14:01:01 <yushiro> #startmeeting fwaas 14:01:01 <openstack> Meeting started Tue Sep 26 14:01:01 2017 UTC and is due to finish in 60 minutes. The chair is yushiro. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:03 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:01:06 <openstack> The meeting name has been set to 'fwaas' 14:01:11 <mlavalle> o/ 14:01:28 <annp_> hi mlavalle, 14:01:54 <yushiro> #chair xgerman_ yushiro 14:01:54 <openstack> Current chairs: xgerman_ yushiro 14:01:57 <mlavalle> annp_: hi there 14:02:09 <annp_> mlavalle: thanks for neutron sticker. :) 14:02:24 <yushiro> mlavalle, hi 14:02:26 <mlavalle> annp_: nice! I am glad youot it 14:02:32 <annp_> mlavalle: it looks very cool. 14:02:34 <SridarK> Hi FWaaS folks 14:02:41 <yushiro> #chair SridarK 14:02:42 <openstack> Current chairs: SridarK xgerman_ yushiro 14:02:50 <SarathMekala> hi SridarK 14:02:53 <SridarK> shall we get started 14:02:58 <yushiro> yes 14:03:03 <SridarK> yushiro: my turn today ? 14:03:10 <xgerman_> o/ - sorry might be a bit distracted… 14:03:20 <SridarK> xgerman_: np 14:03:27 <yushiro> SridarK, Maybe yes :) 14:03:30 <SridarK> ok 14:03:33 <xgerman_> +1 14:03:49 <SridarK> ok lets dive right in 14:04:23 <SridarK> i think we have discussed the PTG priorities quite a bit in detail over the last mtgs 14:04:55 <SridarK> #topic L2 Support 14:05:16 <SridarK> yushiro: pls go ahead 14:05:26 <yushiro> OK 14:05:50 <yushiro> Now, we've removed these 'dependency' for each l2 patches. 14:05:52 <SridarK> #link https://review.openstack.org/#/c/323971/ 14:06:22 <yushiro> SridarK, thanks 14:07:14 <yushiro> Now, in l2-agent patch, I have to fix some minor bug and add UTs. Sorry, I couldn't work last week due to other tasks... 14:07:15 <doude> Hello 14:07:30 <yushiro> There is no blocking issue in this patch. 14:08:02 <yushiro> In order to move forward, we should fix & merge https://review.openstack.org/#/c/504324 14:08:21 <yushiro> Is Inessa here? 14:08:44 <SridarK> yushiro: agreed - i was looking at this too and it looks good to me 14:09:01 <yushiro> I, annp_ and Inessa is discussing 'what noop driver should inherit' 14:09:14 <SridarK> i think we can clarify the usage of the L2 driver base class 14:09:44 <xgerman_> I think we should have unit tests — 14:09:59 <xgerman_> but otherwise great ;-) 14:10:09 <yushiro> SridarK, If so, is it OK to inherit just 'object' and describe usage in docstring? 14:10:55 <SridarK> hmm yes - i was wondering too on that instead of having another version 14:11:05 <SridarK> once we clarify that i think we are good 14:11:31 <SridarK> let me also perhaps make a comment on gerrit 14:12:05 <yushiro> SridarK, OK. And, this approach is temporary. Finally, we should improve neutron's side (Inessa's patch in neutron). 14:12:18 <xgerman_> yep 14:12:27 <SridarK> i see annp_ latest response 14:12:33 <SridarK> and that makes more sense 14:12:51 <xgerman_> +1 14:13:24 <yushiro> OK, so, annp_ do you have bandwidth to add UT for this patch? 14:13:54 <SridarK> ok lets close that conversation on gerrit and we can get this in quickly 14:14:18 <yushiro> Yes, If annp_ or I added UTs in this patch, please review it ASAP :) 14:14:19 <annp_> ok, I will update it in tomorrow. 14:14:29 <yushiro> Thank you. 14:14:33 <SridarK> annp_: great thx 14:14:37 <annp_> No problem. 14:14:51 <yushiro> OK, SridarK , is it OK to continue default fwg patch? 14:15:26 <SridarK> yushiro: ok lets get to that 14:15:41 <SridarK> #link https://review.openstack.org/#/c/425769/ 14:15:59 <yushiro> Thanks Inessa and reedip for this work. 14:16:20 <SridarK> +1 14:16:31 <annp_> +1 :) 14:17:11 <SridarK> In terms of logistics - shall we focus on getting in the L2 agent patches and driver in first ? 14:18:05 <yushiro> SridarK, Yes. sorry for confusing. l2-patch is the highest priority. 14:19:09 <SridarK> yushiro: no no no confusion just stating for the record 14:19:21 <SridarK> and we need to discuss default FWG as well 14:19:25 <yushiro> I mean ... https://review.openstack.org/#/c/447251 and https://review.openstack.org/#/c/323971 and https://review.openstack.org/#/c/504324 14:19:26 <xgerman_> +1 14:19:40 <SridarK> ok perfect 14:20:03 <yushiro> So, let's focus on these 3 patches :) 14:20:08 <SridarK> yushiro: if nothing more on the agent, shall we move to the driver 14:20:10 <SridarK> yushiro: +1 14:20:14 <yushiro> Ok 14:20:22 <yushiro> plz go ahead 14:20:35 <SridarK> ah chandanc not around ? 14:21:37 <SarathMekala> he said he will join 14:21:45 <SridarK> SarathMekala: ok thx 14:22:05 <SarathMekala> was mentioning that the meeting has been shifted to Wednesday 14:22:22 <yushiro> annp_, Did u discuss with Inessa about driver's patch ? 14:22:51 <yushiro> annp_, I'd like to know some blocking issue on this. 14:23:07 <SridarK> SarathMekala: ok no we have not effected that yet 14:23:22 <SridarK> #link https://review.openstack.org/#/c/447251/ 14:23:35 <SarathMekala> SridarK, ok 14:24:13 <SridarK> SarathMekala: but perhaps some confusion - we can come back to this in case he will take some time to join 14:24:30 <annp_> yushiro, I haven't discussed with her yet. 14:24:52 <yushiro> annp_, Ah, I mean the last comment for this patch: https://review.openstack.org/#/c/447251 14:25:03 <SridarK> also as we work on merging the L2 support related patches, lets work thru some testing 14:25:08 <annp_> let me see. 14:25:08 <SridarK> https://etherpad.openstack.org/p/fwaas-v2-l2 14:25:26 <SridarK> we can use this to capture test cases and also results 14:25:27 <SarathMekala> Called up Chandan.. he is in a release meeting it seems.. will sync up with you on mail 14:25:36 <SridarK> SarathMekala: ok no worries 14:25:51 <annp_> because we'll come up with local vlan idead 14:26:12 <annp_> so I think we can handle exception at L2 agent side 14:26:18 <annp_> not at driver side. 14:26:58 <annp_> However, i'm not sure which way is better 14:27:49 <annp_> there, i'd like to get more opinion from folks. 14:27:58 <yushiro> annp_, Yes, previous approach was that. I implemented to get vlan_id from network in l2-side 14:28:30 <SridarK> annp_: i will need to look at the comments 14:28:54 <yushiro> I and chandan discussed and realized that there is race condition to get vlan tag from OVS in driver layer. 14:29:41 <SridarK> yushiro: when a port has been created ? 14:30:08 <yushiro> So, I'd like to try to revert previous approach (getting VLAN id from l2-agent side by using local vlan manager) 14:30:18 <yushiro> SridarK, Yes, this is the timing. 14:30:35 <SridarK> yushiro: ok 14:30:58 <yushiro> SridarK, Previously, after handle_port() has been called, we tried to associate default fwg to that port. 14:31:39 <yushiro> SridarK, In this timing, we couldn't refer VLAN tag from OVS because OVS tag has not been created at that moment. 14:31:51 <SridarK> hmm ok 14:32:12 <SridarK> the workflow will be similar with SG correct ? 14:32:50 <yushiro> Yes, correct. 14:33:03 <yushiro> I'll integrate these patches and try to use it. 14:33:10 <SridarK> yushiro: ok 14:33:12 <xgerman_> mmh, sounds good 14:33:17 <annp_> +1 14:33:26 <SridarK> lets use https://etherpad.openstack.org/p/fwaas-v2-l2 14:33:37 <SridarK> to coordinate as we did for the dashboard 14:33:42 <yushiro> There are several modifications for previously. 14:33:52 <SridarK> lets make sure to do some testing before we merge the patches 14:34:04 <yushiro> Yes, I'll describe local.conf for testing. 14:34:14 <SridarK> yushiro: ok perfect 14:34:33 <SridarK> on the driver, annp_ u will be helping out chandanc as well ? 14:34:53 <annp_> SridarK. yes. Sure. 14:35:06 <annp_> SridarK: my pleasure. :) 14:35:07 <SridarK> annp_: many thx much appreciated 14:35:36 <annp_> you're welcome.\ 14:35:46 <yushiro> annp_, thanks.. 14:36:01 <SridarK> So quick summary of plan is to start kicking the tires on some integration testing for L2 support 14:36:12 <yushiro> Yeah. 14:36:20 <yushiro> ah, I forgot I thing. 14:36:30 <SridarK> yushiro: yes pls go ahead 14:36:52 <yushiro> SridarK, xgerman_ I pushed in-tree docs patch : https://review.openstack.org/#/c/505326/ 14:37:01 <SridarK> yushiro: oh yes 14:37:17 <SridarK> I wanted to ask too 14:37:43 <SridarK> do we want to get this in as a place holder with just the directory structure ? 14:38:33 <SridarK> and were u going to target some doc changes as part of the L2 patches ? 14:38:49 <yushiro> SridarK, Hmm, sorry I'm not sure the effect for place_holder in doc directories. 14:39:07 <SridarK> yushiro: ok lets discuss offline 14:39:14 <yushiro> SridarK, OK. 14:39:39 <SridarK> #topic FWaaS Dashboard 14:39:50 <SridarK> SarathMekala: amotoki: pls go ahead 14:40:17 <SarathMekala> the last build was failing on pep8 and one UT failure.. I have fixed them 14:40:27 <SridarK> #link https://review.openstack.org/#/c/475840/ 14:40:54 <SridarK> SarathMekala: I have not been able to redo the tests on the last set of patches 14:41:24 <SridarK> SarathMekala: If there are no known issues - i will run some tests today 14:41:40 <yushiro> SarathMekala, Now I'm running devstack with you latest patch 14:42:02 <SarathMekala> SridarK, I think all the known issues are fixed.. amotoki and yushiro can confirm 14:42:12 <SridarK> SarathMekala: ok 14:42:38 <SridarK> and IIRC there is one issue from native Horizon 14:42:54 <yushiro> SarathMekala, OK I believe it :) My concern is creating firewall_policy with rule. 14:43:27 <SarathMekala> yushiro, what exactly is the problem you are facing? 14:44:26 <SridarK> Also lets communicate the dashboard etherpad: https://etherpad.openstack.org/p/fwaas-v2-dashboard 14:44:28 <yushiro> SarathMekala, maybe viewing is broken for 'drag & drop' firewall_rule to apply/remove. 14:44:34 <SridarK> so we can all be in sync 14:44:50 <yushiro> Anyway, i'll test and comment on your patch, SarathMekala 14:44:53 <SarathMekala> yushiro, I did not implement drag and drop :) 14:45:08 <SarathMekala> instead, I have provided checkboxes for selecting the rules 14:45:09 <yushiro> Just deployed :) 14:45:19 <SridarK> ok 14:45:39 <yushiro> SarathMekala, Oh, really? I'll check it. Let's sync after meeting. 14:46:03 <SarathMekala> yushiro, sure.. even amotoki had a question on this area earlier 14:46:35 <SridarK> ok good we can sync on etherpad and gerrit - i think we are almost there 14:46:47 <SarathMekala> yeah.. sure 14:46:48 <SridarK> hopefully we can merge this in the next few days 14:46:54 <xgerman_> +1 14:46:58 <SarathMekala> I am hoping so :) 14:47:01 <SridarK> ok lets move on 14:47:04 <yushiro> yeees!! 14:47:28 <SridarK> #topic Pluggable backend driver 14:47:45 <SridarK> i saw doude join and we promised to give some time for this topic 14:47:57 <doude> :) 14:48:05 <SridarK> #link https://review.openstack.org/#/c/480265/ 14:48:09 <SridarK> doude: pls go ahead 14:48:20 <doude> not to much to say, just the patch is there and passed the CI since more 2 weeks now 14:48:27 <SridarK> doude: ok 14:48:51 <SridarK> i think once we get the L2 changes in - we can focus more on this patch 14:49:21 <doude> but I think I'll have a huge rebase work when all l2 stuff will land 14:49:28 <SridarK> one thing to think about - how much impact these current changes will have 14:49:35 <SridarK> exactly u read my mind :-) 14:49:44 <doude> :'( 14:50:14 <SridarK> yes so pls bear with us - i think the L2 changes should start landing in the next 2 weeks - i hope 14:50:22 <doude> the principal thing I do in that patch is to reorganize the code and propose a driver interface 14:51:40 <SridarK> the impact may not be that bad - considering that the L3 and most of the plugin side of things are not majorly impacted by L2 support changes 14:53:32 <doude> yes but the bad thing actually in the fwaas V2 code is the db code is not clearly separated from the plugin 14:53:54 <xgerman_> so we should change that? 14:54:02 <SridarK> doude: yes that i agree - we have just morphed what we started with 14:54:16 <yushiro> doude, Just a question. Is your patch enable to handle pre/post for v2 resources by implementing driver? 14:54:23 <doude> and plugin inherits from the DB mixin 14:54:31 <doude> no 14:54:34 <SridarK> doude: yes exactly 14:54:56 <SridarK> so we will plan for some refactoring 14:54:59 <doude> we did that in bgpvpn 14:55:03 <doude> project 14:55:34 <xgerman_> ok, we should out it into LP then ;-) 14:55:42 <xgerman_> put 14:55:49 <SridarK> ok lets do some thinking along those lines as well, yushiro to ur point above ^^ 14:56:24 <yushiro> OK 14:56:31 <SridarK> xgerman_: lets discuss this a bit so we can articulate it well 14:57:05 <SridarK> ok good - doude let us know if u need anthing else to be discussed and we can make time here or on the channel 14:57:05 <xgerman_> yes. 14:57:15 <SridarK> #topic Open Discussion 14:57:23 <mlavalle> o/ 14:57:26 <mlavalle> hey 14:57:29 <SridarK> mlavalle: hi 14:57:33 <xgerman_> hi 14:57:34 <doude> we should discus and decide before I continuing to work on it 14:57:43 <SridarK> so folks last week i had a chance to meet with mlavalle who was in the area 14:57:56 <SridarK> we had a nice discussion on where things are with the project 14:58:00 <SridarK> mlavalle: pls go ahead 14:58:02 <mlavalle> yeah, nice conversation and lucnh 14:58:15 <SridarK> we only have 3 min 14:58:32 <mlavalle> just wanted to share with the team that we are looking to work on adding logging to FW 14:58:52 <mlavalle> and also explore further https://review.openstack.org/#/c/236840/ 14:59:19 <mlavalle> I just wanted to give the team visibility 14:59:26 <mlavalle> for the time being 14:59:29 <xgerman_> thx 14:59:31 <mlavalle> that's all :-) 14:59:39 <SridarK> thx mlavalle 14:59:46 <yushiro> Thanks mlavalle !! 14:59:59 <SridarK> we have some synergy on the logging work done by yushiro and folks as well for SG 15:00:10 <xgerman_> +1 15:00:19 <SridarK> mlavalle: thx for dropping in :-) 15:00:25 <xgerman_> time — 15:00:28 <yushiro> yeah, thanks 15:00:30 <mlavalle> thanks for having me :-) 15:00:56 <SridarK> ok folks thx for joining and have a great week 15:01:01 <xgerman_> +1 15:01:03 <SridarK> #endmeeting fwaas