15:59:47 <frankbrockners> #startmeeting OPNFV BGS daily release readiness synch
15:59:48 <collabot> Meeting started Thu May 28 15:59:47 2015 UTC.  The chair is frankbrockners. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:59:48 <collabot> Useful Commands: #action #agreed #help #info #idea #link #topic.
15:59:48 <collabot> The meeting name has been set to 'opnfv_bgs_daily_release_readiness_synch'
16:00:08 <frankbrockners> #info Frank Brockners
16:00:10 <pbandzi> #info Peter Bandzi
16:00:16 <morgan_orange> #info Morgan Richomme
16:00:25 <radez> #info Dan Radez
16:01:18 <jose_lausuch> #info Jose Lausuch
16:01:32 <HKirksey> #info Heather Kirksey
16:01:39 <fdegir> #info Fatih Degirmenci
16:02:30 <frankbrockners> Do we have anyone
16:02:40 <frankbrockners> From the Fuel team?
16:03:50 <frankbrockners> Let's get stated
16:03:52 <[1]JonasB> #info Jonas Bjurel
16:04:05 <[1]JonasB> I�m in a parrallel meeting
16:05:14 <frankbrockners> Lets do the usual round table with functest, PODs etc.
16:05:31 <frankbrockners> And focus on remaining issues.
16:05:52 <frankbrockners> Morgan, could you start with functest
16:05:56 <morgan_orange> #topic functest
16:06:17 <morgan_orange> #info first complete sequence run on POD2 last night: 4 testcases run
16:06:31 <frankbrockners> Great
16:06:50 <morgan_orange> #info still too many errors but at least the 4 testcases were run
16:06:52 <morgan_orange> #link https://build.opnfv.org/ci/view/functest/job/functest-opnfv-jump-2/
16:07:07 <morgan_orange> there was an hot discussion on BGS (I was attending test in //)
16:07:35 <morgan_orange> but I think that the Sytsrem Under Test is: Openstack HA + ODL
16:07:42 <morgan_orange> It makes sens to tun Tempest
16:07:55 <morgan_orange> we know that there wxill be errors due to integration ODL/O/S
16:08:22 <morgan_orange> but anyway from the functional point of view, we ask for networks, subnets, ports...
16:09:04 <morgan_orange> #info main concern is that the number of tests/failures vary from one test to another
16:09:27 <morgan_orange> #info on last 5 run never the same figures
16:09:41 <morgan_orange> #link https://wiki.opnfv.org/r1_tempest?&#pod_2
16:10:05 <morgan_orange> there are non error => config error on Rally (adding a paramet fix all the keystone v3 failure tests)
16:10:16 <morgan_orange> but main errors are dealing with network (without surprise)
16:10:31 <pbandzi> #info I'll check also again ODL tests, from console output seems that some of them fails
16:10:42 <morgan_orange> ODL suite and vPing shall be OK troubleshootin on vPing planned tomorrow on my side
16:10:44 <jose_lausuch> pbandzi: I know what the errors are
16:10:49 <jose_lausuch> relative path problems again
16:10:57 <jose_lausuch> I faced the same doing it manually on POD1
16:11:03 <morgan_orange> ok
16:11:26 <morgan_orange> for vPing we have 2 types of possible errors: sometime impossible to start the VM, sometimes timeout on the Ping
16:11:32 <pbandzi> jose_lausuch: i'll check tomorrow some of them passed so  I will see
16:11:56 <morgan_orange> newbee question...does ODL support Ipv6?
16:12:04 <lmcdasm> yes
16:12:08 <jose_lausuch> contacting the API works
16:12:12 <morgan_orange> as far as I can see all network IPv6 tests failed
16:12:35 <jose_lausuch> but then the txt test files fail because of filr-not-found or wrong path
16:12:36 <lmcdasm> i dont think that neutron is configured for IPv6 network definitions
16:12:40 <jose_lausuch> it does not even try to test
16:12:50 <lmcdasm> so while ODL will reflect what Neutron has.. if neutron cant do it right, then you have an issue
16:13:03 <lmcdasm> vxlan tunneling itself should be fine with IPv6
16:13:34 <jose_lausuch> ok
16:14:08 <morgan_orange> #info wiki has been updated but best source for errors are now jenkins console logs
16:14:23 <lmcdasm> can te r1_old_suite page be updated as well?
16:14:46 <pbandzi> lmcdasm: I'll update
16:15:44 <morgan_orange> that is all for me
16:17:19 <[1]JonasB> I will presenting in my other meeting for 30 minutes
16:19:41 <morgan_orange> frankbrockners: still there ...
16:22:00 <morgan_orange> still somebody on board?
16:22:06 <jose_lausuch> hehe
16:22:16 <jose_lausuch> haning
16:22:19 <jose_lausuch> hanging
16:23:31 <frankbrockners> Still here... Sorry got distracted
16:24:30 <HKirksey> Well we’re lacking Chris’s flights of poetic fancy today….perhaps easier to get distracted. :)
16:27:04 <jose_lausuch> so, shall we move on?
16:27:53 <frankbrockners> Yes please.
16:28:40 <frankbrockners> Any progress on POD1 or the ISO for Foreman?
16:29:53 <frankbrockners> radez are you around?
16:30:06 <jose_lausuch> POD1 is installed with O/S and to be worked for ODL
16:30:10 <jose_lausuch> but not sure what
16:30:50 <frankbrockners> Is POD1 operational with ODL now?
16:30:52 <lmcdasm> for POD1 - the testing scope was only clarified this morening
16:31:01 <lmcdasm> POD1 manualy test of ODL was done on monday
16:31:15 <lmcdasm> previous tests on POD1 dont count cause they were done without ODL turned on (that wasnt clear0
16:31:28 <lmcdasm> the script was submitted for POD1 and the steps to set it in ODL mode
16:31:41 <jose_lausuch> suggestion is to run the jobs from jenkins deploy+functest with ODL turned on
16:31:50 <lmcdasm> waiting for test results - however, the questino of what to do / how far to fix and exit criteria from you (Frank) is stil out there
16:32:35 <jose_lausuch> not "suggestion" , focus instead
16:32:57 <radez> #info foreman ISO is still in progress. I'm having trouble reproducing a new environment from it to move forward.
16:33:23 <radez> #info currently comparing an env installed from the git repo to the iso installer to try and find the delta
16:33:40 <frankbrockners> lmcdasm: it's not me to decide the exit criteria. What we jointly agreed is to fix those issues which we can control in the time being and document the failures which we cannot fix. This assumes that we have an understanding why things fail.
16:34:04 <frankbrockners> The real worry is that tests are not fully repeatable
16:37:20 <lmcdasm> pounding in the update for POD1 -sorry i didnt do it before
16:37:24 <lmcdasm> #info  POD1 manualy test of ODL was done on monday
16:37:34 <lmcdasm> #info previous tests on POD1 dont count cause they were done without ODL turned on (that wasnt clear)
16:37:51 <lmcdasm> #info the script was submitted for POD1 and the steps to set it in ODL mode provided
16:38:34 <lmcdasm> #info tests to be run, unsure of the focus to test / resolve - since we know the configurations betwen the two systems are different, so you can expect deltas in results - and you wont know if they are cause of configuration or component errors
16:38:38 <lmcdasm> thats it for me
16:39:55 <frankbrockners> Thanks Daniel.
16:40:38 <frankbrockners> Let's create a quick laundry list
16:40:51 <frankbrockners> What things are still pending?
16:41:50 <frankbrockners> From what I see it it's (a) POD1 deploy with ODL and automatic testing
16:42:03 <frankbrockners> (b) ISO for Foreman
16:43:05 <frankbrockners> (c) Documentation of failing tests (why they fail)
16:43:25 <frankbrockners> (d) finalize vPing
16:43:26 <morgan_orange> + min vPing + ODL OK on both PODs
16:43:55 <frankbrockners> (e) finalize docs
16:44:02 <frankbrockners> Thanks morgan_orange
16:44:10 <frankbrockners> What else are we missing?
16:44:25 * jose_lausuch is troubleshooting POD2 since spawning VMs doesnt work
16:44:42 <jose_lausuch> nothing from my side
16:47:06 <morgan_orange> frankbrockners: fine for me...
16:47:14 <morgan_orange> trevor_intel: functest pres of this afternoon uploaded on the wiki: https://wiki.opnfv.org/?ns=%3A&image=%3Aopnfv-_functest.pdf&do=media&tab_files=files&tab_details=view
16:47:28 <morgan_orange> I have to jump out
16:47:42 <morgan_orange> see you tomorrow and see the results of troubleshooting of vPing and ODL suite
16:47:50 <morgan_orange> CI is working for us every 6 hours
16:47:59 <frankbrockners> Ok
16:48:16 <morgan_orange> we will also see the reproducibility of the errors
16:48:32 <frankbrockners> Let's go through the above laundry list tomorrow
16:48:38 <jose_lausuch> any opnestack guru around?
16:49:00 <jose_lausuch> familiar with this error: {"message": "No valid host was found. ?
16:49:05 <jose_lausuch> sorry for using the meeting for this
16:49:14 <jose_lausuch> I'll ask later
16:49:15 <frankbrockners> morgan_orange: agree, if things are reproducible that would be great
16:49:46 <pbandzi> jose_lausuch: when trying to create VM?
16:49:53 <jose_lausuch> yes
16:50:07 <jose_lausuch> ping me in private so we dont bother the chat
16:50:44 <pbandzi> jose_lausuch: that might indicate problem with compute node... I would check /var/log/nova.log maybe
16:50:56 <jose_lausuch> I did
16:51:00 <jose_lausuch> check private
16:54:56 <frankbrockners> Ok
16:55:17 <frankbrockners> Looks like we're done
16:56:14 <frankbrockners> So let's go through the above list tomorrow again
16:56:21 <frankbrockners> Thanks everyone
16:56:31 <frankbrockners> #endmeeting