14:31:13 <debrascott> #startmeeting Brahmaputra Daily Standup
14:31:13 <collabot`> Meeting started Tue Mar  1 14:31:13 2016 UTC.  The chair is debrascott. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:31:13 <collabot`> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:31:13 <collabot`> The meeting name has been set to 'brahmaputra_daily_standup'
14:31:26 <debrascott> #link agenda https://wiki.opnfv.org/releases/brahmaputra/minutes
14:31:26 <ChrisPriceAB> #info Chris Price
14:31:32 <debrascott> #topic roll call
14:32:03 <ChrisPriceAB> #info Chris Price
14:32:48 <radez> #info  Dan Radez
14:33:17 <fdegir> #info Fatih Degirmenci
14:33:24 <narindergupta> #info Narinder Gupta
14:33:54 <debrascott> were you all able to meet yesterday? sorry for my connection issues
14:34:20 <ChrisPriceAB> nothing yesterday
14:34:24 <debrascott> finally got internet at the new house late yesterday
14:34:49 <debrascott> OK, I don’t see that we have anyone from functest or yardstick on?
14:34:59 <bin_> #info Bin Hu
14:35:15 <debrascott> I looked at the links that morgan_orange had been using for status but they now point to C-release
14:36:06 <debrascott> Since we have installer folks here, let’s start with that
14:36:18 <debrascott> #topic Installers status
14:36:41 <debrascott> radez: how are Apex scenarios for 2nd B release going?
14:36:53 <ChrisPriceAB> can we list the scenario's we are working on and report on those daily.  Thatw ould seem to be useful in helping establish daily plans for brahmaputra.2.0
14:37:27 <debrascott> ChrisPriceAB: that is why I wanted to see the info from morgan_orange’s wiki pages
14:37:35 <radez> debrascott: moving along, we're working through getting more test senarios setup to validate the SDNVPN
14:37:43 <radez> and continuing to investigate getting contrail to work
14:37:45 <joekidder> #info Joe Kidder
14:37:57 <radez> we have a root cause for ONOS tests failing, it's envoronmental to LF
14:38:06 <ashyoung> #info Ashlee Young
14:38:36 <radez> not sure what the solution is yet but we've identified that it's a Firewall configu that dropping the traffic
14:39:27 <debrascott> radez: do you know which scenarios you are troubleshooting?
14:39:42 <debrascott> I’m lost without morgan’s list
14:39:47 <radez> I just indicated them
14:39:53 <radez> sdnvpn, onos, dev on contrail
14:40:21 <debrascott> thanks!
14:41:04 <narindergupta> debrascott: JOID is investigating onos and opencontrail
14:41:34 <debrascott> radez: except contrail these all deploy but have test issues, correct?
14:42:12 <radez> corrrect, SDNVPN still needs to be merged to stable, but it has been merged to master
14:42:21 <radez> and deploys
14:42:32 <debrascott> ok good, thank you
14:42:59 <debrascott> narindergupta: same for JOID right? ONOS deploys but has test issues, contrail working on deployment?
14:43:13 <narindergupta> debrascott: thats correct
14:43:57 <debrascott> radez: is Aric helping you with the LF environment?
14:44:33 <radez> we did troubleshooting together but it's a bit out of my hands now. there needs to be justification to change the LF FW
14:44:49 <radez> or we need to put in an extra piece of hardware between the FW and the PODs
14:45:03 <trozet> radez: is this regards to ONOS?
14:45:06 <radez> I'll continue to keep track with him
14:45:09 <radez> trozet: yup
14:45:10 <ChrisPriceAB> radez that sounds wierd.  Are you not able to get the SW onto the jump server?
14:45:22 <radez> SW?
14:45:22 <trozet> my advice would be to fix the ONOS bug rather than change LF network arch
14:46:25 <debrascott> trozet: agree if it is indeed a bug and not hw setup issue
14:46:26 <radez> ChrisPriceAB: it doesn't have anything to do with the deployment... the traffic leaves our pod and and dropped by the LF firewall
14:46:40 <ChrisPriceAB> software, I would asume once the SW is on the JH the FW should not cause issues.
14:46:44 <radez> so either ONOS needs to arp and send traffic directly back to the jumphost
14:46:54 <radez> or the network env needs to allow the traffic to flow
14:47:10 <ChrisPriceAB> oh ok.  yeah it would seem best to solve that in the platform.
14:47:36 <trozet> ChrisPriceAB: the issue is ONOS tries to send any packet to the LF gateway, even if it is in it's own subnet, rather than arping for the internal address and sending it to that device.  A lot of routers restrict that type of behavior since it can be an attack or cause unnecessary routing traffic
14:47:51 <radez> and by traffic we mean, local subnet traffic through the gateway
14:48:17 <ChrisPriceAB> yep, got it thanks guys.  :D
14:49:15 <debrascott> oh, ok, so it should be fixed in ONOS-
14:49:28 <debrascott> ashyoung: I assume you are on it?
14:49:33 <radez> Bob had indicated this was a design decision, and they intended to work this way... just to put that on the table
14:50:33 <ChrisPriceAB> err. OK.  Well it can be a documented limitation if it will not be solved in B.
14:50:47 <debrascott> hmm, seem counter-intuitiive
14:51:05 <debrascott> but not enough information on the tradeofffs
14:51:40 <debrascott> ChrisPriceAB: agree, if there is no better way to deal with it
14:52:26 <debrascott> narindergupta: is your issue with ONOS the same?
14:52:27 <ashyoung> debrascott: I don't think we're redsigning how ONOS works
14:52:59 <narindergupta> debrascott: we are working on getting the gateway mac configured
14:53:23 <narindergupta> currently charm changes are required and this issue may come back later
14:53:26 <debrascott> ashyoung: agreed, but influence, influence, influence  ;)
14:54:07 <ashyoung> debrascott: In the past, the only way to influence is fork and patch-- then hope they take the patch.
14:55:27 <debrascott> ashyoung: no core developers on your team?
14:55:28 <ashyoung> debrascott: Will see what we can do about this.
14:55:39 <bin_> ashyoung: what is the rational of ONOS to be designed in the way of sending all packets to G/W router, even if the packets is on its own subnet?
14:55:44 <ashyoung> debrascott: Doesn't matter. It all has to be approved by one person.
14:56:17 <ashyoung> debrascott: Guys, I am NOT a part of ONOS. I am the PTL for ONOSFW and ONOS is just 1 component
14:56:38 <debrascott> ashyoung: beer, brownies, equipment….;)
14:57:06 <debrascott> ashyoung: thats what I meant by core developers…sounds like you need someone on the ONOS team
14:57:25 <ashyoung> bin_: You're an equal member-- in fact, moreso. They will do anything for AT&T :)
14:58:06 <ashyoung> debrascott: Please, please, please stop trying to suggest that. We have at least 10 core developers. It's not our call.
14:58:47 <ashyoung> debrascott: Again, we can fork and patch. It's not up to us if they don't merge our patches. But we can certainly do a patched build. But this is out of scope for B release
14:58:58 <ashyoung> debrascott: Sounds like we're adding a feature change
14:59:09 <ashyoung> debrascott: Is this what we're asking for?
14:59:55 <debrascott> Yes, certainly out of scope for B but should be fixed long term if it is not following industry practice
15:00:46 <ashyoung> debrascott: Agreed
15:00:46 <morgan_orange> debrascott: https://wiki.opnfv.org/functest_release_2
15:01:01 <debrascott> morgan_orange: thank you!
15:02:07 <debrascott> morgan_orange: do you have any new status you want to report for functtest
15:02:13 <debrascott> ?
15:02:25 <trevor_intel1> #info Trevor Cooper for Brian Skerry
15:03:08 <debrascott> OK- lets transition to TSC call
15:03:17 <debrascott> thank you for joining
15:03:22 <debrascott> #endmeeting