08:00:15 <morgan_orange> #startmeeting Functest weekly meeting March 14th 08:00:15 <collabot`> Meeting started Tue Mar 14 08:00:15 2017 UTC. The chair is morgan_orange. Information about MeetBot at http://wiki.debian.org/MeetBot. 08:00:15 <collabot`> Useful Commands: #action #agreed #help #info #idea #link #topic. 08:00:15 <collabot`> The meeting name has been set to 'functest_weekly_meeting_march_14th' 08:00:21 <morgan_orange> #topic call role 08:00:24 <morgan_orange> #info Morgan Richomme 08:00:34 <CristinaPauna> #info Cristina Pauna (ENEA) 08:00:39 <juhak> #info Juha Kosonen 08:00:42 <ollivier> #info Cédric 08:00:48 <leo_wang> #info leo wang 08:00:57 <MatthewLi> #info Jun Li 08:01:07 <HelenYao> #info Helen Yao 08:01:09 <morgan_orange> #info agendat for today => https://wiki.opnfv.org/display/functest/Functest+Meeting 08:01:18 <LindaWang> #info Linda Wang 08:01:23 <morgan_orange> #topic action point follow up 08:01:28 <ChrisPriceAB> #info Chris Price (irc only) 08:01:33 <morgan_orange> #info AP1, AP2, AP3, AP4 juhak give feedback on tempest_smoke & rallay_sanity error => see next topic 08:01:38 <morgan_orange> #info AP5 juhak troubleshooting multisite 08:01:44 <morgan_orange> #info AP6: ollivier add a wiki page about review best practices 08:02:00 <ollivier> AP6 still pending 08:02:00 <morgan_orange> #action ollivier add a wiki page about review best practices 08:02:08 <morgan_orange> #info note a page for Euphrates has been initiated 08:02:18 <morgan_orange> #link https://wiki.opnfv.org/display/functest/Functest+Euphrates+page 08:02:39 <morgan_orange> we will nto have time to discuss it this week , but we may plan a topic next time of after the release... 08:02:45 <morgan_orange> #info AP7: morgan_orange ollivier clarify python way for feature projects inhering feature based - see impact on sfc, sdnvpn 08:03:15 <morgan_orange> #info done (security_scan not ready for a cleaner integration => , patch merged) (security scanning OK) 08:03:21 <morgan_orange> #info AP8: ollivier see what is wrong on odl_l3 scenario => see next topic 08:03:26 <morgan_orange> #info AP9: SerenaFeng check testapi fixed are included on testresults.opnfv.org, if not apply them and maybe enable automatic mechanism now.. 08:03:32 <morgan_orange> #info done, testapi is now fully automated from releng to testresults.opnfv.org and artifacts => congratulation to Serena and Rohit 08:03:39 <morgan_orange> #info AP10: morgan_orange => share the mail with scenario owners (any alias?) 08:03:44 <morgan_orange> #info info shared during weekly meeting, I also reported that the wiki page referencing the scenario does not seem fully accurate anymore 08:03:50 <morgan_orange> #info AP11 steve and morgan_orange, see if SNAPS smoke is compatible with lxd (probably not...) if not exclude it 08:04:00 <morgan_orange> #info done, SNAPS smoke excluded for this scenario, JIRA created for adaptation in Euphrates 08:04:06 <morgan_orange> #info AP12 morgan_orange create Danube branch March 10th 2PM CET 08:04:11 <morgan_orange> #info branch created by aric on friday 08:04:20 <morgan_orange> #info jenkins board modified, but no danube CI run seen so far 08:04:26 <morgan_orange> #info only Danube runs come from Cisco labs (not connected to CI but pushing results in DB) 08:04:31 <morgan_orange> #info reporting page update 08:04:36 <morgan_orange> #link morgan_orange adapt reporting to include Danube version 08:04:41 <morgan_orange> #link http://testresults.opnfv.org/reporting/functest/release/danube/index-status-apex.html 08:04:46 <morgan_orange> #info AP13: all close the JIRA that are fixed to clarify 08:04:52 <morgan_orange> #info still pending... 08:04:57 <morgan_orange> #info AP14 morgan_orange clean JIRA for Danube 08:05:01 <morgan_orange> #info in progress 08:05:25 <morgan_orange> ChrisPriceAB: do not worry, it is always IRC only 08:05:34 <morgan_orange> any question on the action point from last week? 08:05:37 <ChrisPriceAB> :) 08:05:56 <morgan_orange> #topic Danube status 08:06:07 <morgan_orange> I ma preparing a mail on the topic, 08:06:22 <morgan_orange> no fuel or apex run sicne several days.... 08:06:45 <morgan_orange> tim reported on issue on danube first try 08:06:54 <morgan_orange> we must regenerate a functest:stable docker 08:07:07 <morgan_orange> I modified the jenkins panel 08:07:35 <morgan_orange> HelenYao: do you know if I have only to initiate the docker-build-push-danube.... 08:07:58 <morgan_orange> danube jobs are using functest:stable (that is currently still colorado based) 08:08:16 <HelenYao> I have to check the jjb first 08:08:29 <HelenYao> the stable is outdated 08:08:35 <morgan_orange> ok can I action you on that (production of a new stable 08:08:45 <HelenYao> sure 08:08:53 <morgan_orange> #action HelenYao generate new functest:stable based on danube 08:08:57 <morgan_orange> thanks 08:09:24 <morgan_orange> #info new danube jobs call functest:stable (outdated still based on colorado) 08:09:36 <morgan_orange> #info danube Cisco run (not connected to CI) still used latest 08:10:09 <morgan_orange> #info as a consequence on danube job we only see fdio scenario today 08:10:18 <morgan_orange> #info still needs to consider master results 08:10:36 <morgan_orange> juhak: regarding the different AP you have on the different scenarios/installer, do you have an update 08:11:14 <juhak> there are a lot of scenarios, still working of this 08:11:17 <morgan_orange> #info apex 08:11:18 <morgan_orange> #link http://testresults.opnfv.org/reporting/functest/release/master/index-status-apex.html 08:11:40 <juhak> apex, failures on os-odl_l3-nofeature-ha and os-odl-bgpvpn-ha. Error details missing, needs https://gerrit.opnfv.org/gerrit/#/c/30187/ (merged 10.3) to get further info 08:12:46 <morgan_orange> it seems that all the tempest_smoke_serial secnarios (whatever the installer) have issues with odl_l3 scenarios => maybe need to restrict the list of tests if they do not make sense (if that is true errors, we shall remain them of course) 08:13:23 <juhak> they may be related to tempest/paramiko bug: https://bugs.launchpad.net/tempest/+bug/1577632 08:13:48 <pma> https://gerrit.opnfv.org/gerrit/#/c/30341/ 08:14:03 <morgan_orange> #action morgan_orange contact bgpvpn team for apex 08:14:20 <MatthewLi> morgan_orange: https://build.opnfv.org/ci/view/functest/job/functest-docker-build-push-danube/ 08:14:42 <MatthewLi> need to trigger this to generate STABLE 08:14:43 <juhak> pma: thanks 08:15:16 <morgan_orange> MatthewLi: yep I added this menu this morning, was just asking Helan to check if running the job is enough or if the job needs to be adapted 08:15:37 <morgan_orange> #info patch submitted for doctor but no apex run on master since more than 5 days.... 08:15:38 <MatthewLi> morgan_orange: think only releng committer have priviledge 08:15:42 <morgan_orange> #info so not possible to test the patch 08:16:35 <morgan_orange> #info warning no run on Master during 5 days 10 days before the release....it is a bit hot.. 08:17:06 <morgan_orange> #info apex scoring will probably decrease on master due ti lack of CI runs 08:17:09 <morgan_orange> #info compass 08:17:20 <morgan_orange> #info new openo scenario sucessfully itnegrated 08:17:36 <morgan_orange> #info issues with odl scenarios (odl_l2 and odl_l3) 08:18:06 <morgan_orange> #info issues also with ocl (but only installer with such sdn controler) 08:18:13 <morgan_orange> #info fuel 08:18:30 <morgan_orange> #info no run on master since 2 days 08:18:55 <morgan_orange> #still lots of scenario, relative good shape 08:19:14 <morgan_orange> #link http://testresults.opnfv.org/reporting/functest/release/master/index-status-fuel.html 08:20:22 <morgan_orange> also tempest_smoke _serial to be checked in several scenarios, juhak any update for fuel? 08:21:12 <juhak> there are different failure in several scenarios, working on this 08:21:21 <morgan_orange> ok thanks 08:21:25 <morgan_orange> #info joid 08:21:35 <morgan_orange> #link http://testresults.opnfv.org/reporting/functest/release/master/index-status-joid.html 08:22:03 <morgan_orange> #info lxd and nosdn scenario stable (only 1 error remaining on rally_sanity for lxd) 08:22:09 <juhak> rally/nova failed, but no data available 08:22:10 <morgan_orange> #info odl scenario also failing 08:22:20 <juhak> "Not possible to push results to artifact: some error happened when using gsutil" 08:22:26 <juhak> I'll follow this 08:22:29 <morgan_orange> there is a pending jira, resutls are not pushed to artifacts ion huawei-12 08:23:32 <morgan_orange> that is for the current status 08:23:58 <morgan_orange> for fuel barometer scenario, patched has been submitted yesterday, wait for CI run to check the effect 08:24:21 <morgan_orange> any update in Daisy? 08:25:01 <morgan_orange> #info no update on Daisy 08:25:11 <morgan_orange> #topic Danube Documentation 08:25:37 <morgan_orange> #info Functest documentation had been updated (first iteration) several weeks ago 08:25:54 <morgan_orange> #info as we are approaching we need to do the second iteration 08:26:10 <morgan_orange> #aciton morgan_orange review user/conf/dev guide and put contributors in review 08:26:34 <morgan_orange> #info there is still also possibility to including other test prohects and review also their documentation 08:27:04 <morgan_orange> #link https://wiki.opnfv.org/pages/viewpage.action?pageId=10289177 08:27:37 <morgan_orange> #info feel free to add your name to review the doc of another project (last time review were very valuable...) 08:28:00 <morgan_orange> #info release note to be initated 08:28:08 <morgan_orange> #action morgan_orange prepare release note 08:28:18 <morgan_orange> #info release note will be validated by José 08:28:18 <ollivier> morgan_orange: issues vping_ssh (Joid) as test cannot establish connection to IP '10.6.15.12'. Not related to functest test. 08:28:39 <morgan_orange> ollivier: ok 08:28:48 <morgan_orange> any question on the documentation? 08:29:10 <ollivier> morgan_orange: I'm facing a similar issue here. I am investigating 08:29:34 <morgan_orange> #topic RefStack 08:30:06 <morgan_orange> #link https://gerrit.opnfv.org/gerrit/#/c/30225/ 08:30:40 <morgan_orange> #info patch ready to be merged, introducing refstack testing as a new test case (i.e. running Refstack suite launched from refstack client) 08:30:56 <morgan_orange> #info test case using testBase abstraction, so results pushed to the DB 08:31:32 <morgan_orange> #info mail this morning refstack could be fine to consume our testAPI but we need to put all the results in the detaisl section (not only the errors or skipped) 08:31:38 <MatthewLi> will amend 30225 soon 08:32:06 <morgan_orange> #info this test will replace defcore created by José but we can try some cycles with both and remove refcore once refstack is fine 08:32:46 <morgan_orange> MatthewLi: I jnow that dovetail roadmap is different from the release, the goal is to include this testcase in Danube, right? 08:32:59 <MatthewLi> morgan_orange: yep 08:33:15 <morgan_orange> in the same time need to https support, HelenYao provided a patch for that https://gerrit.opnfv.org/gerrit/#/c/30253/ 08:34:19 <morgan_orange> MatthewLi: OK so we can target merge today => check results on Master => if OK cherry pick on Danube ...timing is short, but I do not know if the release manage can reasonaby consider the 22 as the target... 08:34:43 <MatthewLi> morgan_orange: looks fine 08:35:00 <morgan_orange> HelenYao: can you explain the patch for https support 08:35:01 <MatthewLi> we can achieve that before 22 :) I think 08:35:22 <HelenYao> There is a request in the tech group 08:35:40 <morgan_orange> MatthewLi: yes but in theory we should have 4 succesful consecutive runs .... it takes some time... 08:36:07 <HelenYao> only os_cacert is required to be added to support https. Also, the check_os is changed a bit to support both http and https 08:36:08 <MatthewLi> morgan_orange: yep, let's see how it works in CI 08:36:58 <morgan_orange> HelenYao: I was a bit surprised, we do not need to deal with ceritificates at this stage? 08:37:01 <HelenYao> after the patch is merged, functest is able to run openstack which has http or https enabled 08:37:04 <MatthewLi> morgan_orange: by the way, the unittest will only to add some basic, since there maybe some adaption after CI checks..... 08:37:22 <MatthewLi> then we can add more unittest... 08:37:27 <HelenYao> morgan_orange: in the mail, functest is asked to support https 08:37:37 <morgan_orange> ok 08:38:46 <morgan_orange> MatthewLi: sorry to bother you with unit tests but we all should now consider unit tests when adding code. Thanks to Ahsish we got now a good base, not sure but probably the most complete of OPNFV project. It ensures some stability...we should not rely on 1 person but be all concerned by that..that is why I asked you to add them.. 08:38:55 <morgan_orange> at least cover the main stuff 08:39:10 <morgan_orange> for Euphrates - we may expect a 80% coverage before merge - 08:39:20 <MatthewLi> morgan_orange: will achieve this 08:39:22 <morgan_orange> ChrisPriceAB: MatthewLi LindaWang, anything else you want to add for refstack? 08:39:47 <morgan_orange> assuming that apex/fuel are not up, we may have to test it either on compass or joid 08:40:10 <morgan_orange> #topic weekly jobs 08:40:26 <morgan_orange> #info weekly job introduce on fuel nosdn-nofeature master 08:40:30 <ChrisPriceAB> Nothing to add other rthan we need to identify how we report from it. 08:40:50 <ChrisPriceAB> And store results (and details) consistently across projects. 08:41:14 <morgan_orange> #link https://build.opnfv.org/ci/job/fuel-os-nosdn-nofeature-ha-baremetal-weekly-master/ 08:41:23 <morgan_orange> #info jobs can be triggered manually from jenkns 08:41:35 <morgan_orange> #info such jobs shall allow the testing of VNF on boarding features 08:41:46 <morgan_orange> #info cloudify_ims tested 08:42:05 <morgan_orange> #info patch merged to avoid exit on vnf tests to see results of orchestra and vrouter 08:42:19 <morgan_orange> #info as soon as fuel is back in town we could try to trigger the job to see that 08:43:06 <morgan_orange> hideyasu: in theory next time we run the job, we should be able to see the vrouter results...I was expecting that over last week end, but fuel installation is failing at the moment... 08:43:34 <morgan_orange> any comment... 08:43:36 <hideyasu> ok, thank you 08:43:50 <morgan_orange> #topic plugfest 08:44:10 <OPNFV-Gerrit-Bot> Michael Polenchuk proposed functest: [tempest/defcore] Configure volume device name https://gerrit.opnfv.org/gerrit/30161 08:44:28 <morgan_orange> #link https://wiki.opnfv.org/display/EVNT/Plugfest+-+Danube+Release 08:44:42 <morgan_orange> juhak: I did not see any participants from Nokia yet, do you plan to join? 08:45:23 <morgan_orange> Several people from Functest will attend => as usual provides support on running the tests + plugfest test DB,... 08:45:24 <juhak> we are planning, but no official decision yet 08:45:56 <morgan_orange> I read yesterday the the number of seats was limitted to 100 and there were for the moment 49 people registered... 08:46:07 <morgan_orange> so if you plan to go, do not forget to register... 08:46:20 <juhak> yep, I know... 08:46:41 <morgan_orange> I asked during the last weekly meeting if it would have been possible to run bottlenecks+ yardstick stress test defined for Danube 08:47:08 <morgan_orange> I sw gabriel_yuyang connected, maybe he has an update on the participation of people from bottlenecks/yardstick to the event? 08:47:25 <morgan_orange> otherwise we will probably discuss the roadmap for Euphrates 08:47:38 <morgan_orange> #info several functesters will attend 08:47:56 <morgan_orange> #info usual job: explain and support functest towards the different SUT 08:48:06 <morgan_orange> #info probably discussion on Functest roadmap 08:48:29 <morgan_orange> HelenYao: will you be in Paris? 08:48:42 <HelenYao> morgan_orange: i am afraid not 08:49:03 <morgan_orange> ok 08:49:08 <gabriel_yuyang> morgan_orange: for plugfest, I may not be able to attend. I am looking for proxy now. 08:49:28 <morgan_orange> hongbo will be there but he will probably deal with dovetail 08:50:19 <morgan_orange> any other info/questions on the plugfest? 08:51:07 <morgan_orange> it could be also the right place for a new demo of the VNF catalogue 08:51:08 <gabriel_yuyang> morgan_orange: could we do demo remotely? 08:51:22 <morgan_orange> Kumar made a demo during the weekly test meeting, we could replay 08:51:29 <morgan_orange> gabriel_yuyang: sure it shall be possible 08:51:50 <morgan_orange> #action morgan_orange initiate a wiki page for Functest@plugfest - plan remote demo for stress tests 08:52:15 <morgan_orange> #info demo on the Open Source VNF catalogue could be proposed (interesing for the end user) 08:52:30 <morgan_orange> #link https://wiki.opnfv.org/display/DEV/VNF+Catalogue 08:52:37 <morgan_orange> #topic AoB 08:52:46 <morgan_orange> #info CFP Beijing Summit 08:53:13 <gabriel_yuyang> morgan_orange: I mean if I can not attend. I connect remotely and do the demo. Or play some video. 08:53:45 <morgan_orange> #info etherpad page initiated to reference possible topics for the testing community 08:53:47 <morgan_orange> #link https://etherpad.opnfv.org/p/TestCommunitySummit 08:54:04 <morgan_orange> #info page shared on the wg mailing list (no comments/remarks so far) 08:54:11 <morgan_orange> #info deadline 27th of March 08:54:24 <morgan_orange> do you see valuable topics for Functest 08:54:39 <morgan_orange> what would be interesting/useful to report to the community 08:54:48 <morgan_orange> it can be things that are fine, or not.... 08:55:12 <morgan_orange> easy or challenging... 08:55:52 <morgan_orange> I think we will give an update on Functest (what is new, what has been changed, vision,...) 08:56:21 <morgan_orange> but we may imagine other topics: difficulty to troubleshoot, evolution to alpine, micro service approach to be consolidated,... 08:56:58 <morgan_orange> ok for Beijing 08:57:05 <morgan_orange> any other topic you want to raise for today? 08:57:36 <morgan_orange> i can see that during the meeting there were 2 attempts to produce the stable functest docker https://build.opnfv.org/ci/view/functest/job/functest-docker-build-push-danube/ 08:57:41 <morgan_orange> without success.. :) 08:58:30 <morgan_orange> MatthewLi: HelenYao will have a look at the job... 08:58:51 <HelenYao> morgan_orange: I kicked off the build 08:59:10 <morgan_orange> ok thanks 08:59:41 <morgan_orange> no other point? I think it will be hard for the release date 09:00:17 <morgan_orange> we can see in the scenarios that we have compared to Colorado some new scenarios but some are no more available (no more onos scenarios with any installer) 09:01:09 <morgan_orange> that is probably life..but it could be seen as a form of unstability and/or difficulty to maintain scenarios in release in a sustainable way... 09:01:15 <morgan_orange> ok it is time 09:01:44 <morgan_orange> José will be back next week as a married man, full of energy... 09:02:15 <morgan_orange> thanks for joining, enjoy troubleshooting, docker production, and last danube patches 09:02:17 <morgan_orange> #endmeeting