08:00:08 #startmeeting Functest weekly meeting 19th Sep. 2017 08:00:08 Meeting started Tue Sep 19 08:00:08 2017 UTC. The chair is jose_lausuch. Information about MeetBot at http://wiki.debian.org/MeetBot. 08:00:08 Useful Commands: #action #agreed #help #info #idea #link #topic. 08:00:08 The meeting name has been set to 'functest_weekly_meeting_19th_sep__2017' 08:00:18 #info Jose Lausuch 08:00:31 #info Agenda https://wiki.opnfv.org/display/functest/Functest+5.+Meeting#Functest5.Meeting-19/09(8UTC) 08:00:49 #info Juha Kosonen 08:01:15 #info Morgan Richomme 08:01:47 #info Linda Wang 08:01:52 #info Cristina Pauna 08:02:39 #info Delia Popescu 08:02:55 #chair morgan_orange 08:02:55 Current chairs: jose_lausuch morgan_orange 08:03:00 #topic action points follow up 08:03:04 #info AP CristinaPauna give an update about supporting alpine on aarch64 08:03:17 what's the status on this? depo CristinaPauna ? 08:03:40 We have 2 changes opened 08:03:49 one patch for multiple dockerfiles 08:04:08 but they are based on functet-core_aarch64 08:04:36 so we need to build either images with architecture as tag 08:04:36 #link https://gerrit.opnfv.org/gerrit/#/c/41981/ 08:04:43 or with architecture in the name 08:04:45 #link https://gerrit.opnfv.org/gerrit/#/c/41677/ 08:04:47 #link https://gerrit.opnfv.org/gerrit/#/c/41677/ 08:04:51 #undo 08:04:51 Removing item from minutes: 08:05:02 you -1 youself 08:05:03 :) 08:05:27 yes, because we don't have functest-core_aarch64 image 08:05:36 we don't build it automatically 08:05:40 #info depo: we need to build either images with architecture as tag or with architecture in the name 08:05:44 shall we create it? 08:05:47 so my change will just break 08:05:47 it can be done today 08:06:08 let's just agree on the image name 08:06:10 #info depo: because we don't have functest-core_aarch64 imag (no automatic build) 08:06:25 whell, I say we create more like keeping the name of the image and set the architecture in the tag 08:06:40 this way we won't have multiple ocker images 08:06:50 I'd say the same, otherwise we need to duplicate all the images... 08:07:04 Im not sure that's the best solution. I must go deeply into details. 08:07:07 so I redo the patch, considering the tag 08:07:14 we can think of euphrates.5.0 and euphrates.5.0.aarch64 08:07:36 or whatever tag we create for euphrates 08:07:43 would that sound ok? 08:07:53 lije euphrates.5.0:aarch64 08:08:08 not sure you can do that 08:08:08 so aarch64 is the tag 08:08:16 no no 08:08:20 euphrates.5.0 is already the tag 08:08:27 a, ok 08:08:30 http://container-solutions.com/multi-arch-docker-images/ 08:08:31 than yes I agree 08:08:34 the image name is functest_core or features or smoke… 08:08:44 jose_lausuch: is it in line with mark and alec discussions (I did not follow, just saw the thread) 08:09:07 didn't read that yet carefully 08:09:09 but I think so 08:09:11 Please consider manifest as I wrote in reviews 08:10:02 #action depo update patches to create the containers with the same image name but different tag 08:10:19 #disagree 08:10:20 ok 08:10:33 ? 08:10:38 ollivier: what do you mean 08:10:45 http://container-solutions.com/multi-arch-docker-images/ 08:10:47 we were thinking to do something like this for functest too :https://hub.docker.com/r/opnfv/storperf-reporting/tags/ 08:11:23 No need to prefix or to add tags... 08:11:37 that's interesting 08:12:02 Yes. Why shouldn't we follow the Docker rules instead of creating OPNFV ones? 08:12:14 depo: can we look into these manifests? 08:12:51 yes, I'll look into it 08:13:18 ollivier: this manifest-tool doesn't come directly from docker, right? 08:13:43 https://docs.docker.com/registry/spec/api/ 08:14:22 ok, didn't know that 08:14:49 #action Delia look also into the possibility of using manifests not to duplicate images or create new tags 08:14:49 As I said in the review, we should go deeply into details. 08:15:45 can we at least merge this https://gerrit.opnfv.org/gerrit/#/c/41981/1 ? 08:16:18 yes, we need that one to red the architecture 08:16:27 it is not depended on the patch 08:16:35 ok 08:16:37 next APs 08:16:38 #info AP ollivier LindaWang create alpine container for RestAPI 08:16:58 Cedric has finished it. 08:17:17 LindaWang: I updated the doc with the new methods 08:17:27 #info DONE https://gerrit.opnfv.org/gerrit/#/c/41819/ 08:18:01 #info AP LindaWang get information from huawei-virtual5 (ram, cpu). Done, last week. Decission to remove vnf from virtual deployments. 08:18:01 morgan_orange Got it. thanks 08:18:31 #info AP LindaWang and cedric, update logging.ini to disable urllib3 log output 08:18:56 #info DONE https://gerrit.opnfv.org/gerrit/#/c/40677/ 08:19:05 Done, but still some warnings for rally and tempest 08:19:15 #info Done, but still some warnings for rally and tempest 08:19:30 #info AP Soumaya push code to repo for final integration 08:19:32 ? 08:19:40 #info done under review 08:19:52 can you paste the review 08:19:58 #link https://gerrit.opnfv.org/gerrit/#/c/41759/ 08:20:01 #link https://gerrit.opnfv.org/gerrit/#/c/41759/ 08:20:03 #undo 08:20:03 Removing item from minutes: 08:20:04 :) 08:20:25 #info AP bring the branching topic to release meeting. Done. Functest was branched the last. 08:20:27 we have a topic for that 08:20:33 #topic Branching Status 08:20:44 #info still lots of code violations + open question on some choices (apt-get install through command is very old style...) 08:20:50 #info Branch stable/euphrates created already 08:21:13 Hi jose .. 08:21:38 #link https://gerrit.opnfv.org/gerrit/gitweb?p=functest.git;a=shortlog;h=refs/heads/stable/euphrates 08:21:59 we will push final commit by oneor two days 08:21:59 #info Jenkins jobs on Euphrates already proposed but need to be adapted 08:22:06 Soumaya: hi, ok, thanks 08:22:34 #link https://gerrit.opnfv.org/gerrit/#/c/42255/ 08:22:45 what are we missing here? 08:22:50 what else do we need to do? 08:22:52 or support? 08:24:18 Have Jenkins jobs of installers on Euphrates been ready? 08:24:24 we are not responsible for that...error on Unknown entry point or macro 'apex-virtual-euphrates-defaults' for component type: 'parameter'. when running /opt/jenkins-ci/jobs/releng-verify-jjb/workspace/jjb/kvmfornfv 08:24:32 ok thanks 08:24:42 We can always blame Functest 08:24:45 the apex jobs are specific, Trevor needs to adapt them... 08:25:05 so,we can't do much for now 08:25:14 it doesn't depend directly on Functest as such 08:25:18 I imagine 08:25:32 fdegir: asked Trevor B. to complete apex stuff 08:25:43 I see some jobs created in fuel for euphrates for instance https://build.opnfv.org/ci/view/fuel/ 08:26:04 and functest is "unresolved" https://build.opnfv.org/ci/view/fuel/job/fuel-os-nosdn-kvm-ha-baremetal-daily-euphrates/ 08:26:04 sorry to crash the meeting 08:26:15 bramwelt fixed apex parts and I rebased the patch 08:26:18 it passed jenkins 08:26:19 https://gerrit.opnfv.org/gerrit/#/c/42255/ 08:26:29 so you can review/+2/submit as you wish 08:26:35 oh 08:26:38 just now 08:26:39 thanks Fatih 08:26:54 merged 08:26:58 thx jose_lausuch 08:27:08 thank you :) 08:27:22 my pleasure :) 08:27:24 fdegir: recheck is enough, no? 08:27:35 ollivier: nope, need to rebase to take apex updates in 08:27:40 so, inm theory the jobs are created and whenever there is an installer run on Euphrates, Functest+Yardstick will be launched 08:27:55 ok. I do it. 08:28:52 ollivier: what about our docker containers? are they prepared to build the image with the packages of the feature projects from euphrates branch? 08:30:56 and all the packaging and so on 08:31:04 I imagine it works 08:31:15 is there any impediment? 08:31:22 Of course not. It requires my patches to be merged ?! 08:31:44 can you list the patches so that we have a look now 08:31:50 and merge asap 08:32:02 ok. You are as reviewers 08:32:44 you mean all these? https://gerrit.opnfv.org/gerrit/#/q/owner:cedric.ollivier%2540orange.com+status:open 08:33:01 ??? 08:34:48 You may update functest-comitter if you want to receive them via suse 08:35:39 please all review those patches so we have euphrates branch working properly 08:35:42 I will review that later 08:35:48 and merge when ready 08:36:00 #topic Euphrates Status 08:36:24 #info No feedback from the community about keeping blue balls in Jenkins for Functest when some tests fail 08:36:49 we are rewritting this part 08:37:05 but people didn't complain so far… 08:37:30 Now we will show the real Functest status 08:37:34 with the patch submitted by Cedric, we distinguish what is our responsibility from third party 08:38:05 I think it's key regarding the last emails. 08:38:33 BTW I just updated jenkins view to integrated Euphrates jobs 08:39:01 ok,thanks. I wanted to do it as soon as we have some jobs running 08:39:05 but they haven't started yet 08:39:22 fdegir: do you know when installer jobs will start deploying on E- branch? 08:40:46 fdegir: I'm not sure it works. I rebased the full list of euphrates changes, add jenkins-ci as reviewer and forced recheck. Nothing... 08:41:03 compass has prepared the E-branch Job 08:41:09 https://build.opnfv.org/ci/search/?q=euphrates 08:41:12 But it has not run in Ci 08:42:18 as far as I can see there in only 1 deploy-euphrates (daisy) 08:42:38 yes morgan_orange https://build.opnfv.org/ci/job/daisy-build-daily-euphrates/ 08:43:21 Michael Pauls proposed functest: fix bug in orchestra: no return code https://gerrit.opnfv.org/gerrit/42383 08:43:41 so, we can't really say much today about this 08:43:51 and the status of master is kind of the same as last week 08:43:54 any addition? morgan_orange ? 08:44:43 Michael Pauls proposed functest: fix bug in orchestra: no return code https://gerrit.opnfv.org/gerrit/42383 08:45:23 ollivier: can you pass me the link please? 08:45:31 jose_lausuch: I'm not sure when the installer jobs start running 08:45:43 jose_lausuch: the schedule/triggers are controlled by installers 08:45:44 jose_lausuch: a little bit better, trends seem to increase 08:45:47 fdegir: https://gerrit.opnfv.org/gerrit/#/c/42321/ 08:45:49 Daisy is almost fully sunny 08:46:07 ollivier: looking... 08:46:08 I could prepare a summary if you want 08:46:17 morgan_orange: I see, 2 scenarios almost sunny, that's good 08:46:22 if you don't mind 08:46:29 then I can report that this afternoon 08:46:33 Could we abandon 2 patches for oblsolete and inactivity ? https://gerrit.opnfv.org/gerrit/#/c/37761/ https://gerrit.opnfv.org/gerrit/#/c/37263/ 08:46:33 without too much detail 08:46:44 fdegir: ok, just in case release management asks us 08:47:06 ollivier: done 08:47:58 #action morgan_orange prepare scenario summary 08:48:12 thanks 08:48:18 ollivier: my rebase comment was about releng change to take apex updates in 08:48:34 ollivier: your recheck comment is correct for your patches 08:48:38 fdegir: I was suprised but I obey :) 08:49:20 regarding feature integration, I see 3 topics (barometer/alpine => Jira created), Parser , and issue with return statement on functest-vnf (orchestra do not push results to DB leading to an error return code and red status on os-nosdn-nofeature) 08:49:22 ollivier: and it was my mistake in the end 08:49:33 ollivier: I created jobs in disabled mode 08:49:47 ollivier: and this fixes it 08:49:51 ollivier: morgan_orange: jose_lausuch: https://gerrit.opnfv.org/gerrit/#/c/42385/ 08:50:24 fdegir: thanks fdegir 08:50:29 morgan_orange: does this fix the problem? https://gerrit.opnfv.org/gerrit/#/c/42383/ 08:51:04 ah no 08:51:09 sorry, just read the comments 08:51:11 I do not think so, that was my patch I abandonned after Valentin's feedback...need to dive a little bit more in the issue 08:51:13 ollivier: please wait few minutes before putting recheck to your patch 08:51:25 morgan_orange: I think you forgot several topics. 08:51:29 ollivier: it takes about 3 minutes for updates to jobs to be reflected to jenkins 08:51:49 ollivier: probably... 08:52:10 morgan_orange: but it's quite good ;) 08:52:11 doctor is also probably on the spot but I am not in review 08:52:24 You are 08:52:27 I saw Domino was PASS 08:52:39 https://gerrit.opnfv.org/gerrit/#/c/41823/ 08:52:40 Promise is still FAIL but no feedback 08:52:49 morgan_orange: https://gerrit.opnfv.org/gerrit/#/c/41823/ 08:53:08 ollivier: you can recheck again now 08:53:15 let's go 08:53:22 oops I was in rreview.. :) 08:53:24 ollivier: and you can recheck all the patches on stable/euphrates since none of them got verified 08:54:05 fdegir: yes I'm doing it. It runs let's see if tox is well configured 08:54:21 ollivier: ok - ping me if something doesn't look right from releng point of view 08:54:44 fdegir: sure. It should be ok. thank you. 08:55:11 #topic AoB 08:55:23 regarding interns, we hope to have the interviews this week 08:55:40 not much to say 08:55:51 do we have enough mentors? 08:56:01 serena for test api 08:56:11 for kubernetes, I can ask Narinder 08:56:19 Unfortunately I would not be able to mentor this time as I will move to ONAP after the release 08:56:21 I can also be, but I have no clue about kubernetes 08:56:53 we also have security groups request, but that's outdated, we will need to ask also for investigation 08:56:57 I think we need at least 2 mentors..we know that usually only one is following but it is safer for the internship 08:57:24 LindaWang: are you mentoring anyone now? 08:57:37 not yet 08:57:42 would you like? 08:58:05 But I am sure if I could do well. :) 08:59:18 you can try :) 08:59:26 we will talk again about this 08:59:29 need to leave now 08:59:39 jose_lausuch: Can you prepare a summary for feature project - no idea the status with bgpvpn, sfc, ... 08:59:39 OK. thanks 08:59:43 Ok I see we don't list any action 08:59:51 if there is nothing else, we close the meeting 08:59:59 Why do we send to features? Do we remove doctor? 09:00:07 What? 09:00:35 morgan_orange: maybe a summary of Functest dev too 09:01:07 what do you mean by Functest dev? 09:02:08 I will write the changelog for the last month 09:02:43 morgan_orange: can you close the meeting when finish? need to run to another meeting 09:03:00 who shall we action for the sync with feature projects? 09:03:19 I think we must action somebody...you or me ..but it must be done.. 09:03:34 morgan_orange: you have done that so well :) 09:04:16 #action morgan_orange prepare sync on status feature project integration 09:04:44 ollivier: did you have time to have a look at the issue reported by barometer? 09:04:56 I commented the review. 09:05:26 #endmeeting