08:00:29 <morgan_orange> #startmeeting  Functest weekly meeting January 12th
08:00:29 <collabot`> Meeting started Tue Jan 12 08:00:29 2016 UTC.  The chair is morgan_orange. Information about MeetBot at http://wiki.debian.org/MeetBot.
08:00:29 <collabot`> Useful Commands: #action #agreed #help #info #idea #link #topic.
08:00:29 <collabot`> The meeting name has been set to 'functest_weekly_meeting_january_12th'
08:00:36 <morgan_orange> #info Morgan Richomme
08:00:55 <amaged__> #info Ahmed Maged
08:00:58 <viktor_nokia> #info Viktor Tikkanen
08:01:38 <morgan_orange> #link https://wiki.opnfv.org/functest_meeting
08:02:12 <morgan_orange> #topic Last week action point follow up
08:02:22 <morgan_orange> #link http://ircbot.wl.linuxfoundation.org/meetings/opnfv-testperf/2016/opnfv-testperf.2016-01-05-08.00.html
08:02:40 <morgan_orange> #info 1: creds issue on joid fixed
08:03:14 <morgan_orange> #info 2: May-meimei, viktor_nokia I think troubleshooting on custom tempest is on going
08:03:21 <morgan_orange> May-meimei: , viktor_nokia last update?
08:05:02 <morgan_orange> #info 3: review of Milestone E. I got the feedback from Jose on the topic. Status has been transmitted to Debra
08:05:10 <viktor_nokia> No updates from my side...
08:05:33 <morgan_orange> #info 4: priorization, I planned to do that today depending on the status of some tasks
08:05:57 <viktor_nokia> I have one pending functest commit regarding cleaning volumes/networks, will finish it today...
08:06:26 <morgan_orange> #info 5: lixiaoguang May-meimei morgan_orange => preparation of onos within Functest doctor done, patch submited, open question on the repo (I put the one recommended for teston installation)
08:06:43 <morgan_orange> #link https://gerrit.opnfv.org/gerrit/#/c/6165/
08:07:23 <morgan_orange> #info 6: peter, amaged__ raghavendrachari any update on the synchro regarding odl test extension?
08:07:54 <morgan_orange> there was a target of something for end of last week but I did not see any patch in gerrit on the topic
08:08:27 <raghavendrachari> hi morgan
08:08:30 <raghavendrachari> gud day
08:08:44 <morgan_orange> hi
08:08:57 <raghavendrachari> i had created docker images with mininet installation on top of it and uploaded ,
08:09:44 <morgan_orange> ok we have a dedicated section to controller, we will come back on the test activities regarding the controllers
08:10:08 <morgan_orange> the point was to be sure that you, Peter and amaged__ were synchronized on the odl task...
08:10:36 <raghavendrachari> no
08:10:43 <jose_lausuch> #info Jose Lausuch
08:10:44 <morgan_orange> #topic B release follow-up
08:10:47 <jose_lausuch> sorry for the late
08:10:55 <morgan_orange> #info feature project code freeze last week
08:11:14 <morgan_orange> #info installer and test project code freeze planned for this week (today I think)
08:12:08 <jose_lausuch> installer as well?
08:12:17 <morgan_orange> #info it means that after this date only bug fixes will be admitted, I hope we still could develop code for integration as the installer did not really produce System Under test allowing the tests, so freezing both at the same time...
08:12:26 <morgan_orange> jose_lausuch: yes theoretically it was both
08:12:52 <morgan_orange> and it is a good point, I do not knwo if all the installers are ready (apex said that it was ok)
08:13:01 <morgan_orange> #info current status
08:13:08 <morgan_orange> #link https://build.opnfv.org/ci/
08:13:34 <jose_lausuch> ok
08:13:51 <morgan_orange> today we are able to run functest on joid/odl (stable & master), fuel/nosdn (stable&master), compass/nosdn (stable)
08:14:07 <morgan_orange> #info  able to run functest on joid/odl (stable & master), fuel/nosdn (stable&master), compass/nosdn (stable)
08:14:18 <morgan_orange> #info apex issue with creds under investigation
08:14:23 <jose_lausuch> morgan_orange: for joid, do we have to do some fixes regarding ips for odl, right?
08:14:39 <morgan_orange> #info all the config => troubleshooting  needed
08:15:01 <morgan_orange> #info especially on odl, as it was supposed to be under refactoring, we did not investigate and the error reasons
08:15:12 <jose_lausuch> https://build.opnfv.org/ci/view/apex/job/functest-apex-opnfv-jump-1-daily-master/lastBuild/console
08:15:44 <jose_lausuch> the IP and the credentials are correctly fetched in apex
08:15:50 <jose_lausuch> but the endpoint seems to be incorrect
08:16:00 <jose_lausuch> it could be due to a bad configuration of the jumphost
08:16:01 <morgan_orange> ok
08:16:04 <jose_lausuch> (can happen)
08:16:43 <morgan_orange> food for today investigation (it used to work and this mechanism did not change so even the patch to execute the docker command in sudo is maybe useless)
08:17:02 <morgan_orange> #info integration of other projects started but not done yet, as no testbed to test on
08:17:23 <morgan_orange> #topic focus on controller
08:17:25 <jose_lausuch> what do we do with ODL?
08:17:32 <jose_lausuch> we leave it as it is
08:17:33 <morgan_orange> #info ODL
08:18:52 <morgan_orange> as no odl patch for refactoring and test extension has been submitted and code is about to freeze, the most reasonable scenario (for me) is to keep the actual suite and troubleshoot it
08:18:56 <jose_lausuch> #info fix issues with IPs (rest api IP and neutron endpoint IP)
08:19:06 <morgan_orange> we may reconsider how to deal with odl extension for a SR1
08:19:16 <jose_lausuch> +1
08:19:24 <morgan_orange> it means that all the JIRA currently assigned to Peter will be de prioretized
08:19:34 <morgan_orange> amaged__: raghavendrachari what do you think?
08:20:06 <morgan_orange> we just could add to the current script the call to the API in order to push results
08:20:38 <raghavendrachari> ok
08:20:43 <morgan_orange> amaged__: ?
08:21:02 <jose_lausuch> well, the only little problem is that the script is bash
08:21:09 <jose_lausuch> and the code we use to push is python
08:21:19 <morgan_orange> it is a rest api....
08:21:21 <jose_lausuch> but I guess there will be similar ways to do it
08:21:21 <jose_lausuch> yes
08:21:22 <morgan_orange> at the end
08:21:27 <jose_lausuch> not difficult
08:22:14 <morgan_orange> shall we vote for this motion "odl test extension deprioritized for B release SR0, troubleshooting on the existing suite + connection to test DB only"
08:22:29 <morgan_orange> strategy for odl extension will be redefined for SR1
08:22:50 <jose_lausuch> ok
08:22:57 <jose_lausuch> formally?
08:22:59 <morgan_orange> we have to assign the tasks (troubleshooting, connection with API and strategy for SR1)
08:23:10 <jose_lausuch> I can take a task
08:23:29 <morgan_orange> I think you have a lot, I would prefer to assign them to amaged__ or raghavendrachari
08:23:44 <raghavendrachari> yes i too can take
08:23:44 <morgan_orange> amaged__: raghavendrachari OK
08:23:49 <raghavendrachari> ok
08:24:12 <morgan_orange> so raghavendrachari we will not use your mininet docker for SR0, but with amaged__ you may consider the odl extension strategy for SR1
08:24:24 <jose_lausuch> raghavendrachari: if you need any help, just ask here in the channel or private
08:24:32 <morgan_orange> ok that is clear for me, I will do the adjustment on Jira
08:24:39 <raghavendrachari> ok
08:24:48 <morgan_orange> raghavendrachari: and if you want to convert bash in python you have 1 day :)
08:24:56 <morgan_orange> we will help for the DB
08:25:03 <morgan_orange> #topic ONOS
08:25:14 <morgan_orange> lixiaoguang: May-meimei are you around?
08:25:54 <morgan_orange> I saw the first onos scenario in CI (fuel even if deployment failed), it means that it will soon ready
08:26:14 <morgan_orange> however if you can already make first integration tests based on the first patch, it would be great
08:26:44 <morgan_orange> this week we freeze the code, week 4 and 5 is planned for troubleshooting and doc and week 6 is the release date target...
08:27:19 <morgan_orange> #action May-meimei lixiaoguang try to test ONOS test run from functest asap, target lab should follow soon
08:27:37 <morgan_orange> #action morgan_orange reorganize Jira for odl
08:27:57 <morgan_orange> as far as I know, onos shall be available on the 4 installers...
08:28:24 <jose_lausuch> ok
08:28:31 <morgan_orange> lixiaoguang: May-meimei it will be good also to complete the doc on onos
08:28:36 <morgan_orange> for the testing
08:28:46 <morgan_orange> test user guide and config/installation guide
08:29:43 <morgan_orange> #action May-meimei lixiaoguang complete onos section in Functest doc
08:29:53 <morgan_orange> #topic on feature project integration
08:30:23 <morgan_orange> #info pre integration of promise done (like onos just preparation of the docker file in order to be able to run the test using run_test.sh -t promise
08:30:35 <morgan_orange> #info patch to be perged today
08:31:17 <jose_lausuch> sounds good
08:31:20 <morgan_orange> promise tests to be done on fuel only => patch created for CI to detect the list of runnable test
08:31:23 <morgan_orange> s
08:31:45 <morgan_orange> #link https://gerrit.opnfv.org/gerrit/#/c/5995/
08:32:07 <morgan_orange> #info patch corresponds to functions to test which test is runnable (installer/controller/scenario/...)
08:32:13 <morgan_orange> to be tested...
08:32:46 <morgan_orange> #info integration of doctor had been initiated, I may start from the last version (including promise) based on r-mibu patch to prepare doctor
08:33:16 <morgan_orange> bryan just left the room, do not know if something is ready for copper
08:33:30 <morgan_orange> #action morgan_orange prepare doctor env (like for onos and promise)
08:33:45 <morgan_orange> any comments/remearks on the integration of the feature tests....
08:33:58 <morgan_orange> still lots of open questions, we will learn by doing and the timing is short...
08:34:08 <jose_lausuch> for doctor, we are still waiting to merge ryota's patch, right?
08:34:23 <jose_lausuch> https://gerrit.opnfv.org/gerrit/#/c/5753/
08:35:04 <morgan_orange> as we made evolution for onos and promise, it is in merge conflict.... usually I go quicker abandon/redo than managing merge conflict...
08:35:17 <morgan_orange> but that is because I am bad in git
08:35:41 <morgan_orange> and there was an error (unexpected creation of a file)
08:35:46 <jose_lausuch> yes, he has to fix it
08:35:50 <jose_lausuch> ok
08:35:54 <jose_lausuch> waiting for that
08:36:00 <morgan_orange> r-mibu: what do you prefer?
08:36:41 <morgan_orange> #topic troubleshooting release B (internal test cases)
08:37:23 <r-mibu> morgan_orange: i will fix the patch asap :)
08:37:38 <morgan_orange> r-mibu: ok thanks
08:37:58 <morgan_orange> #info vPing ok on joid (84s) and fuel (26s), to be checked in CI on other installers (we know it used to work manually but we do consider only tests from CI to give Functest green or red light)
08:38:26 <morgan_orange> main difference joid is with odl, fuel without
08:38:40 <morgan_orange> could be interesting to run functest on joid/neutron
08:39:08 <morgan_orange> David_Orange: could we test it in this config on Orange lab for instance (so CI is not disturbed...)
08:39:24 <jose_lausuch> you mean without controller?
08:39:47 <morgan_orange> yes to see if the time difference is linked to the controller...we should also soon be able to test fuel/odl
08:39:53 <David_Orange> morgan_orange: yes we can. With nosdn parameter
08:40:09 <jose_lausuch> ok
08:40:10 <morgan_orange> but as far as I remember, we did it for Arno and I do not remember suche sifference
08:40:17 <David_Orange> morgan_orange: when the ci on intel pod5 and 6 will be stable maybe
08:40:27 <morgan_orange> ok
08:40:53 <morgan_orange> #info tempest viktor_nokia, you mentioned, only a patch to submit, I think it is properly running on fuel/nosdn (188 OK/0 failed)
08:41:21 <morgan_orange> with joid/odl more errors mainly admin role (a priori fixed) and errors linked to ext network (also a patch in progress)
08:41:32 <morgan_orange> for the other installers we wait for a run on CI
08:42:11 <morgan_orange> I think on manual run there were also more errors on compass (probably due to the configuration / ext network,...)
08:42:32 <jose_lausuch> ok
08:42:35 <morgan_orange> but we may also agree that we troubleshoot only on CI results (so a target SUT)
08:42:35 <viktor_nokia> It is not critical patch, just improving cleanup (not mandatory for B release)
08:42:43 <jose_lausuch> lot of errors were due to lack of ext network
08:42:53 <morgan_orange> it should be fixed
08:43:01 <morgan_orange> it was 50-50
08:43:17 <morgan_orange> ext network and no admin role as the declared role in joid was Admin and not admin
08:43:23 <morgan_orange> it is like services and service ...
08:43:29 <jose_lausuch> ah ok
08:44:02 <morgan_orange> the 4 installers are not homegenous despite our effort (jose created a table long time ago to try to harmonize....)
08:44:11 <morgan_orange> #info rally
08:44:14 <morgan_orange> is juha around?
08:44:23 <boris-42> meeting?)
08:44:25 <viktor_nokia> he is next to me
08:44:37 <viktor_nokia> but not able to participate :)
08:44:45 <viktor_nokia> directly I mean
08:44:54 <morgan_orange> I will try to report what he told me
08:45:50 <viktor_nokia> Juha said that he has one pending patch, will finish today or tomorrow
08:47:32 <morgan_orange> ok this patch shall just allow to move to our current management (run all the scenario the ones after the others) to the certif feature introduced in rally 0.1.1
08:47:44 <morgan_orange> it will be cleaner even if the results at the end doe not change
08:48:12 <boris-42> morgan_orange: btw there is already 0.2.0 release )
08:48:30 <morgan_orange> yes we are still in 0.1.2 ... you go so fast..
08:49:01 <jose_lausuch> I will update the version
08:49:10 <jose_lausuch> boris-42: any major improvements? :)
08:49:29 <boris-42> jose_lausuch: morgan_orange there was serious bug in 0.1.2
08:49:40 <boris-42> jose_lausuch: morgan_orange it didn't create well resource for tempest
08:49:55 <jose_lausuch> boris-42: ok
08:49:58 <boris-42> jose_lausuch: morgan_orange and there is serious improvement in Rally Task framework it self
08:49:59 <jose_lausuch> thanks
08:50:16 <boris-42> but it is realted to HTML reporting and scenario output
08:50:27 <boris-42> not yet used in any of sceanrios that we have
08:51:17 <morgan_orange> #action jose_lausuch move to rally version 0.2.0
08:51:24 <morgan_orange> #info vims use case
08:51:27 <boris-42> morgan_orange: it shouldn't be hard
08:51:34 <morgan_orange> Valentin_Orange: any update...
08:52:01 <Valentin_Orange> no
08:52:22 <morgan_orange> any comment on the time duration issue detected on E/// lab and maybe also in Huawei
08:52:26 <morgan_orange> lab
08:53:22 <Valentin_Orange> If I create a global timeout ..
08:53:23 <jose_lausuch> May-meimei send an email with the same issue
08:53:35 <jose_lausuch> it looks like our DNS problem in Ericsson pod2
08:54:08 <Valentin_Orange> Yes, I just answer them ..
08:54:43 <morgan_orange> yep looks like an integration issue on the lab, the vIMS test case needs access to google DNS..if not ..timeout is very long between N retries so Functest move from 2h to 7h ...wich is not acceptable for CI
08:54:50 <morgan_orange> so it may be a reco for pharos/labs
08:55:03 <morgan_orange> or shall we add a contraint on this test case?
08:55:26 <morgan_orange> 5 minutes left and 2 topics
08:55:26 <jose_lausuch> well, there are 2 options
08:55:29 <Valentin_Orange> leo_wang:  say the vims test-case work only on one environment on huawei platform
08:55:31 <jose_lausuch> req for pharos
08:55:34 <jose_lausuch> or fix it ourselves
08:55:54 <morgan_orange> #topic release C
08:56:12 <morgan_orange> #info we are all focused on release B, but we may start thinking to release C...
08:56:33 <morgan_orange> #info I think it will deal with integration of feature projects, the VNF shown during the summit by Intel,...
08:56:50 <morgan_orange> #action morgan_orange create a page for next release ...
08:57:13 <morgan_orange> and of course a better use of Rally as we still use 5% of the tool...
08:57:15 <morgan_orange> :)
08:57:27 <morgan_orange> #topic AoB
08:57:35 <jose_lausuch> do we start working on release C? :)
08:58:19 <morgan_orange> not really working, but as it will come quickly, I encourage people to start imagine what could be done beyond C....
08:58:50 <morgan_orange> but most of the team will not have time...
08:59:06 <morgan_orange> any other stuff to share for today?
08:59:37 <jose_lausuch> nop
08:59:38 <morgan_orange> if not we close the bridge, so code freeze for us this week, troubleshooting and doc ahead ...
08:59:43 <jose_lausuch> still not clear how to fix the dns problem
08:59:51 <morgan_orange> good luck
08:59:53 <jose_lausuch> this week means today?
09:01:23 <morgan_orange> I think so, looking for the page...
09:03:38 <morgan_orange> got it https://etherpad.opnfv.org/p/steps_to_brahmaputra
09:03:43 <morgan_orange> (not in wiki but in etherpad)
09:03:45 <morgan_orange> no time
09:03:49 <morgan_orange> just a date ..the 12
09:03:58 <jose_lausuch> ok
09:04:11 <morgan_orange> we will see during the release meeting - every day 14h30 UTC on #opnfv-release
09:04:13 <morgan_orange> ok
09:04:17 <morgan_orange> so I close the bridge
09:04:21 <morgan_orange> thanks for your participation
09:04:43 <jose_lausuch> ok
09:04:44 <jose_lausuch> thanks
09:05:11 <morgan_orange> and as Bowie used to sing "We can be Heroes, just for one day"
09:05:17 <morgan_orange> #endmeeting