05:33:29 <soichi> #startmeeting taas
05:33:30 <openstack> Meeting started Wed Feb  1 05:33:29 2017 UTC and is due to finish in 60 minutes.  The chair is soichi. Information about MeetBot at http://wiki.debian.org/MeetBot.
05:33:31 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
05:33:33 <openstack> The meeting name has been set to 'taas'
05:33:37 <yamamoto> hi
05:33:46 <soichi> yamamoto: hi
05:33:57 <soichi> #link https://wiki.openstack.org/wiki/Meetings/taas
05:34:09 <soichi> #topic A talk for Boston Summit (deadline: Feb 6th)
05:34:20 <soichi> i hope we can submit a talk for Boston Summit
05:34:36 <soichi> do you have any idea about contents of our talk?
05:35:00 <reedip> soichi : I have been discussing with tacker team
05:35:44 <reedip> I was trying to find something related to Virtualized Tap interfaces in the current SDN solutions and if we can tie up TaaS with those SDN solutions
05:36:29 <soichi> reedip: it sounds good
05:36:41 <yamamoto> what do you mean by virtualized tap interfaces?
05:39:27 <soichi> i guess taas interface. That is, not using physical tap device.
05:40:00 <reedip> soichi : yes
05:40:21 <yamamoto> use what instead?
05:41:17 <reedip> yamamoto : basically we have the levels defined as User -> Neutron -> OVS for TaaS, isnt it ?
05:41:31 <reedip> I mean the control path moves in that way
05:42:18 <yamamoto> yes
05:42:35 <reedip> Instead of OVS , can we use OpenContrial or ODL  ?
05:43:51 <soichi> i think reedip's idea is a presentation how to use TaaS with SDN solutions such as Contrail, Nuage, and so on
05:44:00 <reedip> Something like that
05:44:11 <yamamoto> is it more than a new driver?
05:46:41 <reedip> Currently I think it may just be a new driver
05:46:52 <soichi> be that as it may, but i'm not sure
05:48:51 <soichi> does it make sense we can say taas is applicable not only OVS but also Contrial and etc?
05:49:34 <yamamoto> midonet, plumgrid, and vmware already have its taas driver and they are SDN.  i'm not sure what's new for opencontrail and odl.
05:50:02 <yamamoto> maybe policy-based something?  i'm not sure.
05:51:18 <yamamoto> (well, i'm not sure about plumgrid.)
05:52:12 <reedip> yamamoto : do all of these solutions provide similar performances ?
05:52:18 <reedip> or functionalities?
05:53:47 <yamamoto> i'm familiar with midonet but i don't know about others.
05:54:41 <yamamoto> midonet provides more than the reference implementation. any port can be tapped, for example.
05:55:23 <soichi> performance comparison sounds interesting for me (but i think it is difficult to prepare testbed)
05:57:02 <yamamoto> yes.  1-node sdn performance doesn't make much sense. :-)
05:59:09 <reedip> :- )
06:01:08 <reedip> so what else do ytou suggest soichi ! :D
06:01:25 <soichi> i'd like to consider a little bit more whether we can say something about using taas with SDN solutions.
06:02:01 <soichi> let's disscuss offline because the deadline is 6th Feb (next Monday)
06:04:11 <yamamoto> reedip: what kind of input from tacker folks you got?
06:04:36 <reedip> In discussion with them , they look positive
06:05:05 <yamamoto> they have some interesting way to use taas?
06:05:31 <reedip> yamamoto : they are familiarizing with taas, so they have not given complete inputs yet
06:07:53 <soichi> Can we go on next topic (Open Discussion)?
06:08:06 <yamamoto> ok. i'm interested in their opinion.  (as i have ~no knowledge about nfv)
06:08:48 <yamamoto> soichi: sure.  i have no topic.
06:08:58 <soichi> #topic Open Discussion
06:09:28 <soichi> reedip: do you have topic?
06:14:04 <soichi> it seems we have no more topics for today.
06:14:48 <soichi> so i'd like to close today's meeting.
06:15:09 <yamamoto> +1
06:15:11 <soichi> #endmeeting