18:00:47 #startmeeting networking_policy 18:00:47 Meeting started Thu Aug 28 18:00:47 2014 UTC and is due to finish in 60 minutes. The chair is SumitNaiksatam. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:00:48 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 18:00:50 The meeting name has been set to 'networking_policy' 18:00:53 hi 18:00:59 s3wong: no worrries we will keep it short 18:01:06 rms_13: thanks for joining 18:01:33 #info agenda: https://wiki.openstack.org/wiki/Meetings/Neutron_Group_Policy 18:01:48 i thought we would keep the agenda open today 18:02:01 but since we have s3wong and rms_13 here, lets get a quick update from them 18:02:05 Incubator update/thoughts ? 18:02:08 #topic SG mapping driver 18:02:16 rms_13: hang on, we will get to that 18:02:34 s3wong: are you planning any updates to the patch? 18:02:51 yes 18:03:04 Ann has some comments, and I will update today or tomorrow 18:03:58 update the patch, that is 18:04:06 s3wong: ok great 18:04:06 s3wong: what about the update related changes? 18:04:23 before next Thursday? 18:04:25 s3wong: the update operation on the rules/contract 18:04:56 s3wong: and persisting the association between the SG rules and the policy rules 18:05:19 s3wong: I intend to review this in detail, but focusing on juno-3 patches right now 18:06:34 SumitNaiksatam: rkukura: with incubator, we are still on hook to get everything done by next Thursday 18:06:46 s3wong: not sure about that 18:06:46 s3wong: ? 18:06:54 or actually, getting it landed by next Thurs 18:07:14 s3wong: we are still going by the J3 time line though 18:07:16 rkukura: that was a question, not a statement, sorry :-) 18:08:50 SumitNaiksatam: yeah, these changes will take time. Not sure I can have them done by next Thurs 18:09:08 s3wong: ok, lets sync up on these 18:09:33 s3wong: so i am assuming the part that you have in right now is tested with UTs, and also functionally? 18:09:48 SumitNaiksatam: yes 18:10:00 s3wong: ok cool 18:10:47 s3wong: so you are able to drive an entire workflow from the creation of ep/epgs to associating with contracts containing rules? 18:11:16 SumitNaiksatam: the unit test actually does this 18:11:27 s3wong: okay :-) 18:11:46 then its a functional test, anyway, good to know though 18:11:52 SumitNaiksatam: SG association cannot happen until EPG/EP is associated 18:12:29 yes, i think we should do some functional testing with the point that we are at 18:12:35 so it was tested as such 18:12:52 s3wong: lets sync up, and make sure what we have all works in unison 18:13:05 SumitNaiksatam: sure 18:13:10 s3wong: thanks for the update, while you are at the other “real” meeting :;-) 18:13:19 #topic Horizon update 18:13:22 Horizon: Basically no progress. Review is open. Nobody has picked up yet. Knocked couple of doors with no luck. That particular patch is the baseline of how we want to organize templates. Once thats approved, one can move forward with doing grunt work of supporting for remaining objects. (l2, l3 etc) 18:13:23 rms_13: hi 18:14:09 * SumitNaiksatam gives rms_13 the prize for the quickest typed update ever! :-) 18:14:20 :). 18:14:35 rms_13: i will review 18:14:36 rms_13: link to the review? 18:14:45 LouisF: great, thanks 18:14:49 https://review.openstack.org/#/c/93590/ 18:14:56 LoiusF: Thanks 18:15:07 #link https://review.openstack.org/93590 18:15:10 rms_13: thanks 18:15:23 typically horizon tends to lag the other projects 18:15:46 of course, if the feature had already landed in neutron we would have seen a lot more activity 18:15:49 I figured 18:16:13 rms_13: thanks for persisting! 18:16:20 Hoping to make it one day :) 18:16:49 rms_13: i think that is the Neutron theme - “hoping to make it one day!” :-) 18:17:22 any questions for rms_13 on Horizon? 18:17:30 SumitNaiksatam: Private +2 18:17:56 I would like to ask few questions here once we are done with Horizon 18:18:01 rms_13: yes 18:18:09 #topic Heat 18:18:13 susaant: hi 18:18:38 susaant: i noticed you updated the wiki 18:18:48 #link https://review.openstack.org/#/c/111417/ 18:18:56 #link https://review.openstack.org/#/c/111419/ 18:19:28 SumitNaiksatam: Yes finally got around to that :) 18:20:14 susaant: good, thanks again to you as well for persisting and staying on message 18:20:51 any questions for susaant on the Heat patch 18:21:23 #topic CLI/Client 18:21:30 songole: anything to discuss here? 18:21:45 nothing much. no updates. 18:21:55 songole: thanks 18:22:03 #topic API intercept 18:22:16 #link https://review.openstack.org/#/c/109901 18:22:50 i reviewed some of this yesterday, and kevinbenton did a very neat job in trying to achieve this with minimal changes 18:23:03 rkukura: would you have time to look at this? 18:23:58 yes, I’ll make time 18:25:04 rkukura: thanks 18:25:09 #topic Open Discussion 18:25:45 so i had put a patch #link https://review.openstack.org/#/c/109901 to test the waters on how we can decouple a feature like GBP 18:25:56 so as not to have to make changes to the existing neutron “core” code when adding a new feature like GBP 18:26:19 i wil be revising that patch based on feedback, hence marked it WIP 18:26:49 did not mean to any major surgery of the extension loading code, just wanted to make it surgical engough 18:27:00 SumitNaiksatam: Did you give the correct link? 18:27:19 rms_13: my bad, thanks for catching 18:27:23 #link Avoid changing constants module when adding new service plugin 18:27:37 #link https://review.openstack.org/116996 18:27:42 wrong copy buffer! 18:28:12 anyway that just to lead up to your questions, rms_13 18:28:19 go ahead, you wanted to ask something earlier 18:28:24 Yes 18:28:52 I want to know incubator update wrt GBP 18:29:09 2. I want to know nova interatction plans for GBP 18:29:16 *interaction 18:29:52 rms_13: 1. i would like to know that as well, i am told to follow the mailing list, i think there are two or three threads going on 18:30:12 rms_13: i personally dont have an update beyond that 18:30:18 if anyone else knows more, please feel free to chime in 18:30:36 SumitNaiksatam: Ok. Unbelieavable but thanks 18:31:04 rms_13: which part is “unbelievable”? 18:31:28 SumitNaiksatam: Time it takes to reach conclusions 18:31:35 Anyways 18:31:53 I am not clear how the neutron incubator will play with heat ... 18:31:55 How do we want to do Nova integration? Do we have one plan in place 18:32:16 is there any clarity on that? 18:32:22 rms_13: good questions, i think those have been raised before, but fee free to raise them on one of the on going threads 18:32:39 err sorry, i meant to respond to susaant 18:33:08 *but feel free to raise them again 18:33:28 rms_13: on 2. in the first iteration, no changes to nova are required 18:33:29 SumitNaiksatam: ok. 18:33:54 rms_13: since on the nova side we will still be using the port 18:34:00 SumitNaiksatam: Ok. 18:34:17 rms_13: past that, there are obviously thoughts on how we can integrate better 18:34:18 For 2nd iteration: nova boot --gbp group-id=x? 18:34:43 rms_13: but i think for that we need the GBP code to stabilize so that we can take the discussion to the nova team as well 18:35:10 rms_13: but duly noted that you are interested in this discussion, and we will follow up once we are past this phase 18:36:20 SumitNaiksatam: I think that effort can be made parallelized while we stabilizes GBP. Atleast we can get the discussion going with the Nova team once we have full consensus amongst us 18:36:31 Post 9/4 maybe? 18:37:12 rms_13: yeah definitely post 9/4, we have to consider the right time after that 18:37:53 rms_13: i think our priority would be the get the first iteration settled, and not distract other teams from what their focus is on in terms of Juno deliverables 18:37:54 I would like to read thoughts on that so far, if there are any captured. 18:38:06 Absolutely. 18:38:15 I am not proposing that at all. 18:38:33 rms_13: yes, you will have to read our minds :-) 18:38:43 SumitNaiksatam: I just want to discuss that early as I think that will be one beast of an integration if we do it right 18:38:59 SumitNaiksatam: Very difficult. Will pass than :) 18:39:12 Anyways thank you 18:39:15 rms_13: “beast” - really? 18:39:27 SumitNaiksatam: i had the question on order of services in a service chain used by gbp contract 18:39:35 LouisF: yes sure 18:40:21 if chain={si, s2} are they applied from consumed epg to produced epg, or vice versa? 18:40:50 it is not stated anywhere 18:42:46 this is a follow on to my earlier query on the service chain bp 18:42:58 LouisF: yes, and thanks for the follow up 18:43:46 regarding order of services and direction of traffic 18:44:02 LouisF: with regards to the GBP model we are only saying consumed to produced 18:44:08 LouisF: and i believe at that point the semantics of the service chain take over 18:44:55 sounds good can we add that to the gbp spec 18:46:42 LouisF: i think the GBP spec already captures that in terms of the redirect action, but we can certainly add clarifications if required 18:46:50 LouisF: good point to bring up though 18:47:29 anything else for today? 18:47:49 or we can get back 13 minutes to focus on J3 reviews 18:49:04 alrighty, thanks everyone for attending 18:49:10 bye 18:49:12 #endmeeting