18:31:06 #startmeeting Networking FWaaS 18:31:07 Meeting started Wed Jul 30 18:31:06 2014 UTC and is due to finish in 60 minutes. The chair is SumitNaiksatam. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:31:08 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 18:31:10 The meeting name has been set to 'networking_fwaas' 18:31:30 yes 18:31:38 sumit:hi 18:31:47 badveli: hi 18:32:00 #topic Action items review 18:32:02 sorry was discussing some ting 18:32:33 our pending action item is a logistical one, we had to set up a f2f with the DVR team 18:32:44 SridarK: we were waiting for you to get back 18:33:00 SumitNaiksatam: yes i am good lets figure out a day 18:33:24 i had got two servers for my installation 18:33:36 badveli: great 18:33:52 #action SumitNaiksatam to start a thread in fwaas team to schedule f2f meeting with DVR team 18:33:55 sridar we can try the installation, going through the wiki 18:33:59 badveli: thats great 18:34:05 thanks sumit 18:34:11 SumitNaiksatam: so maybe badveli and i will spend some time on this later this week 18:34:21 badveli: can you hold to that discussion when we bring up that agenda item? 18:34:25 so we will have some more context b4 the mtg with DVR 18:34:29 SridarK: ^^^? 18:34:33 oops sorry 18:35:05 though we can definitely treat badveli’s activity as a pending action item ;-) 18:35:15 SridarK: no worries 18:35:37 SridarK: btw, belated, but welcome back! :-) 18:35:47 SumitNaiksatam: glad to be back :-) 18:35:59 SridarK: the team definitely missed you last week 18:36:05 #topic Bugs 18:36:07 SumitNaiksatam: thx 18:36:21 i did not find anything critical or high that is pending 18:36:24 SumitNaiksatam: quick scan on bugs nothing new 18:36:34 SridarK badveli: did you get a chance to bug scrub? 18:36:36 SumitNaiksatam: yes 18:36:39 SridarK: ah ok 18:36:55 looked at the link, 18:37:03 badveli: ok good 18:37:34 at some point we have to make a push towards get a grasp on how the pending bugs are going to be closed 18:37:54 i am saying in the context of juno-3 18:38:22 SumitNaiksatam: yes once we get our patches moving - we will try to wrap up on some of these bugs 18:38:28 we will have time to get big fixes in even after juno-3 but they will have to only critical fixes 18:38:37 SridarK: that sounds like a good plan 18:39:03 SumitNaiksatam: mostly we have some low priority issues now 18:39:13 SridarK: i was suggesting that we need to track down the folks who are not in this meeting but are owning the patches for those bugs 18:39:16 yes sumit, after the patches we can wrap the bugs 18:39:21 SridarK: yes, thankfully 18:39:30 SumitNaiksatam: let me take a stab at that 18:39:39 SridarK: great 18:39:41 did we get any feed back from sridargaddam 18:39:57 we can essentially split the bugs between us, and follow up with the relevant folks 18:40:04 SridarK: i know you are already doing that 18:40:07 SumitNaiksatam: sounds good 18:40:12 badveli: we did, and he posted a new patch as well 18:40:20 ok, thanks sumit 18:40:37 yes we can split up the bugs and go after the relevant bugs 18:40:59 i believe Sridargaddam needs to post a new rev 18:41:07 to respond to carl_baldwin’s -1 18:41:18 ok anything else on bugs? 18:41:37 SumitNaiksatam: nothing from me - i am not caught up on last week yet either 18:41:41 we were discussing #link https://review.openstack.org/#/c/90575/7 18:41:47 SridarK: ok sure 18:41:59 #topic FWaaS support for DVR 18:42:09 SridarK: badveli yisun: ? 18:42:28 badveli: you already mentioned that you set up the servers, thats great 18:42:28 yes sumit 18:42:36 badveli: so you have a working DVR setup? 18:42:52 SumitNaiksatam: i believe the plan was badveli and i will get together to go over the changes once we have a DVR setup 18:43:36 sumit:i have the two servers , initially it was tough to have the servers i think we needed atleast two. so got it and trying to set up 18:43:36 badveli: so you have a DVR installation? 18:44:05 sumit: i did not had enough servers, just got two servers 18:44:07 badveli: nice, completely appreciate the difficulty in getting the two servers, its not easy with resource constraints and what not 18:44:20 badveli: i am guessing two should be enough? 18:44:26 yes sumit 18:44:35 badveli: SridarK we should check with Swami on this 18:44:52 SumitNaiksatam: yes 2 was recommended by Rajeev 18:45:00 also regarding the f2f meeting, i think its better if we do some amount of home work before we schedule the meeting 18:45:11 SumitNaiksatam: i think once we this moving we can reach out to Swami 18:45:16 that way we can have more informed conversation 18:45:20 SumitNaiksatam: yes that will be beneficial 18:45:21 SridarK: great 18:45:51 sridar: i think once we set up the things and start playing with it we might get more idea 18:46:06 SumitNaiksatam: badveli ideally once we have DVR we can look at the FWaaS failure cases 18:46:13 as a first step 18:46:17 badveli: SridarK so you plan to sync up this week? 18:46:25 yes we will 18:46:27 yes sumit 18:46:32 ok great, thanks 18:46:43 also do we have any dependencies on the DVR team? 18:46:56 if so, we can start tracking them here 18:47:18 SumitNaiksatam: good question - while doing the spec 2 or 3 patches were yet to merge 18:47:29 SumitNaiksatam: will follow thru on that 18:47:30 SridarK: ah ok 18:47:43 i believe the suggestion was that we used the private repo 18:47:48 for that reason 18:47:55 Sumit: patches from dvr, should we go through them to understand more, there were some couple of changes we might need to understand more 18:47:56 SumitNaiksatam: yes that is correct 18:48:20 SumitNaiksatam: particularly the L3 agent was still in review but i need to check now 18:48:40 ok 18:49:03 SumitNaiksatam: my sense is that our changes are going to be in the Agent and the driver - we will try to scope this out to get a sense of the scope 18:49:18 SridarK badveli can we create a new wiki page for FWaaS DVR support and link it from the FWaaS wiki page? 18:49:32 SumitNaiksatam: sure will do 18:49:52 yes sumit, sridark we can add details independently 18:49:52 badveli: and i can discuss this 18:50:02 on that wiki page, lets list provide links to our patches and to all the dependencies 18:50:05 fine sridark 18:50:21 SumitNaiksatam: sounds good 18:50:33 once this is done, lets send an email to the mailer, so that its known to all what we are working on, and what are dependencies 18:50:55 SumitNaiksatam: ok 18:51:19 #action SridarK badveli to setup FWaaS DVR support wiki page, provide spec/impl/dependency links, send email to -dev with pointer to this wiki page 18:51:22 sridark is already on top of it, we can add it 18:51:30 badveli: great 18:51:41 anything else to discuss on this topic? 18:51:44 any blockers here? 18:51:45 thanks sridark and sumit 18:52:01 SumitNaiksatam: no blockers seen as of now 18:52:11 the scope of the code changes 18:52:16 SumitNaiksatam: i am sure something will come up 18:52:32 SridarK: badveli thanks for the update 18:52:39 #topic Service Objects 18:52:55 i am working on the reference implementation 18:53:08 #link https://review.openstack.org/#/c/67784/ 18:53:25 ah seems like i have the wrong link 18:53:26 sumit 18:53:34 badveli: can you post the link to your patch here 18:53:38 i am tryiing to update the link in the wiki 18:53:47 not able to update it 18:53:58 badveli: was just going to say 18:54:09 everyone please keep: #link https://wiki.openstack.org/wiki/Neutron/FWaaS/JunoPlan updated 18:54:23 badveli: you might have not logged in before you tried to update 18:54:36 badveli: else i dont see any reason why you would not be able to update 18:55:00 https://review.openstack.org/#/c/105873/ 18:55:15 not sure i tried some time back 18:55:19 i will update again 18:55:51 https://review.openstack.org/#/c/106274/ 18:56:25 https://review.openstack.org/#/c/106918/ 18:56:29 three patches 18:56:41 the reference implementation is not yet done 18:56:55 i am working on it and would be finishing sooner 18:57:04 badveli: good 18:57:05 badveli: i have not looked yet either - will do so 18:57:16 thanks sridark and sumit 18:57:21 badveli: sorry for having to take the drastic step of -2ing the patch 18:57:30 no problem 18:57:38 i was a bit confused 18:57:52 badveli: if i did not, someone else would have done it for that reason 18:57:52 on the process, for some reason 18:58:03 ok, no problem sumit 18:58:07 badveli: also your second patch needs to be dependent of the first one 18:58:20 badveli: the process is documented in the gerrit workflow 18:58:21 there was some last minute updates that happened 18:58:37 thanks sumit 18:58:51 badveli: you shoud have been able to search this on google, but here is a ready reference: #link https://wiki.openstack.org/wiki/Gerrit_Workflow 18:59:02 yes sumit the second is dependent on first 18:59:07 badveli: also, i had earlier sent you a link on how the commit messages need to be formatted 18:59:25 badveli: please follow those guidelines, as i still see some descrepancies 18:59:35 sumit, i followed it for the second patch 18:59:40 badveli: i know this is not as much technical, but it does turn off reviewers 18:59:58 yes sumit, the first and last patch were uploaded before 19:00:07 i will also correct them 19:00:10 badveli: for one, “Implements: blueprint fwaas-customized-service for customized service” is not a suggested a commit message title 19:00:36 badveli: ah i see, you updated https://review.openstack.org/#/c/106274/ 19:00:39 badveli: good 19:00:49 yes sumit 19:00:53 Add service group as a firewall customized service 19:00:57 badveli: i can also help offline if u need - SumitNaiksatam has explained a lot of this to me earlier on the dependencies etc 19:01:09 the second one was updated after you send the link 19:01:27 i am adding the reference implementation so will update the rest of the two 19:01:44 but i am a bit not sure how does all this patches work 19:01:46 badveli: great, i also notice you added the UTs in the first patch 19:01:56 badveli: i have removed my -2 19:02:07 thanks sumit 19:02:13 initially at the last minute 19:02:17 badveli: please go ahead, why do you say - “i am a bit not sure how does all this patches work" 19:02:33 yes, you mentioned the first patch 19:02:43 is cli and it will go in after the patches in 19:02:49 neutron 19:03:10 sumit: you had mentioned it will be reviewed after the patches in neutron 19:03:29 badveli: you mean the client patch? 19:03:32 for example i will try to get the reviewers in second patch 19:03:46 yes, the cli part 19:03:57 if some reviewers approve second patch 19:04:04 badveli: yeah, there is no way that the CLI will be approved and merged before the neutron patches 19:04:21 badveli: because it does not make sense to have CLI for a feature that does not exist 19:04:36 badveli: typically the rewiew happens in parallel 19:04:57 badveli: but you will recieve more attention on the CLI patches once its clear that the neutron patches are almost ready 19:05:28 fine sumit: how does this dependency is explicit 19:05:29 badveli: but you do have to keep the CLI patch ready, since people wil ltry to install devstack and test the neutron patches using your CLI patches 19:05:44 badveli: you cannot setup a dependency across projects 19:06:03 i meant to say if i get some approvals for the second and third patch 19:06:05 badveli: which means, in this case, you cannot setup a dependency from your CLI patch on your neutron patch 19:06:18 oh..thanks sumit 19:06:25 badveli: i am a little cofused as to which are your second and third patches 19:06:33 badveli: lets not refer to them that way 19:06:44 does this mean we cannot ask the reviewers for reviewing the cli 19:06:47 badveli: i believe you have one CLI patch and two neutron patches 19:06:51 ok 19:06:52 yes 19:06:55 badveli: i already answered that 19:07:13 badveli: you can ask, and you should, but realistically you will not get much attention 19:07:38 fine sumit, i will actively work on the neutron patches 19:07:46 for the reviews 19:07:57 to happen 19:07:59 badveli: the way you want to go about is by saying - here are the neutron patches, please review them; and here is the CLI patch which will help you to review those 19:08:23 got it sumit 19:08:30 badveli: that way you point to both these sets of patches, with due priority given to the neutron patch 19:09:05 fine sumit 19:09:25 also the wiki link has some tempest and horizon 19:09:30 another suggestion, and pretty for everyone here - we can assign some identifier names to our patches 19:10:03 for example in the group policy work, i go GP-API-1, GP-DB-1,…, GP-API-2, GP-DB-2, etc 19:10:18 that way it becomes easier to refer to patches when we are talking 19:10:28 fine sumit 19:10:29 these can be mentioned in the commit message 19:10:49 you can see the group policy patch: #link https://review.openstack.org/#/c/95900/20 19:11:04 this is just a personal preference, there is no precedence in neutron for this 19:11:13 so feel free to ignore 19:12:21 fine sumit 19:12:39 anything else to discuss on service objects? 19:12:52 my fingers are aching now! :-( 19:12:54 the wiki link is also capturing about the heat 19:13:15 sumit, i think we are done but last one 19:13:24 the wiki link has tempest horizon 19:13:25 badveli: yeah sure 19:13:27 SumitNaiksatam: :-) we need voice to text convertors for IRC 19:13:29 badveli: yes 19:13:36 SridarK: :-) 19:13:42 badveli: very good point to bring up 19:13:57 wikik has horizon,tempest,heat 19:13:59 badveli: we need the tempest API patch to be in review as well 19:14:09 ok, this is what i need 19:14:15 badveli: the neutron patch will not merge without the tempest API patch 19:14:25 badveli: let me hook you up with miguel 19:14:32 thanks sumit 19:14:40 appreciate your help 19:14:44 badveli: and ideally we should have a tempest scenario patch as well 19:15:22 i will go through it, but the scenario patch is compulsary 19:15:23 #action SumitNaiksatam to reach out to mlavalle regarding tempest patches for upcoming fwaas features, cc rest of the fwaas team 19:15:24 ? 19:15:37 badveli: lets check with mlavalle 19:15:44 thanks sumit 19:15:55 badveli: but even if its not, its better to have it, since it wil strengthen our case 19:16:24 badveli: lets do all we can (assuming we have the time) so that our reviews get smoother 19:16:24 fine sumit, let me go through it 19:16:39 yes sumit 19:16:45 #topic FWaaS metering 19:16:51 is prad here? 19:17:10 SridarK: i believe prad got his spec approved in ceilometer 19:17:12 SridarK: thats great 19:17:23 SumitNaiksatam: oh great 19:17:33 SumitNaiksatam: i will sync with prad 19:17:36 if prad, needs any help we are here to help 19:17:46 #topic open discussion 19:17:51 SumitNaiksatam: to make sure on any requirements on neutron side 19:17:53 #undo 19:17:54 Removing item from minutes: 19:18:01 SridarK: yes 19:18:11 #topic Vendor blueprints 19:18:22 natarajk: SridarK: anything you want to bring up? 19:18:24 SumitNaiksatam: sorry done with what i had to say on the pref topic 19:18:33 *prev 19:18:59 SumitNaiksatam:Nothing much. We are fixing our CI server to l3 plugin. 19:19:12 to -> for 19:19:21 natarajk: ah ok, sorry i havent been able to get to it 19:19:35 natarajk: review i mean, i have a big stack 19:19:42 SridarK: anything at your end? 19:19:45 SumitNaiksatam: not much to report on our side - we need the vendor dependent patch to merge 19:20:05 SumitNaiksatam: we will have the FWaaS piece ready 19:20:54 SridarK: sweet 19:21:01 #topic open discussion 19:21:21 SumitNaiksatam: will reach out to u and s3wong to discuss service insertion 19:21:31 gduan: are you tracking the flavor impleemenation from enikanorov__? 19:21:41 SridarK: sure, we are a bit behind on that 19:21:59 SumitNaiksatam: will discuss over email 19:22:26 badveli: can you check with gduan? 19:22:34 fine sumit 19:23:24 badveli gduan: we need to keep track since we might not get a whole lot of time to react once the spec is approved 19:24:53 Hi 19:25:07 gduan: ^^^ 19:25:09 I am actually fine with the patch 19:25:14 ok good 19:25:16 but spec is not approved yet 19:25:21 gduan: yeah 19:25:38 gduan: we have to be ready on the fwaas side to jump in when the spec is approved 19:25:48 gduan: probably this weekend 19:25:53 anything else to discuss? 19:25:55 sure 19:26:04 the patch is the data model patch 19:26:23 gduan: okay 19:26:40 the implementation is not submitted yet, 19:27:05 I mean enikanorov's patch 19:27:06 gduan: for what? 19:27:31 enikanorov's current patch is for data model only 19:28:03 I guess dispatching part will be another one 19:28:23 gduan: ah ok, lets check with him, on provide comment on the patch 19:28:36 gduan: glad that you have been tracking ;-) 19:28:44 okay if nothing else, lets wrap up 19:28:50 ok. I will discuss with him. 19:28:57 gduan: thanks 19:28:59 thanks all! 19:29:00 bye 19:29:02 bye 19:29:04 thanks all 19:29:04 #endmeeting