08:00:31 <ollivier_> #startmeeting Functest weekly meeting 10 July. 2018
08:00:31 <collabot> Meeting started Tue Jul 10 08:00:31 2018 UTC.  The chair is ollivier_. Information about MeetBot at http://wiki.debian.org/MeetBot.
08:00:31 <collabot> Useful Commands: #action #agreed #help #info #idea #link #topic.
08:00:31 <collabot> The meeting name has been set to 'functest_weekly_meeting_10_july__2018'
08:00:39 <ollivier_> #Info Cédric Ollivier
08:00:52 <viktor_t> #info Viktor Tikkanen
08:01:16 <depo> #info Delia Popescu
08:01:55 <ollivier_> #topic Action point follow-up
08:02:19 <ollivier_> #info tempest smoke runs subtests in parallel
08:02:37 <ollivier_> #link https://jira.opnfv.org/browse/FUNCTEST-989
08:02:54 <ollivier_> It seems ok https://build.opnfv.org/ci/view/functest/job/functest-fuel-baremetal-daily-master/lastFailedBuild/console
08:03:02 <viktor_t> Great!
08:03:21 <ollivier_> The tempest-based testcases was a little bit refactored.
08:03:48 <ollivier_> option can be set via testcases.yaml and then we reuse only one class
08:04:00 <ollivier_> (+regex)
08:05:02 <ollivier_> #link https://gerrit.opnfv.org/gerrit/#/c/59491/
08:05:50 <ollivier_> It makes sens to backport after few runs (without renaming the testcases)
08:06:26 <viktor_t> OK
08:07:02 <viktor_t> BTW, what is the status of refstack_defcore test cases? I saw in some log(s) that quite many of them are skipped since not found in tempest test case list...
08:07:22 <viktor_t> Is this refstack_defcore up to date?
08:07:24 <ollivier_> Yes. It's skipped
08:08:03 <ollivier_> refstack defcore list has be be synced before runnig them. But the code from refstack_client is deprecated
08:08:40 <ollivier_> We decided to implement the synch mechanism in Functest to be independant
08:09:19 <ollivier_> #link https://jira.opnfv.org/browse/FUNCTEST-977
08:09:39 <viktor_t> OK, thanks for the link
08:10:44 <ollivier_> Be free to work on it if you're interested.
08:12:01 <viktor_t> Let's see, I need to catch up first anyway...
08:12:21 <ollivier_> ok.
08:12:31 <ollivier_> #topic Functest Fraser
08:12:50 <ollivier_> #info snaps_smoke currently fails if external network is not owned by testing user
08:13:15 <ollivier_> Few tests has been temporarily bypassed.
08:13:25 <ollivier_> #link https://gerrit.opnfv.org/gerrit/#/c/59597/
08:13:43 <ollivier_> #link  https://gerrit.opnfv.org/gerrit/#/c/59595/1
08:14:11 <ollivier_> It seems fine: https://build.opnfv.org/ci/view/functest/job/functest-apex-baremetal-daily-fraser/257/console
08:15:18 <ollivier_> I hope the subtests will be fixed soon. Then we will revert the two previous changes.
08:15:28 <ollivier_> #info all scenarios were successfully backported
08:15:53 <ollivier_> They closed https://jira.opnfv.org/browse/FUNCTEST-970
08:16:23 <ollivier_> (vping_ssh/vping_userdata fail with fraser)
08:18:05 <ollivier_> any point? Else we can switch to Gambia.
08:18:23 <depo> not from me
08:18:37 <viktor_t> seems to be OK
08:18:55 <ollivier_> #topic Functest Gambia
08:19:06 <ollivier_> #info only rally_sanity and shaker failed vs Compass (shaker could run successfully next CI run)
08:19:35 <ollivier_> (I precise rally_sanity ran as expected)
08:19:50 <ollivier_> #link https://build.opnfv.org/ci/view/functest/job/functest-compass-baremetal-daily-master/180/consoleFull
08:20:02 <ollivier_> #link http://artifacts.opnfv.org/logs/functest/huawei-pod1/2018-07-09_05-43-41/rally/opnfv-nova.html
08:20:37 <ollivier_> It means that Functest is already very good.
08:21:33 <ollivier_> Fuel results are not so bad https://build.opnfv.org/ci/view/functest/job/functest-fuel-baremetal-daily-master/lastFailedBuild/console
08:22:19 <depo> pma is working on barbican
08:22:30 <depo> I will make time to look on shaker
08:22:54 <ollivier_> Great. I think the issue is related to the external connectivity (no download).
08:23:24 <depo> ok, I'll check
08:24:12 <ollivier_> But at least, lots of testcases already pass.
08:24:18 <depo> yes
08:24:51 <ollivier_> #info pma is working on barbican
08:25:09 <ollivier_> #action depo look on shaker results
08:25:21 <ollivier_> #info vmtp has been added as a new testcase
08:25:34 <ollivier_> #link https://gerrit.opnfv.org/gerrit/#/c/58401/
08:25:46 <ollivier_> #link https://jira.opnfv.org/browse/FUNCTEST-980
08:27:03 <ollivier_> I hope we will get results soon but master jobs are mostly stopped.
08:27:46 <ollivier_> #new opnfv/functest-healthcheck is well verified (except XCI due to its conception)
08:27:58 <ollivier_> #info new opnfv/functest-healthcheck is well verified (except XCI due to its conception)
08:28:14 <ollivier_> #link https://gerrit.opnfv.org/gerrit/59561
08:29:06 <ollivier_> Increasing to 10 min the boot duration (Cirros) seem too much.
08:29:45 <ollivier_> Even in case of "emulation"...
08:29:53 <depo> agree
08:30:25 <ollivier_> great
08:30:55 <ollivier_> #info vping_ssh is now fully covered
08:31:15 <ollivier_> it may ease covering cinder_test
08:31:18 <ollivier_> (model)
08:31:39 <depo> yes, I saw it, I follow the model and working on it
08:31:44 <depo> thank you very much
08:31:47 <ollivier_> yes. I can help you this week.
08:32:24 <depo> I wil send you an email if I get stuck and with what
08:32:28 <depo> thank you
08:32:34 <ollivier_> ok.
08:33:08 <viktor_t> BTW, how about vmtp failure (in https://build.opnfv.org/ci/view/functest/job/functest-fuel-baremetal-daily-master/lastFailedBuild/console), does it wait 2 min for server to boot?
08:34:08 <ollivier_> I may be improved in a second change. For the time being, vmtp boots the 2 VMs.
08:34:43 <ollivier_> then it selects the timeout.
08:35:24 <ollivier_> http://artifacts.opnfv.org/logs/functest/lf-pod2/2018-07-10_06-33-32/vmtp/vmtp.conf
08:35:37 <ollivier_> We could decrease generic_retry_count
08:36:55 <ollivier_> But yes, we must follow the same rules if possible.
08:37:08 <ollivier_> 1 could be enough compared to the other testcases.
08:38:37 <ollivier_> #info Morgan declared the new testcases in DB
08:40:04 <ollivier_> A new scenario has been published this morning (cloudify). It will avoid duplicating code between vims and vrouter.
08:41:02 <ollivier_> I will do my best to refactor them before the end of this week.
08:42:26 <ollivier_> any additional point about Gambia?
08:43:59 <depo> it's fine with me
08:44:03 <viktor_t> not from my side
08:44:35 <ollivier_> #topic AoB
08:45:08 <ollivier_> I think I will be on holidays next week (for 3 or 4 weeks)
08:46:23 <depo> goot to know, I will bother you this week:)
08:46:30 <ollivier_> Yes!
08:47:08 <viktor_t> If you have time, check https://etherpad.openstack.org/p/ECG_Keystone_Testing (it's about Keystone federation)
08:47:26 <viktor_t> There is a question there: "What the OPNFV team should have on their testing agenda to avoid overlapping?"
08:47:45 <viktor_t> It may concern more EdgeCloud or Yardstick projects, but anyway...
08:47:59 <ollivier_> sure.
08:48:20 <ollivier_> It's an hot topic here ;)
08:49:47 <depo> one question, who you be your backup during vacation? if we need to merge anything
08:51:13 <ollivier_> I will merge patches if really required. I will check who is here. Viktor, Valentin and Juha are allowed to do so.
08:51:29 <depo> ok, thanks
08:51:50 <ollivier_> It's the installer blackout. We are not closed to the release :)
08:52:11 <viktor_t> ?
08:52:36 <depo> yes, I know, just checking
08:52:49 <ollivier_> I mean they are few CI runs.
08:54:20 <ollivier_> I take the action.
08:54:27 <ollivier_> Anything else?
08:54:36 <depo> not from my side
08:54:45 <viktor_t> no
08:55:11 <ollivier_> Thank you very much. Great work!
08:55:14 <ollivier_> #endmeeting