15:04:12 <raphael> #startmeeting nic_weekly 15:04:12 <odl_meetbot> Meeting started Fri May 1 15:04:12 2015 UTC. The chair is raphael. Information about MeetBot at http://ci.openstack.org/meetbot.html. 15:04:12 <odl_meetbot> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:04:12 <odl_meetbot> The meeting name has been set to 'nic_weekly' 15:05:26 <dlenrow> colindixon: master of meeting tools 15:06:51 <colindixon> #topic agenda bashing 15:07:01 <colindixon> ah, raphael #chair me? 15:07:09 <raphael> #chair colindixon 15:07:09 <odl_meetbot> Current chairs: colindixon raphael 15:07:16 <colindixon> #topic agend bashing 15:07:54 <dlenrow> colin's buying monitors for the house :) 15:07:57 <colindixon> #link https://wiki.opendaylight.org/view/Network_Intent_Composition:Main#May_1.2C_2015 the agenda 15:08:35 <colindixon> #topic initial YANG / interface models 15:09:46 <colindixon> #link https://github.com/opendaylight/nic/tree/master/intentengine-api/src/main/yang the current YANG interfaces 15:10:40 <colindixon> #info it seems like the major changes are that we’ve split out primitive actions, and factored thigns apart into multiple yang files 15:11:32 <colindixon> #info dbainbri says we’d like to freeze this API in the next few days and wants feedback on that idea 15:11:40 <dlenrow> my microphone is not working. I would suggest we freeze a very skinny (minimum for SFC) for Li, and then immediately keep evolving our IM in parallel, including in Li timeframe 15:12:07 <dlenrow> Don't let Li freeze compete with continuing to iterate on what is community agreed "final" model 15:12:29 <colindixon> #info dbainbri and colindixon seem to agree that focusing on just allow and deny 15:12:52 <dlenrow> Don't we need a redirect/steer verb for Mathieu's SFC work? 15:13:19 <colindixon> #agree barring a major uprising, we’ll scale back the primitive actions to allow/deny, with agumentations for thigns like SFC (and maybe redirect steer for mlemay’s work) 15:13:20 <dlenrow> Need somebody with microphone to point dbainbridge at my comments 15:13:29 <colindixon> dlenrow: working on it 15:13:52 <colindixon> #info dlenrow raises concerns about just allow/deny for now 15:15:04 <colindixon> #action dbainbri to make the changes to trim down the primitives in the repository 15:15:39 <raphael> #info Now we have a high level intent model, but we're going to focus on actions allow and deny for lithium. 15:15:54 <colindixon> #action mlemay to work with dbainbri to make sure that the revised model will work for his issues in the Lithium time-frame 15:16:03 <dlenrow> Li IM is throw-away, fast fail learning opportunity 15:17:28 <colindixon> #info colindixon asks if we should have a “YANG expert” (e.g., Ed, Robert, etc.) look through things and make sure we’re not making any obvious mistakes 15:17:30 <raphael> #info raphaelamorim to push karaf console extensions with add, delete and list support today 15:17:52 <raphael> #action raphaelamorim to push karaf console extensions with add, delete and list support today 15:18:34 <colindixon> #undo 15:18:34 <odl_meetbot> Removing item from minutes: <MeetBot.ircmeeting.items.Action object at 0x19d72d0> 15:18:36 <colindixon> #undo 15:18:36 <odl_meetbot> Removing item from minutes: <MeetBot.ircmeeting.items.Info object at 0x1b78c90> 15:18:42 <colindixon> #topic Karaf CLI 15:18:51 <colindixon> #action raphaelamorim to push karaf console extensions with add, delete and list support today 15:18:58 <raphael> thanks colin 15:19:03 <raphael> wrong topic 15:19:04 <colindixon> #topic providers 15:19:31 <colindixon> #info there are two providers we’re looking at in Lithium: VTN and GBP 15:20:08 <dlenrow> I think Mathieu was seeking path of least resistance and staffing from VTN side made that the easiest path to SFC 15:20:26 <colindixon> #info dbainbri asks if we have any commitment from those projects, because at some point we will really want that 15:20:53 <dlenrow> Current POR Commitment is from innocybe to build SFC on VTN 15:22:22 <hideyuki> dlenrow: We don't target SFC. I mean that we planned to implement a VTN provider for SFC in Lithium. But, we don't have enough time for Lithium now. Therefore, we plan to implement a very simple VTN provider for simple scenario. 15:22:44 <dlenrow> Li needs architectural placeholder. OPNFV 2.0 will be built on Li and needs some SFC hooks/NBI to ship in Li 15:23:36 <dlenrow> Planning to support one use case SFC for Li. There will be an NBI and an implementation. 15:24:16 <dlenrow> Doing with neutron doesn't meet goal of demonstrating that NIC NBI works for SFC. 15:24:59 <dlenrow> It's about common NIC API on ONOS and ODL. ONOS will support OPNFV SFC for Rel 2.0. 15:25:16 <dlenrow> If Li ODL doesn't have it, ONOS wins as OPNFV SDN controller 15:26:16 <dlenrow> As of now, SFC demo is a committed Li deliverable AFAIK 15:26:29 <colindixon> #info colindixon asks if there’s a simple way that we could map NIC onto Neutron in Lithium so that we would have a provider 15:27:43 <colindixon> #info dlenrow says that we need to deliver an SFC demo on top of ODL and ONOS in the Lithium timeframe, it appears as though mlemay is the person who is staffing that effort 15:27:52 <dlenrow> Clearly hear that you need to see ODL code commits to start building confidence that we will get something in. 15:28:22 <dlenrow> Community might help best by pursuing parallel conversation on getting to extended IM with community concensus. 15:30:41 <colindixon> #info dbainbri say she hasn’t seen any evidence of that work in the Lithium release, which concerns him, mlemay says he will try to reshuffle resources to make that happen 15:30:59 <raphael> #info hideyuki is working on VTN provider and focusing on alow/deny action for lithium timeframe 15:31:04 <colindixon> #undo 15:31:04 <odl_meetbot> Removing item from minutes: <MeetBot.ircmeeting.items.Info object at 0x1b78b50> 15:31:06 <colindixon> #undo 15:31:06 <odl_meetbot> Removing item from minutes: <MeetBot.ircmeeting.items.Info object at 0x1b78d90> 15:31:12 <raphael> #info hideyuki is working on VTN provider and focusing on alow/deny action for lithium timeframe 15:31:13 <colindixon> #info dbainbri say he hasn’t seen any evidence of that work in the Lithium release, which concerns him, mlemay says he will try to reshuffle resources to make that happen 15:31:28 <colindixon> thanks raphael 15:31:33 <raphael> :D 15:31:39 <colindixon> #topic documentation 15:31:39 <dlenrow> I could help with doc\ 15:31:43 <hideyuki> raphael: Thank you! 15:31:50 <colindixon> #action dbainbri to work on docs for the model 15:32:28 <colindixon> #info dbainbri would love help with docs and especially adding anything for things for providers if we have any 15:33:19 <mlemay> Ok 15:33:23 <colindixon> #action raphael to work on integration testing for NIC 15:33:32 <colindixon> #topic trello board review 15:33:56 <colindixon> #link https://trello.com/b/L4GssBld/opendaylight-nic 15:36:14 <colindixon> #topic deliverable review 15:36:37 <colindixon> #link https://wiki.opendaylight.org/view/Network_Intent_Composition:Lithium_Release_Plan the release plan 15:37:13 <colindixon> #info we seem to be in good shape on: use cases, the YANG model, and the high-level design 15:37:26 <colindixon> #info the prototype and ONF SFC seem much, much scarier 15:39:10 <dlenrow> Will work with innocybe and HP team's to either present detailed plan to include SFC/VTN provider demo or will agree to miss Li milestones. Report back by email to lists early next week 15:39:15 <colindixon> #info in particular we don’t seem to have worked out either (a) if the model is sufficient for those and (b) if there’s a provider for the model that will be sufficient in the Lithium release 15:39:30 <colindixon> #action dlenrow Will work with innocybe and HP team's to either present detailed plan to include SFC/VTN provider demo or will agree to miss Li milestones. Report back by email to lists early next week 15:39:41 <dlenrow> Will work with innocybe HP, NEC team's to either present detailed plan to include SFC/VTN provider demo or will agree to miss Li milestones. Report back by email to lists early next week 15:40:08 <dlenrow> Apology, left out Hideyuki before 15:41:04 <dlenrow> Hoping Li has the architectural hooks for SFC, Maybe separate that from including demo in Li 15:41:59 <dlenrow> +1, giant disclaimer in Li that IM will change 15:42:47 <dlenrow> Info Model 15:43:51 <colindixon> #info there is a longer discussion around worries about missing delivering on our Lithium release plans 15:44:04 <dlenrow> Restart meetings for YANG wars in parallel to Li delivery 15:44:47 <mlemay> Yes... 15:44:58 <mlemay> Sorry got disconnected.. 15:45:30 <colindixon> #info colindixon and dbainbri say that there seems to be a real likelihood of shipping no NIC providers in Lithium, although it seems most likely that hideyuki and VTN will deliver one of the options 15:45:45 * dbainbri strongly disagrees with a restart of model wars. 15:46:01 <colindixon> #endmeeting