16:00:02 <frankbrockners> #startmeeting FDS weekly Thursday synch
16:00:02 <collabot`> Meeting started Thu Jun  2 16:00:02 2016 UTC.  The chair is frankbrockners. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:02 <collabot`> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:00:02 <collabot`> The meeting name has been set to 'fds_weekly_thursday_synch'
16:00:13 <frankbrockners> #topic roll call
16:00:22 <frankbrockners> #info Frank Brockners
16:00:35 <jlinkes> #info Juraj Linkes
16:00:38 <frankbrockners> Folks, could you #info in if you're around
16:00:38 <mmarsale> #info Maros Marsalek
16:01:23 <frankbrockners> anyone else around?
16:01:38 <fpan> #info Feng Pan
16:01:45 <ashyoung2> # info  Ashlee Young
16:01:50 <martin_sunal> #info Martin Sunal
16:03:14 <frankbrockners> as agenda for today I just had (1) brief 1 line status on key projects (GBP, HC, Test, Install) (2) discuss status of current integration between components (3) Open questions
16:03:24 <frankbrockners> anything else to cover?
16:03:30 <alagalah> Thats a lot :)
16:03:37 <frankbrockners> it is
16:03:48 <frankbrockners> so let's get going
16:04:01 <alagalah> FYI Martin, Maros and some of the GBPSFC guys just did a TOI on the env
16:04:16 <frankbrockners> #topic 1-line status on GBP, HC, Test, Install
16:04:27 <frankbrockners> Who can speak for GBP?
16:04:31 <martin_sunal> I can
16:04:43 <frankbrockners> martin_sunal go ahead
16:04:58 <michal-cmarada> #info Michal Cmarada
16:05:11 <martin_sunal> we are testing code for creating of vhost user interface. this part is expected to be done this week
16:05:21 <frankbrockners> let's info that
16:05:35 <frankbrockners> #info martin_sunal reports: we are testing code for creating of vhost user interface. this part is expected to be done this week
16:05:57 <martin_sunal> #info we want to start with VBD integration from next week
16:06:25 <frankbrockners> martin_sunal - is GBP code complete now?
16:06:34 <frankbrockners> (for CLUS)
16:07:01 <martin_sunal> frankbrockners: no, it will be complete after VBD integration
16:07:08 <frankbrockners> ok - thanks
16:07:14 <frankbrockners> let's touch on VBD next
16:07:25 <frankbrockners> tylevine - could you give us a brief status?
16:07:42 <tylevine> Sure
16:08:35 <tylevine> Andrew and I have validating our VBD implementation, but so far we have seen the topology manager working correctly, need to test VXLAN tunnel creation and management today
16:08:57 <frankbrockners> #info per tylevine: Andrew and I have validating our VBD implementation, but so far we have seen the topology manager working correctly, need to test VXLAN tunnel creation and management today
16:09:13 <frankbrockners> alagalah - dare to comment on your test efforts?
16:09:20 <alagalah> frankbrockners: Sure....
16:10:04 <alagalah> frankbrockners: Last week had AFPACKET + VXLAN + BD working with HC, so checked that box in lieu of hearing anything from VBD.... I had VHOST connecting but couldn't get the iface recognised in the guest but ... thats an implementation issue w QEMU
16:10:32 <alagalah> frankbrockners: This morning did a handover to Maros/Martin ....
16:10:36 <frankbrockners> #info per alagalah: Last week had AFPACKET + VXLAN + BD working with HC, so checked that box in lieu of hearing anything from VBD.... I had VHOST connecting but couldn't get the iface recognised in the guest but ... thats an implementation issue w QEMU
16:10:43 <frankbrockners> thanks alagalah
16:10:47 <alagalah> frankbrockners: That was recorded and I have
16:10:49 <alagalah> Wait
16:10:54 <alagalah> let me info what I was about to say
16:11:59 <frankbrockners> in parallel --- tylevine - are you still testing on AWS - or did you start using the CLUS demo servers that Viliam setup?
16:12:04 <alagalah> #info alagalah tested VLAN -> bridge domain between two bare metal boxes at home last night, had ping between netns working ... couldn't test HC as Linux Foundation were doing maint. taking nexus offline .... but given I had it working, I will finish that up in "spare time" as I have to move onto other things... I'm here as a resource to Martin/Maros and their team for things integrating with VPP and testing
16:12:40 <tylevine> Andrew reached out to Viliam to get access to his test bed, but I have not heard back from him yet
16:12:57 <tylevine> So until further notice we are still testing in AWS
16:13:01 <alagalah> #link https://drive.google.com/open?id=1S6GHSDsuyLk_30EEPDfpvtofdjNFjkN9WUarZGO_1s0 alagalah's hand off to MArtin/Maros - will evolve
16:13:18 <frankbrockners> jlinkes - could you touch base with tylevine?
16:13:25 <frankbrockners> on lab access
16:13:35 <frankbrockners> thanks alagalah
16:13:46 <frankbrockners> let's move to HC status
16:13:54 <frankbrockners> mmarsale - quick update?
16:13:54 <alagalah> frankbrockners: Can I ask a Q first ?
16:14:02 <frankbrockners> please alagalah
16:14:02 <mmarsale> frankbrockners: sure
16:14:10 <jlinkes> tylevine: here's our page for the CLUS testbed https://wiki.cisco.com/display/OPNFV/Cisco+Live+Demo+Architecture
16:14:15 <alagalah> mmarsale: hold on 1 man, before we move onto HC
16:14:21 <mmarsale> alagalah: waiting
16:14:37 <alagalah> tylevine: Did you see Martin needs VBD things by Monday
16:14:59 <tylevine> jlinkes: thanks
16:15:11 <alagalah> tylevine: ie I believe he will need to see at a min. your postman collection for doing BD+VXLAN so he can mock APIs
16:15:35 <alagalah> tylevine: And a accessable testbed as per jlinkes is providing is a good place to point too
16:15:40 <tylevine> alagalah: Yes I have set Friday as our target for finishing testing, I should have a postman collection to martin by then
16:15:54 <jlinkes> tylevine: there were some connection issues with the testbed, David, the admin, is working on it, so the testbed might not be usable atm
16:15:59 <alagalah> tylevine: PERFECT! Just wanted to make sure you were aware of the critical path... thanks
16:16:09 <alagalah> frankbrockners: Thanks dude, I'm done
16:16:19 <mmarsale> here comes HC
16:16:20 <mmarsale> #info HC working on 1. VLAN model upgrade 2. L2 ACLs basic support 3. NETCONF(RESTCONF) notifications from VPP
16:16:20 <mmarsale> #info HC items in progress ETA: early next week
16:16:20 <mmarsale> #info HC will provide any support to GBP/VBD in case there are any issues with HC features already available required in FDS
16:16:23 <tylevine> alagalah: I will try to make as must use of the testbed as possible, with falling back to AWS if things are broken
16:16:33 <frankbrockners> thanks guys - tylevine - please send  details to the alias once you have the postman config avail
16:16:53 <frankbrockners> thanks mmarsale - that was quick :-)
16:16:55 <tylevine> frankbrockners: no problem
16:17:08 <alagalah> mmarsale: re: VLAN model upgrade
16:17:26 <alagalah> mmarsale: Does the discussion this morning re: POP on RX defining iface TX action change any other what you are doing >?
16:17:37 <alagalah> *any of what...
16:17:51 <frankbrockners> ah - before we move on...
16:17:55 <mmarsale> alagalah: not really ... you might just have to update the requests in your ptivate collection slightly
16:18:05 <mmarsale> alagalah: but behavior will stay the same
16:18:13 * alagalah waits for frankbrockners
16:18:15 <frankbrockners> on VBD - Woj proposed this "hack" to VBD to get external connectivity
16:18:15 <mmarsale> alagalah: we will give heads up with examples
16:18:44 <alagalah> frankbrockners: Where is Wojciek ?
16:18:47 <frankbrockners> tylevine - did you synch with Woj on that?
16:19:08 <frankbrockners> Woj had a conflict (family afaik/) - couldn't make it
16:19:14 <tylevine> frankbrockners: No we have not synched on that yet
16:19:41 <frankbrockners> could you please do so - Woj is usually avail later today (evening hrs in europe)
16:19:53 <alagalah> frankbrockners: my understanding is that "external access" is normally provided by a "provider network" in Neutron, corect ?
16:20:29 <alagalah> yamahata: Is that correct?
16:20:32 * alagalah waves yamahata
16:20:35 <frankbrockners> alagalah - Woj proposed a hack for CLUS - just looking for the most simple solution for now before we have proper L3
16:20:44 <alagalah> frankbrockners: nothing to do with L3
16:20:54 <alagalah> frankbrockners: Its important because...
16:20:56 <frankbrockners> in a way it does
16:21:14 <alagalah> frankbrockners: well I mean, I'm not shoehorning L3 into this discussion :)
16:21:29 <alagalah> frankbrockners: What I am trying to say is "provider network" in the docs / diags that Martin drew ...
16:21:42 <alagalah> frankbrockners: Use VLAN tagging (not VXLAN) on the wire... and VBD does not support that today....
16:22:01 <alagalah> frankbrockners: So we were under the impression that VBD = tenant network and GBP talks to HC direct for vhost and VLAN ....
16:22:07 <alagalah> frankbrockners: ie provider.
16:22:12 <martin_sunal> alagalah: it will work only in case if nova instances will have interfaces in both netwroks (tenant and provider)
16:22:16 <frankbrockners> alagalah - yes - this is the issue that woj and tyler need to synch up on
16:22:45 <alagalah> frankbrockners: So if VBD is going to do some provider stuff.... then we need to make sure GBP doesn't have too much contrived logic to decide when to use VBD vs HC is all
16:22:50 <alagalah> frankbrockners: martin_sunal cool.
16:22:52 <frankbrockners> could we take this up over email - and use Woj's email as reference?
16:22:57 <alagalah> frankbrockners: ack
16:23:29 <frankbrockners> let's quickly get a status on install and test
16:24:01 <jlinkes> #info per jlinkes: the l2 tenant/provider testcases will be implemented by fd.io people - https://jira.fd.io/browse/CSIT-64
16:24:07 <frankbrockners> jlinkes - could you speak for viliam? Viliam's made available a "raw" server setup with VPP included (though not configured)
16:24:34 <jlinkes> #info per vluc: we should have a deployment with vpp installed
16:24:46 <frankbrockners> please see the email that Viliam
16:25:00 <frankbrockners> sent. This should allow us to test things on real hardware
16:25:10 <frankbrockners> and we can use the system to incrementally wire things up
16:25:29 <frankbrockners> ie. for VBD / HC / VPP testing - GBP / HC / VPP testing
16:25:34 <jlinkes> #info per jlinkes: OVS deployment was redeployed to Mitaka, but there were failures due to connectivity issues
16:25:39 <frankbrockners> Neutron NB / GBP testing etc.
16:25:49 <jlinkes> #info per jlinkes: contacted David who's fixing it
16:25:58 <frankbrockners> thanks jlinkes
16:26:16 <frankbrockners> so status on the lab right now https://wiki.cisco.com/display/OPNFV/Cisco+Live+Demo+Architecture
16:26:53 <frankbrockners> two setups: 1st setup ("VPP") is Ostack liberty, ODL Be
16:27:22 <frankbrockners> 2nd setup is planned to be with Mitaka and vanilla OVS
16:27:43 <frankbrockners> we'll move the VPP setup to Mitaka as well as soon as we have the install issue on the OVS-setup resolved.
16:28:07 <frankbrockners> we're getting close to the bottom of the hr
16:28:17 <frankbrockners> any other issues that we need to bring up here?
16:29:10 * frankbrockners wonders whether we're done...
16:29:28 <jlinkes> nothing from me
16:29:37 * frankbrockners thinks we're done
16:29:47 <frankbrockners> thanks everyone - I'll post the log to the wiki
16:29:55 <frankbrockners> #endmeeting