16:05:06 <frankbrockners> #startmeeting BGS daily synch
16:05:06 <collabot> Meeting started Thu May  7 16:05:06 2015 UTC.  The chair is frankbrockners. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:05:06 <collabot> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:05:06 <collabot> The meeting name has been set to 'bgs_daily_synch'
16:05:14 <frankbrockners> #info Frank Brockners
16:05:33 <pbandzi> #info Peter Bandzi
16:05:43 <fdegir> #info Fatih Degirmenci
16:05:44 <trozet> #info Tim Rozet
16:06:42 <lmcdasm> #info daniel smith
16:08:52 <frankbrockners> anyone knows whether Jonas is joining?
16:09:34 <frankbrockners> anyway.. let's see who'll join late
16:09:41 <frankbrockners> let's do the usual round table
16:10:15 <trozet> want me to go?
16:10:21 <frankbrockners> anyone interested in starting - updates on deployments to pod1/2, testing, etc.
16:10:26 <frankbrockners> trozet: yes please
16:11:07 <trozet> #info update on deploy.sh: virtual is *almost* working, nodes are deployed as vagrant VMs, one issue with nova-compute that I'm trying to fix
16:11:37 <trozet> #info update on deploy.sh: pbandzi has tested and figured out proxy support which will be added
16:12:16 <morgan_orange> #info Morgan Richomme
16:12:16 <trozet> #info update on LFPOD2: jenkins deployed again with success.  morgan_orange re-ran part of the test suite which triggered the ODL bugs
16:12:38 <trozet> #info ODL issue not seen with only the tempest smoke test suite.  pbandzi to try ODL test suite to see if that triggers the bug
16:13:03 <trozet> #info trozet to submit install guide to gerrit today
16:13:28 <trozet> thats the end of my status
16:13:33 <frankbrockners> thanks trozet
16:13:34 <fdegir> question to trozet
16:13:42 <fdegir> when do you think Foreman scripts/files (deploy.sh, clean.sh, inventory) will fully be in OPNFV Gerrit, merged to master?
16:13:56 <fdegir> we currently take stuff from github, not from OPNFV gerrit
16:14:14 <trozet> let me check what was merged
16:14:49 <trozet> looks like we need to get those other patches merged in gerrit
16:15:14 <fdegir> so it is work in progress
16:15:15 <fdegir> thanks
16:15:24 <trozet> fdegir: I can work on that with frankbrockners and radez and send you an email tmrw when it is done
16:15:32 <trozet> fdegir: then you can switch to genesis repo
16:15:35 <trozet> is that OK?
16:15:41 <fdegir> it would be really good to do that
16:15:43 <frankbrockners> fdegir: sounds good
16:15:47 <fdegir> so we can remove those messy workarounds
16:15:53 <fdegir> that's all from me
16:16:02 <frankbrockners> we can get those merged asap - will look into this right after the meeting
16:16:45 <frankbrockners> one addition to the test update:
16:17:23 <frankbrockners> #info Discussion on the release readiness call earlier today: Focus would be to document which tests pass and which tests fail
16:17:56 <morgan_orange> #link https://wiki.opnfv.org/list_of_tempest_smoke_tests
16:18:00 <frankbrockners> #info For those tests which fail, let's understand and document the root cause (i.e. test case wrong, setup wrong, component has an issue)
16:18:47 <morgan_orange> #info on POD2 SStill 30 errors but seem linked to 5 errors at the end, some look trivial (ambiguous network => tests done on non fresh system then several network starting with the same name test-*
16:18:56 <frankbrockners> #info We have common understanding that we won't be able to fix component issues for now. So focus would be on documenting the issues (e.g. if there are issues with Neutron-ODL interaction - let's just explain what those are)
16:19:15 * ChrisPriceAB sorry I'm late
16:19:20 <frankbrockners> that was my little addition
16:19:26 <frankbrockners> Morgan please add...
16:20:22 <morgan_orange> well, I summarized in the page the list of smoke tests and if you remove the test in bold you have the green list (all tests passezd on Ericsson POD)
16:20:28 <frankbrockners> ok - cool
16:20:31 <morgan_orange> #not full match with my test on POD2
16:21:20 <trozet> #info one other thing related to BGS, but not required for Arno.  I saw some commits for the hardware adapter stuff the Fuel guys have been working on.  If we can make that common, that would be great.
16:21:24 <frankbrockners> morgan_orange: did we run the same sequence of tempest - robot - rally tests on Ericsson POD as we do on LF POD or was the sequence different?
16:21:35 <frankbrockners> Sequence seems to matter from what I understood from trozet
16:21:56 <morgan_orange> we performed tests on both
16:22:18 <frankbrockners> exact same sequence?
16:22:48 <trozet> sequence means - tempest smoke test, odl tests (pbandzi executes), and rally tests
16:22:53 <morgan_orange> #link https://wiki.opnfv.org/opnfv_functional_testing
16:23:31 <morgan_orange> see the table
16:23:31 <frankbrockners> yup - and individual sequence of tests within each of these test suites.
16:24:08 * frankbrockners thanks - looking at the table now
16:24:10 <morgan_orange> we have difference on rally and on tempest
16:24:21 <morgan_orange> maybe I shall explain, I think it was confusing
16:24:24 <frankbrockners> which might be the reason...
16:24:36 <morgan_orange> we documented 54 testcases (wait for vIMS but no feedback so far)
16:25:12 <morgan_orange> 1) Rally => rally are bench scenario running N times basic tests on Openstack module and providing KPI (time for bootin, ..)
16:25:25 <morgan_orange> that is more perfo tests on light scenario
16:25:42 <morgan_orange> 2) Tempest => functional suite for OpenStack
16:25:48 <morgan_orange> we can run it with several options
16:25:52 <morgan_orange> smoke, all, ..
16:26:05 <morgan_orange> we focus on smoke and would like to be as close as possible to the 100% OK
16:26:15 <morgan_orange> 3) ODL => Peter
16:27:05 <morgan_orange> 4) vPing => basic python script uploading an image, creating a network booting 2 VM getting IP of 1 and in cloud inti performing ping towards this IP, watching console to see if PING is OK, then cleaning everithing
16:27:05 <pbandzi> ODL I ma having issues now with connection to neutron... investigating
16:27:21 <morgan_orange> we worked on vPing today
16:27:55 <morgan_orange> manually was able to do almost everything except the cloud init , seems to have an issue reaching meta data
16:28:32 <morgan_orange> did not have the problem on our labs but we already had networks up & running not a fresh install
16:28:44 <morgan_orange> for all these test cases we have a test guide
16:29:21 <morgan_orange> that shall be corrected but is already well advanced
16:29:22 <frankbrockners> thanks for the details morgan_orange
16:29:43 <frankbrockners> let's move to POD1 deployment
16:29:55 <frankbrockners> pbandzi, ChrisPriceAB?
16:30:30 <pbandzi> #info I saw mail that Setfan is proceeding with installation
16:30:44 <pbandzi> #info we had issue with vlan0 but hav workaround now
16:31:25 <pbandzi> #info and issue wit public network - foreman set dhcp server on it disturb fuel checks
16:31:45 <pbandzi> #infe sent mail to Konstanti asking for public net separation
16:31:50 <pbandzi> that's all
16:32:16 <frankbrockners> thanks pbandzi
16:32:32 <frankbrockners> we'll need to get the public networks isolated asap
16:33:14 <frankbrockners> anything additional updates from anyone? Anything else we should discuss?
16:33:34 <fdegir> just a reminder regarding Jenkins access
16:33:53 <fdegir> all: please provide info if you think you need job triggering permissions
16:34:01 <fdegir> you have details in the mail
16:34:15 <frankbrockners> thanks fdegir - this is important
16:34:54 <frankbrockners> ok ... looks like that's it for today
16:35:13 <frankbrockners> thanks everyone! And see you tomorrow at 9am PDT
16:35:17 <frankbrockners> #endmeeting