18:43:54 #startmeeting Networking FWaaS 18:43:55 Meeting started Wed May 7 18:43:54 2014 UTC and is due to finish in 60 minutes. The chair is SumitNaiksatam. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:43:56 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 18:43:58 The meeting name has been set to 'networking_fwaas' 18:44:11 so we discussed the priorities over the last week 18:44:22 SumitNaiksatam, I'll stay for this one :-) 18:44:35 sballe: sure, most welcome :-) 18:44:43 SumitNaiksatam, thx :) 18:45:35 #info etherpad for summit: https://etherpad.openstack.org/p/juno-fwaas 18:45:44 SumitNaiksatam: have put in some basic stuff 18:45:49 SridarK: thanks much for the updates! 18:46:10 #topic Integration with Flavor Framework 18:46:13 we can update on the specific BP links shortly 18:46:33 gduan: you are going to take this up, or enikanorov will do it? 18:46:49 i am guessing enikanorov is swamped 18:46:52 i think STF is still pending 18:47:02 enikanorov: i agree 18:47:05 that needs to be merged to start working with flavors integration 18:47:17 enikanorov: ok, so we dont have consensus on that? 18:47:31 enikanorov: we will also have more discussion on this in ATL on STF ? 18:47:33 enikanorov: sorry, i forgot to bring it up in the adv services meeting! 18:47:38 i think we do. gduan: could you update the patch with new version? 18:48:06 SridarK: i think i'll cover it briefly on design track 18:48:07 enikanorov: you mean sync? 18:48:16 enikanorov: i believe from the last meeting, the PTL agreed to the STF direction? 18:48:31 gduan: i don't remember the most recent version, so did you remove provider attr from the public API? 18:48:49 enikanorov: ok. not yet. I will do that 18:49:00 gduan: ok, that was the most important change 18:49:27 enikanorov: sure. will do it today/tomorrow and submit spec too 18:49:29 SumitNaiksatam: i think Kyle agreed, but markmcclain's concerns need to be addressed 18:49:31 (if any) 18:49:50 enikanorov: ok, i think we need to get a move on 18:49:58 SumitNaiksatam: one sec 18:50:15 enikanorov: if required lets have a separate meeting and plough through any objections 18:50:22 SumitNaiksatam: i guess we keep the same BP 18:50:44 SridarK: we can keep the same bp, but we will need to create a bp spec in review 18:50:45 SumitNaiksatam: okay 18:50:57 SumitNaiksatam: i'll talk with Mark 18:51:17 SumitNaiksatam: ok 18:51:19 enikanorov: we can invite the PTL, markmcclain1 and whoever else is interested 18:51:32 enikanorov: thanks 18:51:36 SumitNaiksatam: this will be good 18:51:46 SridarK: good point to bring up from process perspective 18:52:16 so once the STF issue is settled, who is doing the flavor’s patch for FWaaS? 18:53:05 ok perhaps we need to get STF patch in first 18:53:16 gduan: you need to file a new bp for STF, right? 18:53:31 gduan: assumign we have thumbs up for using STF 18:53:48 SumitNaiksatam: just spec for review, right? 18:53:55 gduan: yeah 18:54:05 SumitNaiksatam: will do it today 18:54:12 SumitNaiksatam: but it will be very short 18:54:15 gduan: thanks! 18:54:21 gduan: yes sure 18:54:28 enikanorov: btw, is there a new spec for STF? 18:54:44 enikanorov: you mentioned changes earlier, where have those been made? 18:54:48 SumitNaiksatam: not yet, haven't managed to prepare it 18:54:53 enikanorov: ah ok 18:55:01 enikanorov: but you have a patch in review? 18:55:18 SumitNaiksatam: i meant the changes of gduan's patch 18:55:24 enikanorov: ok 18:55:28 where public STF attributes need to be removed 18:55:48 enikanorov: ah got it, to prepare it for flavors integration 18:56:30 gduan: so no dependency for you as far as this is concerned 18:57:45 so lets get the STF in first, and circle back to the flavors patch 18:57:59 #topic FWaaS Service Insertion 18:58:15 so we have to create a bp for this as well 18:58:36 and it will have a dependency on the service insertion stuff discussed earlier 18:58:51 we also need to check with kanzhe if he is planning to add support for this 18:58:55 in fwaas 18:59:23 but this is our priority for juno to be able to insertion on a specific router 18:59:36 SumitNaiksatam: ok was just going to ask if that will only cover the base infrastructure or have a sample implementation 18:59:57 #action SumitNaiksatam to check with Kanzhe on service insertion support for fwaas 19:00:03 SumitNaiksatam: If Kanzhe does that - we will get it for free :-) 19:00:17 SridarK: yeah that was my understanding :-) 19:00:53 SumitNaiksatam: so we should not refer to the old BP on this in the ether pad - i will remove that 19:01:01 ok 19:01:04 let be there for now 19:01:09 ok 19:01:14 since there is no insertion bp 19:01:23 ok got it 19:01:29 in fact let the old one be there, and we can add a second one 19:01:37 ok 19:01:40 it will help people to know the history 19:01:46 SridarK: thanks for bringing that up 19:02:00 np 19:02:10 #topic FWaaS Service Objects 19:02:19 beyounn: we are on track for this? 19:02:34 beyounn: submit the bp in gerrit that is? 19:03:21 gduan: can you check with beyounn? 19:03:40 we need the bp to be in gerrit before the summit, so that we can discuss 19:04:22 #topic FWaaS Zones 19:04:36 SridarK: i know you have been busy with the critical bug 19:04:53 SumitNaiksatam: working on the write up based on last weeks discussions 19:05:01 SridarK: ok 19:05:10 will have it gerrit in tne next day or two 19:05:33 SridarK: lets also keep an eye on the service insertion discussion since these are interpendent 19:05:49 SridarK: i would recommend first putting in google doc 19:05:55 SumitNaiksatam: yes very much so - will indicate dependency on that 19:06:04 SridarK: it might be faster and easier to draw diagrams 19:06:06 SumitNaiksatam: ok that makes more sense 19:06:15 will do that 19:06:24 SridarK: you can then fill the gerrit template at leisure 19:06:32 sounds good 19:06:51 SridarK: we want validate the proposal first 19:07:11 ok we can discuss first then 19:07:25 SridarK: i know we spent a lot of time earlier on this, but i still dont have a crisp idea in my mind 19:07:53 SridarK gduan: one question i have though, is it always possible to “infer” zones? 19:08:11 SumitNaiksatam: ok and in any case there are the dependencies and that influences a lot of how we specify 19:08:40 SumitNaiksatam: hmm! infer - i am not sure we can do that always 19:08:49 meaning if the particular firewall backend implementation always requires a zones configuration, can the driver map the neutron constructs to a zone? 19:08:56 SridarK: ok 19:09:09 SridarK: its ok if we cannot handle all the cases 19:09:26 SumitNaiksatam: u are thinking of the default case kind of scenario ? 19:09:37 SridarK: yeah 19:10:10 SumitNaiksatam: that i think we can do by having something like a any zone to any zone 19:10:20 SridarK: ah ok 19:10:43 SridarK: i was hoping a little more specific than that, but i myself dont have a clear idea in mind 19:10:52 SridarK: so i am just loud thinking 19:10:53 i probab did not say that very well but basically like a regular firewall 19:11:07 SridarK: ok 19:11:21 SumitNaiksatam: will clarify that in the doc 19:11:34 SridarK: do we know if any of the vendor folks who use zones are going to be around for the discussion? 19:12:15 SumitNaiksatam: not sure on that - i am trying to troll our mktg guys for some user scenarios that they can share 19:12:23 SridarK: :-) 19:12:37 SridarK: if they dont want it, we are off the hook?!? :-P 19:13:05 moving on 19:13:09 #topic FWaaS Ceilometer Requirements 19:13:10 SumitNaiksatam: i think i need to rephrase my ask in that way - i am sure i will get a quick response :-) 19:13:19 SridarK: :-) 19:13:34 SridarK: is prad around? 19:13:57 you have updated the etherpad with the requirements 19:14:08 yes added the basic ones 19:14:19 my understanding was the same in speaking to him 19:14:21 i will need to also reflect the usage - metrics that we can do 19:14:38 who is filing the bp for this? 19:14:40 or rather we dont have to do anything for that 19:14:57 He has a BP on the Ceilometer side 19:15:03 will prad have the bandwidth to do it in neutron? 19:15:08 at least file the bp 19:15:10 not sure on that 19:15:17 will follow up on that 19:15:23 i would rather prefer that the requirements are driven from ceilometer 19:15:46 i mean someone from that team files the bp on the neutron side as well 19:16:08 yes makes sense - will let him know and we can prioritize the implementation based on bandwidth 19:16:10 i believe RajeshMohan mentioned he would have time to pursue this if needed 19:16:15 SridarK: thank 19:16:18 s 19:16:23 ok cool 19:16:44 #topic Vendor drivers 19:17:02 gduan: there? 19:17:30 Hmm seems like gduan dropped off 19:17:56 But he did mention that they may be doing a refactor of the existing 19:17:59 impl 19:18:09 ok, i wanted to track the bp 19:18:19 it needs to be added 19:18:33 I also see Cisco driver 19:18:37 who is doing it? 19:18:53 i mean who is creating the bp for it? 19:18:58 Yes i will file the BP and push for Gerrit review 19:19:03 SridarK: ok good 19:19:19 on implementation perhaps others will join in as well not clear 19:19:31 SridarK: yes, good, better to get help on that 19:19:39 SridarK: we will need you on core fwaas 19:19:55 SumitNaiksatam: :-) 19:20:18 SridarK: though it should be made clear that blueprints expire at the end of the release cycle 19:20:33 at least thats my current understanding 19:20:36 SumitNaiksatam: may need some vendor extensions on that BP will get ur thoughts on that offline 19:20:45 SumitNaiksatam: ok 19:21:32 SridarK: ok sure 19:21:37 #topic open discussion 19:21:58 i think we will have our plates overflowing with the above in Juno 19:22:14 SumitNaiksatam: yes i think Juno will be a tight release 19:22:19 so i think this is a good plan to be discussed in the summit 19:22:33 look fwd to that - it will be good to have something shipped in Juno 19:22:35 SridarK: yeah i dont know how far we can get with the above 19:22:53 i guess it all depends on the reviewer turn around! 19:23:03 SumitNaiksatam: we will give it our best push 19:23:10 SridarK: absolutely (for a change) ! :-) 19:23:33 also we have to go through two rounds of reviews 19:23:41 one for the bp and one for the code 19:23:48 With the BP reviews - it is good to get folks plugged in early and hopefully will help on the code reviews 19:23:56 hopefully the bp review wont take the entire cycle :-P 19:24:01 :-) 19:24:14 alrighty lets wrap it! 19:24:21 SumitNaiksatam: sounds good 19:24:25 anything else? 19:24:31 no nothing 19:24:37 see u at ATL 19:24:52 will send u email on the zones doc 19:24:57 SridarK gduan beyounn sballe: thanks for joining, see you in ATL 19:24:58 bye 19:25:01 #endmeeting