldorf :) 08:59:59 he will be in the hague this year 09:00:10 I will not 09:00:48 if needed I can organize a meetup in Orange (as we did in Espoo) 09:01:11 but not sure we will have as many people as in Barcelona...so it could be an argument to send you there :) 09:01:41 #info Functets meetup planned during OpenStack Summit Barcelona 09:02:10 #info if needed regional meetup could be organized in Orange (Lannion or Paris if more convenient) 09:02:31 any other info/questions? 09:02:46 ok have a good week 09:02:52 and enjoy all your AP 09:02:56 #endmeeting