18:03:19 #startmeeting networking_policy 18:03:21 Meeting started Thu Apr 7 18:03:19 2016 UTC and is due to finish in 60 minutes. The chair is SumitNaiksatam. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:03:23 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 18:03:25 The meeting name has been set to 'networking_policy' 18:03:43 #info agenda https://wiki.openstack.org/wiki/Meetings/GroupBasedPolicy#April_7th.2C_2016 18:03:46 rkukura: no worries 18:04:02 so just a quick heads up at the outset 18:04:45 the gate is broken because of a new version of a dependency lib (fixtures) 18:05:00 #link https://review.openstack.org/#/c/302737/ 18:05:20 its being fixed in neutron, and possibly in upper-constraints as well, but until then our py27 will fail 18:05:33 OChi 18:05:34 its failing on the stable branches as well 18:05:38 igordcard: yeah 18:06:10 and barring a few, most tests fail due to this 18:06:31 it might be possible to pin that dependency in your local repo but i havent tried it 18:06:47 SumitNaiksatam: oh is that the same that is blocking my devstack patch? 18:07:07 but your devstack patch was failing on the previous day as well, right? 18:07:19 igordcard: let me check 18:07:37 i rechecked yesterday 18:07:41 its fine now 18:07:58 i might have rechecked before this gate breakage 18:08:05 anyway 18:08:16 hemanthravi: still there? 18:08:23 yes, 18:08:34 #topic Design Spec - NFP 18:08:35 have couple of patches that failed. 18:08:40 not sure if it's due to this 18:08:44 hemanthravi: ok, need to check 18:09:06 the wierd thing is that my sanity job passed on our master (which is mitaka( 18:09:17 so perhaps mitaka is capped properly 18:09:32 kilo and liberty are definitely failing for that same reason 18:09:58 hemanthravi: thanks for updating the spec in response to the last set of comments 18:10:24 igordcard: i believe you had +1’ed this earlier, and hemanthravi has responded to your pending comments post that 18:10:30 addressed all the comments so far, let me know if anyging is missing 18:11:03 igordcard: are you okay with the current state of the spec? 18:12:24 probably igordcard has slow connection 18:12:37 i think most of my comments were answered 18:12:57 since this is a large feature, it would have been good to have more details documented 18:13:09 sorry, 18:13:20 however, the spec is place to facilitate the design discussion 18:13:33 I didnt have major issues with it, no 18:13:43 and the implementation details should be captured in the devref documentation 18:13:49 igordcard: okay thanks 18:14:00 rkukura: did you get a chance at all to look at the spec? 18:14:26 hemanthravi: i believe osm has reviewed this spec? 18:14:37 yes, he has 18:14:37 not recently 18:14:45 hemanthravi: and magesh? 18:14:50 yes 18:15:05 hemanthravi: okay can you please get them to comment/vote? 18:15:07 rkukura: okay 18:15:10 will do 18:15:17 will do 18:15:24 songole: ah there you are 18:15:29 have also posted a number of patches for the impl 18:15:35 hemanthravi: right 18:16:05 rkukura: igordcard: at this point my proposal is that we move forward with this spec unless there are any major objections to anything particular 18:16:41 and expect that finer implemetation details will be documented in devref documents that will accompany the impl patches 18:16:45 does that sound okay? 18:16:47 SumitNaiksatam: OK - I will re-review it today if possible 18:17:05 rkukura: okay cool, thanks 18:17:20 rkukura, thanks 18:17:26 SumitNaiksatam: yeah 18:17:34 igordcard: okay, thanks 18:17:50 #topic GBP Devstack Plugin 18:18:07 #link https://review.openstack.org/#/c/300557 18:18:20 hemanthravi: thanks for the NFP discussion 18:18:27 igordcard: thanks for posting this patch 18:18:35 igordcard: i have tested it and it works well! 18:18:59 igordcard: i put a comment about the log confguration 18:19:22 SumitNaiksatam: cool! I'll add some ocumentation and a few todos I have annoted 18:19:33 igordcard: i did notice that the logs were getting populated in /opt/stack/logs by default, but i didnt find the stack.sh 18:19:39 igordcard: hence i put that comment 18:19:58 SumitNaiksatam: alright, I will check and see 18:20:07 its quite possible that i was looking through blurry eyes late in the night, and i might have missed it! :-) 18:20:15 please add it if its required 18:20:44 igordcard: and yeah, TODOs sounds good 18:21:02 hemanthravi: can you or your team take a look at this devstack? 18:21:31 hemanthravi: my thinking is that we should build the NFP devstack (targeted for the summit) on top of this 18:21:33 SumitNaiksatam, igordcard started looking at this 18:21:50 hemanthravi: yeah, the GBP devstack plugin is ready 18:21:51 for the same to use it for nfp too 18:21:58 mean the same approach 18:22:21 hemanthravi: are you asking if igordcard is looking at it, or are you saying that igordcard has already started looking at it? 18:23:10 i meant dhuldev who did the nfp devstack srcript is going to use igordcard patch to modify his script to follow the same approach 18:23:22 hemanthravi: perfect 18:23:59 Has there been any testing of this devstack plugin on CentOS yet? If not, I’ll try that. 18:24:00 hemanthravi: my estimate is that it should be a matter of just moving the scripts 18:24:11 hemanthravi: it's probably only a change of 1 or 2 lines in the devstack plugin, to enable the additional nfp plugin and maybe some env vars 18:24:13 rkukura: no, and that would be awesome! 18:24:25 igordcard: yeah 18:24:27 igordcard, thanks 18:25:01 hemanthravi: we can either merge the devstack plugin now and then you/others create a new one on top, or we can delay this a little a bit and incorporate everything in my current patch 18:25:09 should we merge the nfp script into gbp plugin or keep it separate for now 18:25:11 ? 18:25:11 rkukura: just to be totally clear, i have only tested GBP devstack till date only on Ubuntu 18:25:28 igordcard: no, we first merge your patch 18:25:36 rkukura: no, I've only tried in Ubuntu 14.04 with Liberty cloud archive and stable/mitaka devstack branch 18:26:01 hemanthravi: lets keep the NFP devstack part outside the tree for now 18:26:29 hemanthravi: our you could add it to your patch series 18:26:31 SumitNaiksatam: same here 18:26:33 *or 18:26:38 ok 18:26:39 SumitNaiksatam: about ubuntu 18:26:47 igordcard: right 18:27:23 rkukura: also, igordcard is reusing most of the scripts we already had for the GBP devstack we were installing earlier 18:27:46 rkukura: he has put it into the right plugin hooks, so that we dont have to manually run those scripts and patch 18:27:48 yes yes 18:28:06 rkukura: i say this to highlight that the scripts have mostly been already tested 18:28:25 so there should not be any big alarms here, this is not a totally new devstack 18:28:33 *devstack installation 18:28:44 SumitNaiksatam: right - there was an edit needed with the old scripts for them to work on CentOS, so I’m hoping that will get resolved 18:28:51 rkukura: ah ok 18:29:17 SumitNaiksatam: I just think maybe some testing should be done around customizing other variables in local.conf, that might be necessary for some reason, and seeing if the devstack plugin still works well 18:29:17 igordcard: thanks a ton for getting this done 18:29:27 igordcard: okay 18:29:50 igordcard: i would prefer to merge this first, and then add fixes in follow up patches if required 18:30:04 assuming everyone else in the team is on board with taht 18:30:40 hemanthravi: will appreciate if you, songole, magesh can review/test this devstack plugin 18:30:55 SumitNaiksatam: alright, I'll add basic doc on enabling the plugin today and then it will be ready to merge 18:31:00 will do 18:31:00 since rkukura is testing on CentOS, i need one more core to do this on ubuntu 18:31:04 SumitNaiksatam: I'll also investigate the stack.sh log issue 18:31:09 igordcard: nice! 18:31:20 igordcard: thanks again for the update, and all the work 18:31:34 #topic Design Summit Prep 18:31:46 sorry this was not on the documented agends 18:31:49 *agenda 18:32:00 but since we have a little more time 18:32:37 so regarding the workshop - i shared a high level script with the authors on the workshop 18:33:29 rkukura: i will share with you and others in the team as well, did not want to share it prematurely if there were going to be changes based on initial feedback 18:33:52 ok 18:33:52 one of the things we really need to have heading into the summit is to beef up our devref documentation 18:34:19 so i requested igordcard to take a look a our existing specs and try to see if those can be massaged into devref documents 18:34:39 SumitNaiksatam, started working on the service developer workflow - will try to get that done by tue 18:34:43 as a first pass, mu proposal, was to merely translate the content from the specs to corresponding docs 18:34:49 hemanthravi: nice! 18:35:20 but this is a general call to action to the entire team 18:35:32 igordcard: would you be willing to coordinate this activity? 18:36:43 SumitNaiksatam: I don't know if I have enough time these next weeks to "coordinate" it, but I surely can port at least a part of the specs to devref 18:37:14 igordcard: nice, any progress we can make here, helps 18:37:21 SumitNaiksatam: I can keep in touch with the rest of the folks involved so each one can focus on specific parts 18:38:02 igordcard: right, so if you can let us know which specs you are starting with, others can jump in as well 18:38:20 igordcard: dont have to decide now 18:38:31 igordcard: we can sync up offline on this 18:38:49 yeah, but feel free to start first and tell me about it, as well 18:38:55 igordcard: sure 18:39:22 #topic Open Discussion 18:39:40 rkukura: i am guessing there wasnt much to discuss on the packaging topic, right? 18:39:53 nothing noew 18:39:55 new 18:39:59 rkukura: okay 18:40:06 we need to release mitaka 18:40:44 i am just waiting to get feedback on the drivers, if they are working correctly 18:41:51 anything else for today? 18:42:45 alrighty, thanks all for joining 18:42:51 thanks SumitNaiksatam! 18:42:55 thanks, bye 18:42:56 hemanthravi: and igordcard, i know its late for you, good night! 18:43:01 bye 18:43:05 bye all 18:43:09 bye 18:43:15 #endmeeting