15:02:54 <dneary> #startmeeting OpenDaylight-OPNFV interlock meeting
15:02:54 <odl_meetbot> Meeting started Thu Feb 12 15:02:54 2015 UTC.  The chair is dneary. Information about MeetBot at http://ci.openstack.org/meetbot.html.
15:02:54 <odl_meetbot> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:02:54 <odl_meetbot> The meeting name has been set to 'opendaylight_opnfv_interlock_meeting'
15:03:06 * dfarrell07 is here
15:03:11 * tbachman is watching
15:03:17 <tbachman> is it IRC only?
15:03:17 * regXboi hears crickets
15:03:17 * dfarrell07 can help with minutes if needed
15:03:22 * flaviof here
15:03:26 <dneary> #chair dfarrell07 tbachman flaviof
15:03:26 <odl_meetbot> Current chairs: dfarrell07 dneary flaviof tbachman
15:03:27 <dfarrell07> no, there's a call
15:03:29 <regXboi> tbachman: no
15:03:41 <dneary> regXboi, tbachman: No, phone too
15:03:45 <tbachman> dneary: ah
15:03:46 <tbachman> thx
15:03:51 <tbachman> I think that was in the email
15:03:52 <regXboi> dneary: I know, I'm on the call (on mute)
15:03:57 <dfarrell07> US Toll-Free Dial-In Number: 800 451 8679; Intercall conference code: 915 507 3783#
15:03:59 * tbachman goes to consult ovsdb mail archives
15:04:04 <tbachman> dfarrell07: thx!
15:04:50 * tbachman has another meeting he probably will have to run to — but will dial in for now
15:05:01 <flaviof> tbachman: https://lists.opendaylight.org/pipermail/ovsdb-dev/2015-February/001143.html
15:05:07 <dfarrell07> #info goal of meeting is to introduce OPNFV and ODL communities
15:05:12 <tbachman> flaviof: thx!
15:05:26 <dfarrell07> #info also going to talk about an issue we ran into recently, details of how it was fixed
15:06:00 <edwarnicke> #info edwarnicke
15:06:10 <dfarrell07> #info dfarrell07
15:06:10 <regXboi> #info regXboi (Ryan Moats)
15:06:13 <dneary> #info Daniel Smith, Ericsson
15:06:16 <tbachman> #info tbachman
15:06:21 <Prem> #info Prem, Ericsson
15:06:28 <shague> #info shague Red hat, odl ovsdb project
15:06:29 <ebrjohn> #info ebrjohn Brady Johnson, Ericsson, SFC Project Lead
15:06:42 <dneary> #info Uli Kleber, Huawei
15:06:51 <dneary> #info Brady Johnson, Ericsson
15:06:55 <frankbrockners> #info Frank Brockners, Cisco
15:06:57 * tbachman hears binary exponential backoffs
15:07:06 <flaviof> #info flaviof Red Hat, Boston
15:07:35 <trozet> #info trozet, Red Hat
15:07:56 * dfarrell07 is not from OVSDB, btw dneary :)
15:08:08 <dneary> dfarrell07, My mistake
15:08:11 <dfarrell07> np :)
15:09:10 <dneary> #info From OPNFV BGS: trozet, frankbrockners
15:09:10 <dfarrell07> #info trozet discovered most recent issue, gives overview
15:09:30 <dneary> #info From OVSDB project in OpenDaylight: flaviof, shague
15:10:11 <dfarrell07> #info tl;dr is that there was a timing issue involving feature loading in ODL, trozet made changes in quickstack Puppet mod to work around it
15:10:42 <flaviof> #info trozet changed quickstack to wait for odl to have bundled features loaded before having neutron connect to odl
15:11:47 <dfarrell07> #info there's another issue that ODL is sending 200 OKs when it shouldn't (details in open bugs)
15:13:11 <regXboi> #info I actually argue that "another issue" is not an issue - it is things working by design
15:13:42 <flaviof> #info trozet waits for the net-virt feature to show bundle active in order to work around issue #1
15:13:51 * regXboi asks for chair privs
15:14:07 <dfarrell07> #chair regXboi
15:14:07 <odl_meetbot> Current chairs: dfarrell07 dneary flaviof regXboi tbachman
15:14:08 <tbachman> #chair regXboi
15:14:08 <odl_meetbot> Current chairs: dfarrell07 dneary flaviof regXboi tbachman
15:14:10 <tbachman> lol
15:14:11 * regXboi can't believe he just did that
15:14:13 <dneary> #info From the ODL Controller: Ryan Moats, Ed Warnicke
15:14:29 <flaviof> #chair regXboi
15:14:29 <odl_meetbot> Current chairs: dfarrell07 dneary flaviof regXboi tbachman
15:14:39 <dfarrell07> #info From ODL Integration and ODL Puppet mod, Daniel Farrell (dfarrell07)
15:14:44 <regXboi> #info correction: Ed Warnicke is from Controller, Ryan Moats is a general gadfly
15:14:49 <dneary> Thanks dfarrell07
15:14:54 <dfarrell07> np dneary :)
15:15:16 <dneary> regXboi, Who were the 3 committers you mentioned?
15:15:35 <dfarrell07> #info discussion about making sure we're waiting on the right "things are loaded and gtg"-type event from ODL
15:16:00 <regXboi> dneary: the proposed committers are myself, ed, and flavio
15:16:15 <dneary> regXboi, So you were in te right place :-)
15:16:57 <regXboi> dneary: yes - I *had* declined because I had a dental appointment for my daughter in this time slot - but she's sick, so we postponed and .... so I'm here
15:17:25 <dneary> Sick + dental appointment would be a nightmare layering
15:17:48 <regXboi> dneary: but I've been lurking behind the scenes on these issues since they came to light
15:18:54 <dfarrell07> #info question on ODL start order from dneary
15:19:18 <dfarrell07> ^^and going to ask regXboi to note his answer down if he wants
15:19:49 <dneary> #info 1. Neutron -> ODL ML2 agent (OpenStack side)
15:20:10 <flaviof> #info neutron in odl uses whiteboard pattern so an other odl bundle -- ovsdb being one of them -- can get neutron events
15:20:11 <dneary> #info 2. ML2 agent -> Neutron service (part of Controller)
15:20:57 <dneary> #info 3. Controller SB services get "offered" Neutron request, if all say "yes" when offered, the controller responds "OK 200"
15:21:22 <dneary> #info 4. SB services then commit to handling the request SB asynchronously
15:22:03 <dneary> #info flaviof: "At this point, there are no SB modules, no modules say no, therefore request is accepted, but can't be handled"
15:23:16 <frankbrockners> question: is the issue also showing up if there are connected OVSs? I.e. OVSDB SB plugin should respond?
15:30:38 <trozet> guys my scenario is ODL is fully up, 2 OVS nodes ( one on compute, one one network),  if you remove br-int from network, ODL will still add vxlan tunnels/flows to br-int on compute and act like its creating networks, even though the other side of the tunnel is missing
15:30:57 <trozet> if you remove br-int from compute side nova will error out when you try to bring up the instance, but this doesnt happen on the other side
15:31:26 <trozet> if ODL gets a request from Neutron, and it is missing its br, it should error back on the REST call imo
15:31:49 <trozet> adding ports relies on 2 pieces of info as Sam said, so that case I'm fine with, but ovs bridge should be there
15:32:23 <shague> there is an original problem on bootstrapping whcih then causes further issues
15:35:25 <dneary> frankbrockners, It is for me, but I think I have dome something wrong
15:35:48 <dneary> #info trozet: my scenario is ODL is fully up, 2 OVS nodes ( one on compute, one one network),  if you remove br-int from network, ODL will still add vxlan tunnels/flows to br-int on compute and act like its creating networks, even though the other side of the tunnel is missing
15:36:03 <dneary> #info trozet: if you remove br-int from compute side nova will error out when you try to bring up the instance, but this doesnt happen on the other side
15:36:18 <dneary> #info trozet: if ODL gets a request from Neutron, and it is missing its br, it should error back on the REST call imo
15:36:18 <dneary> adding ports relies on 2 pieces of info as Sam said, so that case I'm fine with, but ovs bridge should be there
15:38:41 <dfarrell07> *generic reminder to mute if you're not talking*
15:40:34 <dneary> #info Discussion of when it's appropriate to be synchronous in ML2 responses and whether it's possible to do async notifications via (for example) OpenStack's MQ
15:41:17 <dneary> Generic question: I can see ML2 requests coming in with netdump (HTTP request on port 8080)
15:41:29 <dneary> How can I see the hand-off from the ML2 plug-in to the controller?
15:41:49 <dneary> And a trace of the interactions between the controller and the SB modules?
15:42:27 <shague> dneary: tcpdump/wireshark shows neutron coming to odl. From there you will need to enable the neutron and ovsdb logging
15:42:27 <dneary> And then, how can I see the actual actions of the SB modules (say, OVSDB)
15:42:43 <dneary> shague, Plus controller?
15:43:00 <shague> plus controller?
15:43:24 <dneary> I'm assuming you need trace on the controller too to see the "whiteboard pattern" calls
15:44:01 <shague> From there you will need to enable the neutron and ovsdb logging on odl
15:45:10 <dneary> shague, Thanks
15:45:45 <flaviof> #info regXboi proposes the implementation that uses config and operational states
15:51:58 <dneary> #info Some action items for OpenDaylight: If there are no SB modules available, return a fail on any NB requests
15:53:48 <dneary> #info Some debate about whether to have a fire/notify model or a 2 pass approach ("request then check") but general opinion expressed by edwarnicke that there's a mismatch between Neutron's synchronous request model & a back-end async model
16:00:07 <dfarrell07> #info question from OPNFV: "we need a mostly-working deployment in the next few weeks. possible at all?" answer from shague: "trozet has a working setup today. it doesn't address all of the details described on today's call, but it does the basics"
16:00:45 * dfarrell07 found this community connection very helpful
16:01:20 <mlemay> +1 to ryan's comments :)
16:01:56 <frankbrockners> +1 for continued discussions
16:02:03 <dfarrell07> Can just add a weekly "OVSDB and ODL integration call"
16:02:14 <dfarrell07> s/OVSDB/OPNFV
16:02:16 <dfarrell07> my bad
16:02:36 <frankbrockners> probably plus the new neutron project
16:02:36 <dfarrell07> ^^yes!
16:04:55 <dfarrell07> thanks for the great work, @everyone :)
16:05:08 <flaviof> #endmeeting