08:01:10 <ollivier> #startmeeting Functest weekly meeting 15 May. 2018 08:01:10 <collabot`> Meeting started Tue May 15 08:01:10 2018 UTC. The chair is ollivier. Information about MeetBot at http://wiki.debian.org/MeetBot. 08:01:10 <collabot`> Useful Commands: #action #agreed #help #info #idea #link #topic. 08:01:10 <collabot`> The meeting name has been set to 'functest_weekly_meeting_15_may__2018' 08:01:13 <ollivier> Hi 08:01:19 <ollivier> #info Cédric Ollivier 08:01:30 <juhak> #info Juha Kosonen 08:01:37 <viktor_t> #info Viktor Tikkanen 08:03:03 <LindaWang> #info Linda Wang 08:04:52 <djkonro> #info Konrad Djimeli 08:05:40 <ollivier> Hello. Let's start. Cristina or Delia may connect later. 08:05:46 <ollivier> #topic Action point follow-up 08:06:06 <ollivier> #info refstack_client is still incompatible with OpenStack Queens (then OPNFV Gambia) 08:06:55 <depo> #info Delia Popescu 08:07:11 <ollivier> As far as I know nothing has been merged but the topic is still opened in OpenStack refstack-client. 08:07:28 <ollivier> LindaWang: do you confirm? 08:07:43 <ollivier> BYW, the dummy virtualenv is being removed. 08:07:45 <LindaWang> yes, it is still under review 08:07:55 <LindaWang> #link: https://review.openstack.org/#/c/562956/ 08:08:21 <ollivier> #link https://review.openstack.org/#/c/564080/ 08:09:54 <ollivier> other actions will be described in topics 08:09:59 <ollivier> #topic Functest Fraser 08:10:33 <ollivier> I propose a full overall status before Fraser.1 (10 days) 08:10:48 <ollivier> #info Daisy passed Functest gate 08:11:00 <ollivier> #link https://build.opnfv.org/ci/view/functest/job/functest-daisy-baremetal-daily-fraser/lastSuccessfulBuild/console 08:11:13 <ollivier> #info cloudify_ims failed in both fraser and master jobs (CloudifyClientError: 400: URL ... not found - can't download blueprint archive) 08:11:41 <ollivier> A fix has been proposed but I'm not sure the testcase is fine. 08:12:22 <ollivier> #link https://github.com/Orange-OpenSource/opnfv-cloudify-clearwater/commit/abd39d772c77317a25745feaaf312165e7588346 08:12:33 <ollivier> #link https://gerrit.opnfv.org/gerrit/#/c/57321/ 08:12:59 <ollivier> Let's check the next CI runs 08:13:06 <ollivier> #info only rally_sanity fails when verifying Compass 08:13:17 <ollivier> #link https://jira.opnfv.org/browse/COMPASS-597 08:13:48 <ollivier> #info only snaps_smoke fails when verifying Apex 08:13:59 <ollivier> #link https://jira.opnfv.org/browse/APEX-598 08:14:11 <ollivier> #info only cloudify_ims failed when verifying Fuel (amd64) 08:14:21 <ollivier> #link only cloudify_ims failed when verifying Fuel (amd64) 08:14:24 <ollivier> #undo 08:14:24 <collabot`> Removing item from minutes: <MeetBot.ircmeeting.items.Link object at 0x298a950> 08:14:37 <ollivier> #link https://jira.opnfv.org/browse/FUEL-373 08:14:48 <ollivier> #info 2 subtestcases failed when verifying Fuel arm64 08:14:57 <ollivier> #link https://jira.opnfv.org/browse/ARMBAND-384 08:15:05 <ollivier> #link https://jira.opnfv.org/browse/ARMBAND-385 08:15:26 <ollivier> I don't publish any report about Joid. gsutils are not installed. 08:16:15 <ollivier> Be free to comment if you can complete with Installer data. Except maybe IMS, Functest Fraser is working very well. 08:17:22 <ollivier> The question is more do we backport shade updates? it cleans all resources in master. 08:18:11 <ollivier> We have to care about https://jira.opnfv.org/browse/FUNCTEST-965 a simple API chage between Shade Pike and Shade Queens 08:18:40 <ollivier> What do you think? I would suggest to backport. 08:19:26 <LindaWang> Backport to Fraser? 08:19:31 <ollivier> yes 08:20:06 <LindaWang> Snaps works well for Fraser. 08:20:16 <LindaWang> But fine with me. 08:20:36 <depo> 727742 08:20:46 <ollivier> True. I'm just afraid about tenancy in VNF and cleaning. But from the time being, we can keep it asis. 08:20:46 <juhak> fine for me too 08:22:51 <ollivier> Thank you. Let's close first https://jira.opnfv.org/browse/FUNCTEST-965 and then we can backport it if needed. 08:23:10 <ollivier> #topic Functest Gambia 08:23:38 <ollivier> depo:is 727742 related to Functest? Or is a typo issue? 08:23:54 <ollivier> #info lots of testcases have been updated to shade 08:24:01 <depo> type issue, sorry :) 08:24:28 <ollivier> Great work done on vping, tempest, rally and odl. 08:24:54 <ollivier> It only remains vnfs now. 08:25:16 <ollivier> depo: no problem 08:25:32 <ollivier> depo: Could you please confirm that vping is now working on ARM? 08:26:09 <ollivier> #link https://gerrit.opnfv.org/gerrit/#/q/status:merged+project:functest+branch:master+topic:shade 08:26:19 <depo> no, I cannot. I tested it before sending the patch, but I don't have CI runnings. It should work , though ... 08:26:32 <ollivier> ok. 08:26:50 <ollivier> #info rally selects now the right Gnocchi endpoint 08:27:01 <ollivier> #link https://bugs.launchpad.net/rally/+bug/1770593 08:27:02 <juhak> thank you Cedric for detecting the problem 08:27:11 <ollivier> #link https://review.openstack.org/#/c/567832/ 08:28:16 <ollivier> no pb :) That's great to help OpenStack too. 08:28:52 <ollivier> juhak: I think they are now releasing multiple rally-openstack version. Should we continue to pull master or could we leerage on pypi versions? 08:29:35 <juhak> let's continue to pull master still for a while 08:29:47 <ollivier> ok. thank you 08:30:58 <ollivier> juhak: Would it be possible to fully test Rally subtests even if one service is missing? Fuel is closed to be fully verified master. 08:31:25 <ollivier> #link https://jira.opnfv.org/browse/FUNCTEST-966 08:31:37 <ollivier> #link https://build.opnfv.org/ci/view/functest/job/functest-fuel-baremetal-daily-master/182/console 08:31:55 <juhak> hmm, need to check 08:32:29 <ollivier> ok. The testcase must be failed but it would check nova, neutron. Or we can simply put Gnocchi at the end. 08:33:41 <juhak> putting end could be the quick solution 08:34:25 <juhak> but I'll check this 08:34:38 <ollivier> Thank you. 08:34:53 <ollivier> #info live migration (rally_sanity) failed on Orange ONAP Openlab (we are investigating it) 08:35:04 <juhak> Live migration has failed on Orange ONAP Openlab also earlier, 08:35:11 <juhak> at that time the problem was gone along with OpenStack reinstallation 08:35:18 <juhak> Morgan's comments on https://jira.opnfv.org/browse/APEX-564 08:36:25 <ollivier> yes. It seems on Orange side too but it seems related to the status of VMs after migration. 08:37:17 <juhak> they are not becoming ACTIVE? 08:37:39 <ollivier> Yes. VOLUME_RESIZED or something else. 08:37:51 <juhak> ok 08:38:20 <ollivier> I just had a very quick look and most of Orange contributors were on holidays for 2 weeks :) 08:39:36 <ollivier> #info tempest post processing was updated 08:39:39 <ollivier> #link https://gerrit.opnfv.org/gerrit/#/c/57049/ 08:40:42 <ollivier> LindaWang: Could you confirm it works now at least in fraser? If I understood correctly, master postprocessing is failing. 08:41:06 <LindaWang> ollivier: yes, it works for Fraser already. 08:41:21 <ollivier> and failing in master? 08:41:50 <LindaWang> tempest.log does not include the info of each sub tests in master 08:42:39 <LindaWang> so we can not obtain the `details` (success, failures, skipped tests) from tempest.log 08:43:08 <ollivier> Rally may provide it too. no? 08:43:47 <LindaWang> it seems some update with "rally verify start ..." 08:44:31 <LindaWang> but tempest-report.html works well and it is enough for functest 08:44:54 <LindaWang> the `details` info is required by Dovetail 08:45:21 <ollivier> Yes. But it makes sense to put that details in the DB. 08:45:46 <ollivier> +Dovetail 08:46:05 <LindaWang> exactly. xudan from Dovetail is working on that issue. 08:46:23 <ollivier> ok. Thank you. 08:46:43 <ollivier> #topic Functest Kubernetes 08:46:56 <ollivier> #info we can hardly define default values with the current Xtesting design 08:47:01 <ollivier> djkonro: Thank you. 08:47:28 <ollivier> #link https://gerrit.opnfv.org/gerrit/#/c/56625/ 08:47:48 <ollivier> I will update Xtesting to limit env vars for K8s too. 08:48:35 <djkonro> @ollivier: ok, thanks. 08:48:35 <collabot`> djkonro: Error: "ollivier:" is not a valid command. 08:48:43 <ollivier> BYW, even EXTERNAL_NETWORK is useless in case of Functest. All testcases can be ran without any Functest env vars. 08:49:06 <djkonro> ollivier: ok, Thanks 08:49:41 <ollivier> djkonro: I think we would have to load the default values from an external files. Else It doesn't fit well the classical inheritance. 08:50:01 <ollivier> From the time being, the testcases are working well. 08:50:28 <ollivier> #link https://build.opnfv.org/ci/view/functest/job/functest-compass-baremetal-daily-master/lastSuccessfulBuild/ 08:50:29 <djkonro> ollivier: yes, that seems fine. 08:51:40 <ollivier> LindaWang: I think deployments are still failing in Fraser. Do you have any input from Compass? 08:52:18 <ollivier> djkonro: If I remember well your presentation, we could increase the subtestcases ran in healthcheck. 08:53:01 <LindaWang> ollivier: Compas Fraser deployment is fine as far as i know 08:53:29 <ollivier> #info https://build.opnfv.org/ci/view/functest/job/functest-compass-baremetal-daily-fraser/96/console (stor4vnf) 08:53:51 <LindaWang> oh, you mean k8 scenario 08:54:01 <ollivier> Yes ;) 08:54:27 <djkonro> LindaWang: yes, we can increase the subtest. I would look into it. 08:54:40 <ollivier> I haven't seen any full success vs Compass master 08:54:46 <ollivier> Compass Fraser 08:55:26 <ollivier> djkonro: thank you. 08:55:40 <ollivier> #topic AoB: 08:56:31 <ollivier> I forgot the new testcase under review https://gerrit.opnfv.org/gerrit/#/c/57257/ 08:57:07 <depo> I am now making the changes you mentioned 08:57:32 <ollivier> great. thank you. Else it seems fine. 08:57:39 <depo> thank you 08:58:10 <ollivier> I am not sure we have to duplicate a new config entry. Cirros could be reused in all testcases except vnfs. 08:58:28 <ollivier> (vping, cinder, tempest) 08:59:23 <ollivier> That's not big deal. We can keep it asis and refactor later. 08:59:29 <ollivier> if needed 09:00:22 <ollivier> #info I will send you an introduction about AUTO/ONAP. 09:01:59 <ollivier> (we will setup dedicated meetings if needed about the possible interactions between AUTO and Functest) 09:02:15 <ollivier> Anything alse? 09:02:35 <LindaWang> no from my side 09:03:54 <ollivier> Great work! 09:03:57 <ollivier> #endmeeting