18:01:32 #startmeeting networking_policy 18:01:34 Meeting started Thu Oct 16 18:01:32 2014 UTC and is due to finish in 60 minutes. The chair is SumitNaiksatam. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:01:35 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 18:01:37 The meeting name has been set to 'networking_policy' 18:01:41 #info agenda: https://wiki.openstack.org/wiki/Meetings/Neutron_Group_Policy 18:01:56 glebo is gone... 18:02:08 any big announcements or info that anyone would like to share upfront? 18:02:17 s3wong: he will back soon :-) 18:02:37 one piece of information i wanted to share was the GBP design summit session 18:02:51 #link http://kilodesignsummit.sched.org/event/98dc4255384e340682137c8a7ee7e60d#.VD_6XIt4r4w 18:03:15 this is a separate session in the “other projects” track (in the design summit) 18:03:48 i am of course talking in the context of the upcoming Paris summit (should have mentioned that before) ;-) 18:03:50 cool 18:03:56 s3wong: yeah :-) 18:04:04 SumitNaiksatam: do we know if we are getting Neutron design summit slots? 18:04:13 s3wong: not sure 18:04:27 s3wong: we will leave that to community to decide the neutron priority 18:04:37 okay we can discuss the details of the session in a separate agenda topic 18:04:40 SumitNaiksatam: sure 18:05:01 any other info/announcements? 18:05:36 #topic Specs in review 18:05:43 #link https://review.openstack.org/#/q/status:open+project:stackforge/group-based-policy-specs+branch:master,n,z 18:06:11 some of the ones mentioned above already have implementation patches posted against them 18:06:24 we are little behind on the reviews 18:06:56 SumitNaiksatam: I should be able to make some progress on these today. 18:07:04 It took time for jenkins to be happy... Glad he is smiling now :) 18:07:15 rkukura: thanks 18:07:35 SumitNaiksatam: Sorry, I actually meant the code reviews, but I need to review the specs as well. 18:07:38 rms_13: yes, but we are currently discussing the specs, implementation patches are next :-) 18:07:47 rkukura: yeah, this is the specs review 18:08:14 LouisF: good to see your specs, thanks for posting promptly! 18:08:15 need review of the gbp chaining spec https://review.openstack.org/#/c/125876/ 18:08:41 hemanthravi: sorry i am behind on that 18:08:56 SumitNaiksatam, when you get some time 18:09:38 i see that only kevinbenton has reviewed 18:09:44 the ones that were assigned to him 18:09:57 ivar-lazzaro: rkukura do you have issues with kevinbenton’s comments? 18:09:59 SumitNaiksatam: thanks 18:10:20 LouisF: you want to update the commite message title to reflect the name of the spec 18:10:21 SumitNaiksatam: nope, I'm going to push the new version of the patch in seconds :) 18:10:26 ivar-lazzaro: sweet 18:10:39 SumitNaiksatam: i will be updating them 18:10:43 SumitNaiksatam: I think I clarified kevinbenton’s concern about L2P, and am updaing the extension driver spec. 18:10:49 LouisF: great, thanks 18:10:56 rkukura: ok great 18:11:07 rkukura: sorry i was not following the discussion thread 18:11:15 glebo: welcome back! 18:11:25 *sorry he's late, irc client connectivity issues, dkw* 18:12:11 #nick glebo 18:12:15 can i also plug a request to review the spec update that i put: #link https://review.openstack.org/127913 18:12:56 i am happy to spend a few minutes here if someone wants to dig deeper on the specs that are in review currently 18:13:08 on -> into 18:13:23 if not we will move to the implementation patches 18:14:07 ok moving on 18:14:19 #topic Implementation patches in review 18:14:36 wait 18:14:38 #undo 18:14:39 Removing item from minutes: 18:14:52 #topic Current DB approach: SSDC 18:15:09 want to bring this to everyone’s notice since this patch was already approved 18:15:42 SumitNaiksatam: so no problem with the release folks? 18:16:04 per discussion in the last meeting (and on friday in the review meeting), ivar-lazzaro posted a new patch for the single schema different series approach (SSDS) 18:16:10 s3wong: i am not sure 18:16:20 ivar-lazzaro: rkukura: s3wong’s question ^^^ 18:16:43 s3wong: we don't know yet, but we needed this to be in place to fix dependency issues 18:16:50 ivar-lazzaro: +1 18:17:01 yes, installation was broken without a fix 18:17:13 so in the interest of expediency, i approved this patch 18:17:21 s3wong: we can always go back to the DSDC approach if the problem arises 18:17:31 i believe at least rkukura approves of this 18:17:40 please let us know if there are concerns 18:17:52 ivar-lazzaro: sure 18:17:55 But in general I would suggest to start investigating the issue with the packagers before it's too late :) 18:17:57 i have tested that devstack install now works with the fix in place 18:18:42 ivar-lazzaro: can you quickly summarize for the benefit of everyone what the current approach is? 18:18:54 SumitNaiksatam: sure 18:18:55 ivar-lazzaro: sorry to put you on the spot 18:19:23 The current approach is to use a different "version table" in the same database/schema Neutron is using 18:19:47 so we are basically still tied to the Neutron's database, but we run a completely different migration chain 18:19:54 ivar-lazzaro: good summary 18:20:06 which means that Neutron can upgrade just fine as long as table names are unique 18:20:16 (same goes for gbp ofc) 18:20:17 ivar-lazzaro: and thanks so much for working on this so expeditiously, it was a blocking issue 18:20:38 okay any more questions on the current approach? 18:20:51 so for anyone posting a new migration, nothing changes 18:21:03 Nope 18:21:12 just change the HEAD in the gbp repo, and point to the right down migration 18:21:24 this approach is much cleaner than DSDC 18:21:29 ivar-lazzaro: did I summarize it correctly? 18:21:38 but less safe packaging wise 18:21:45 SumitNaiksatam: that's perfect :) 18:21:47 Why less safe? 18:22:12 certainly simpler! 18:22:20 rkukura: because we don't know if using the same database is acceptable overall 18:22:39 rkukura: once we know that for sure I'll be happier :-) 18:22:43 I think that is more of an architectural issue than a packaging issue 18:22:50 But we can ask around 18:22:55 rkukura: sure 18:22:57 ok moving on 18:23:08 For Juno, this is a service plugin within neutron-server, after all 18:23:08 ivar-lazzaro: thanks again! 18:23:15 rkukura: agreed! 18:23:16 SumitNaiksatam: np! :-) 18:23:21 #topic Implementation patches in review 18:23:34 so we had a long review meeting (i believe it was Oct 14th) 18:23:55 nick/ glebo_ 18:23:56 this was an IRC meeting, you can find the logs here: #link https://wiki.openstack.org/w/images/7/7f/Gbp-patch-reviews-10142014.pdf 18:24:14 we got out of the meeting with some AI and review owners 18:24:45 this has been captured here: #link https://docs.google.com/spreadsheets/d/1VAcYODD5fA_t2yleb95pTlIxD-btnUllZTN-uhNrsh4/edit#gid=0 18:25:30 lo mestery 18:25:55 so we are still a bit behind on the reviews here 18:26:07 we merged GPM-RMD-1, did we? 18:26:15 all: having serious irc issues this morning, so Guest96350 = glebo 18:26:34 *hope y'all can live w/ it* 18:26:37 Guest96350: np 18:26:51 ok we did 18:27:07 on GP-API-2, rms_13 you have a -1 18:27:08 *will repost the mapping once in a while to help any new-comers* 18:27:19 i believe i have addressed your comments 18:27:20 I just updated the gpm-rmd-1 review with links to the LP bugs I filed. 18:27:35 Ok. Will take a look 18:27:44 rkukura: thanks 18:28:04 rms_13: i addressed them yesterday evening :-P 18:28:06 rkukura: gpm-rmd-1 is already merged, right? 18:28:11 s3wong: yes 18:28:19 ivar-lazzaro, does hierachical contracts patch need a rebase 18:28:22 s3wong: yes, just wanted to record that the bugs were filed 18:28:24 s3wong: i think rkukura added as record 18:28:31 rkukura: sure 18:28:40 ivar-lazzaro, service-chain patches dependent on this are failing too 18:28:41 hemanthravi: yep 18:28:48 hemanthravi: i believe it does, i did not rebase it yesterday when i was rebasing others 18:29:00 hemanthravi: okay i did not realize that 18:29:13 hemanthravi: there are merge conflicts, will address right after the meeting 18:29:13 i can do it later in the day if ivar-lazzaro doesnt get to it 18:29:22 ivar-lazzaro, thanks 18:29:47 for GP-DB-2 18:30:17 hemanthravi: and ivar-lazzaro you +2’ed 18:30:33 if no one has objections, can someone +A this? 18:30:47 SumitNaiksatam: there's still a dict/set issue (just like the contract patch) but didn't want to block it 18:30:50 sorry to get into the weeds here with the specific patches 18:31:05 SumitNaiksatam: I can approve it 18:31:38 ivar-lazzaro: ah okay, i believe you added that comment a couple of mins back :-) 18:31:47 SumitNaiksatam: yes :) 18:31:53 ivar-lazzaro: but my bad, i should have checked for it when i was fixing the other stuff 18:31:59 ivar-lazzaro: you can file a bug 18:31:59 reminder: Guest96350 = glebo *suffering from irc client disfunction syndrome this morning* 18:32:08 Guest96350: ;-) 18:32:21 Guest96350: we at least know you are not masquerading! ;-) 18:32:33 \ME starts doubting that Guest96350 really is glebo :-) 18:32:43 ivar-lazzaro: i can assure you he is 18:32:52 GP-PLG-3 18:32:52 SumitNaiksatam: j/k :) 18:32:58 SumitNaiksatam: well, reverse masquerading, aka parading 18:33:03 who is the second core on this? 18:33:30 sorri, i meant GP-PLG-2 18:33:43 rkukura: that would be you ;-) 18:33:53 ivar-lazzaro: who is glebo anyway, really? 18:33:53 yep 18:34:02 hopefully you can take a quick look at it 18:34:17 Guest96350: maybe we all are! 18:34:30 GP-API-3 18:34:35 SumitNaiksatam: Yes, I hope to make some progress on the code reviews today 18:34:42 rms_13: you are guilty again ;-P 18:34:55 what did I do now? 18:35:03 :) 18:35:09 rms_13: just pulling your leg 18:35:57 rms_13: you put the -1 saying that jenkins was not happy, except that jenkins had not voted 18:36:06 rms_13: but it has now and its happy :-) 18:36:22 ivar-lazzaro: *hears familiar tune, then lyrics, "your own... personal... glebo" 18:36:41 rkukura: i believe you will be looking at GP-API-3 as well 18:37:03 May be the session was not upto date... Ya its passing so I will do +2. 18:37:16 SumitNaiksatam: yes 18:37:26 for GP-DB-3 i believe i have addressed the comments, ivar-lazzaro and rms_13 can you take a look? 18:37:35 SumitNaiksatam: sure 18:37:37 Rebase after Rebase and that after +2 ... :) 18:37:40 Sure I will 18:37:53 rms_13: great, thanks! 18:38:14 GP-PLG-3 18:38:46 i think this is good to go, we can wait for the preceeding patches to merge and then +A this 18:39:05 btw, lets merge the patches in order (when we come to it) ;-) 18:39:24 i am not sure Jenkins will allow a merge otherwise 18:39:36 that brings us to GPM-RMD-SG 18:39:53 Ya that is the reason I did not +A after Hemanth did another +2 on one of them. 18:39:57 so myself and rkukura are on the hook, we will get to it once the backlog clears 18:40:11 rms_13: sure 18:40:35 and we also have the hierarchical contracts patch 18:40:45 thats Hemanth and me 18:40:52 i think hemanth has reviewed to some extent 18:40:56 i havent 18:40:58 will get to that 18:41:18 so once we are done with these foundation patches, we have a bunch of service chain patches to get to 18:41:33 perhaps doesnt make sense to spend time on discussing those right now 18:41:47 but please keep those patches in mind as well 18:42:00 SumitNaiksatam: there's another one 18:42:04 SumitNaiksatam: lost in the bunch 18:42:11 ivar-lazzaro: ah sorry, did i miss something 18:42:13 SumitNaiksatam: which rkukura will certainly appreciate :) 18:42:23 #link https://review.openstack.org/#/c/128148/ 18:42:51 since we are now depending from Neutron only for UTs, we can forget about pypi and directly pull from their repo 18:42:56 ivar-lazzaro: oh my bad, yeah that is critical bug fic 18:42:58 the official one 18:42:59 *fix 18:43:23 ivar-lazzaro: So it does work? 18:43:48 rkukura: yeah now it does, because we don't need to pull the repo anymore when running the installation script 18:44:00 rkukura: it's only needed for UTs which have always worked 18:44:11 rkukura: need to point to stable/juno though 18:44:12 ivar-lazzaro: oh sorry, i mixed up 18:44:24 ivar-lazzaro: was just going to mention stable-juno 18:44:29 ivar-lazzaro: i thought you were referring to the delete object chain bug 18:44:50 SumitNaiksatam: that's another one, I didn't have the chance to test it yet 18:44:51 #link https://review.openstack.org/#/c/126740 18:45:15 yes, we have lots of things in flight, if you are still wondering! :-) 18:45:24 okay, so moving on 18:45:40 #topic GBP Design Summit session 18:45:54 as i mentioned earlier - #link http://kilodesignsummit.sched.org/event/98dc4255384e340682137c8a7ee7e60d#.VD_6XIt4r4w 18:46:00 we need to plan for this 18:46:17 not sure that we need to discuss this today 18:46:34 we have time until the summit 18:46:57 s3wong: you were mentioning something about this earlier? 18:47:26 SumitNaiksatam: 2 wks away now, not too much time, really 18:47:35 fyi, the GBP conference track session is #link https://openstacksummitnovember2014paris.sched.org/popular#.VD5YAYvF_3B 18:47:39 Guest96350: true 18:47:49 Guest96350: we just got notification that we have been allocated this slot 18:48:11 Guest96350: so in the process of planning, and the rest of the team just got to know about it as well 18:48:16 Guest96350: but point well taken 18:48:30 Guest96350: in general, this will be a “design” discussion 18:48:36 Guest96350: i will get the etherpad started for it 18:48:47 Guest96350: so that we can start adding the topics that we want to discuss 18:49:10 and over time we can add structure 18:49:17 does that sound good to everyone? :-) 18:49:29 SumitNaiksatam: +1 18:49:29 my earlier comments not specifically targeted at Guest96350 :-) 18:49:34 SumitNaiksatam: sorry, was out for a moment 18:49:34 * Guest96350 sees SumitNaiksatam take his 842nd action item of the mtg ;-) 18:49:43 Guest96350: :-) 18:50:07 SumitNaiksatam: we should plan for the conference presentation, as it happens on the first day of the summit :-) 18:50:11 s3wong: no worries 18:50:16 s3wong: ah on that 18:50:26 s3wong: i was planning to bring that up in the open discussion 18:50:33 s3wong: i did update the slides 18:50:40 s3wong: at least with those from the policy summit 18:50:44 s3wong: i will send them out 18:50:47 SumitNaiksatam: and of course the "other project" design summit also :-) 18:50:55 SumitNaiksatam: OK 18:50:56 s3wong: i believe others are working on the slides as well 18:51:09 s3wong: so will send it as soon as i get that input 18:51:12 SumitNaiksatam: others? I am not :-) 18:51:22 s3wong: mandeep and mikeC 18:51:44 ok thats a good way to segue to the open discussion 18:51:48 #topic Open Discussion 18:51:50 SumitNaiksatam: so those will be sent to ML? 18:52:07 SumitNaiksatam: slides, I meant 18:52:18 Guest96350: perhaps not the ML right away, they will posted on the openstack site after the presentation 18:52:47 Guest96350: we expect to iterate until the presentation day, so might not be worthwhile sending the draft versions to the ML in advance 18:53:00 Guest96350: can certainly share within the team for feedback and review 18:53:03 SumitNaiksatam: so, sent unicast to "interested parties for reivew"? 18:53:11 Guest96350: yes, absolutely 18:53:24 ok we are in open discussion 18:53:26 SumitNaiksatam: cross-talk; ack 18:53:45 LouisF: i noticed you and cathy_ were following up on the Congress integration? 18:53:54 LouisF: thanks for following up on that thread 18:54:06 LouisF: any update you would like to give to the rest of the team 18:54:24 or perhaps on the blueprints you posted? 18:55:24 perhaps LouisF is away 18:56:07 s3wong: you have any updates on that front? 18:56:26 s3wong: you mentioned the operators track as well for socializing GBP? 18:57:23 i think we lost a few people since we are getting close to lunch time here on the west coast :-) 18:57:36 We are dividing the propose BluePrint in 3 parts. 18:57:37 SumitNaiksatam: ahah 18:57:53 nbouthors: nice to know, and good to see you here :-) 18:58:46 hi 18:59:05 I work with Louis on the BP 18:59:23 nbouthors: yes of course, sorry i did not realize you were in the meeting today 18:59:47 Also, it is nice to see that OVS is starting to implement L7 classification. 18:59:48 so the blueprint which LouisF posted is a joint work between nbouthors LouisF and cathy_ 18:59:55 nbouthors: right? 19:00:00 SumitNaiksatam: yes 19:00:20 nbouthors: yes, noticed your pointer earlier today regarding the OVS work 19:00:25 nbouthors: will follow up 19:00:41 THought it could be of interest 19:00:59 nbouthors: i think the team will be happy to get some pointers to docs or references if you have any 19:01:11 nbouthors: we can take it offline 19:01:13 okay we are over time 19:01:15 SumitNaiksatam: Wil do 19:01:16 thanks everyone 19:01:19 nbouthors: thanks 19:01:22 Bye 19:01:24 SumitNaiksatam: sorry, again stepping out 19:01:35 until next week (we might have another review meeting if we dont make enough progress) 19:01:41 bye all 19:01:45 bye 19:01:45 #endmeeting