14:01:02 <xgerman> #startmeeting fwaas 14:01:04 <openstack> Meeting started Tue Apr 4 14:01:02 2017 UTC and is due to finish in 60 minutes. The chair is xgerman. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:06 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:01:08 <openstack> The meeting name has been set to 'fwaas' 14:01:15 <xgerman> #chair SridarK yushiro xgerman njohnston 14:01:16 <openstack> Warning: Nick not in channel: yushiro 14:01:17 <openstack> Current chairs: SridarK njohnston xgerman yushiro 14:01:30 <xgerman> #topic Announcements 14:01:50 <xgerman> So I saw that TC elections are coming up; so vote, vote,vopte 14:02:22 <SridarK_> sorry got bounced 14:02:38 <SridarK_> I have a conflict with another meeting so will lurk around 14:02:42 <xgerman> #chair SridarK_ 14:02:43 <openstack> Current chairs: SridarK SridarK_ njohnston xgerman yushiro 14:03:11 <reedip> o\/o 14:03:19 <xgerman> also P-1 is coming up: #link https://releases.openstack.org/pike/schedule.html 14:03:35 <yushiro> Hi. Sorry for late ;( 14:03:42 <xgerman> #chair yushiro 14:03:42 <openstack> Current chairs: SridarK SridarK_ njohnston xgerman yushiro 14:03:58 <xgerman> no worries, was just going through announcements ;-) 14:04:02 <xgerman> anyhting else? 14:04:04 <reedip> xgerman : whome to vote ??? 14:04:09 <xgerman> no idea 14:04:13 <yushiro> OK, thanks xgerman 14:04:34 <hoangcx> Boston forum ideas? 14:04:55 <xgerman> mmh, nova says they are not very informative 14:05:21 <reedip> who are not informative ? 14:05:24 <reedip> nova? 14:05:35 <xgerman> the forums where you meet with end users/operators 14:06:19 <xgerman> #link https://www.mail-archive.com/openstack-dev@lists.openstack.org/msg103433.html 14:07:05 <xgerman> #link http://forumtopics.openstack.org 14:07:41 <xgerman> I guess any exposure is good so hoangcx if you want to propose something I will show up (if my travel gets approved) 14:08:02 <reedip> oh ok ... 14:08:38 <hoangcx> xgerman I am waiting for travel approval too :-( 14:08:38 <xgerman> moving on… 14:09:15 <xgerman> #topic Pike 14:09:34 <xgerman> #link https://etherpad.openstack.org/p/fwaas-pike 14:10:26 <xgerman> I have seen movement on shared/public - reedip/yushiro? 14:10:31 <reedip> yep 14:10:50 <reedip> xgerman : vks1 already proposed a patch in neutron-fwaas and I proposed in neutron-lib 14:10:59 <xgerman> awesome! 14:11:04 <yushiro> yup. but sorry I am so swamped and not pushed yet into python-neutronclient 14:11:24 <reedip> yushiro : I can help if you want 14:11:24 <xgerman> well, it’s supposed to be osc anyway ;-) 14:11:26 <SridarK_> +1 i think this was a bit of taking a wrong turn early on 14:11:32 <yushiro> vks1, reedip Thank you so much for your great work :) 14:11:35 <SridarK_> at least we have clarity 14:11:43 <SridarK_> now 14:11:50 <xgerman> +1 14:11:51 <reedip> For review : https://review.openstack.org/#/q/status:open+branch:master+topic:bug/1676922 14:12:58 <reedip> xgerman : its OSC layer in NeutronClient :D 14:13:27 <xgerman> LBaaS has started some plugin… 14:14:11 <xgerman> also you need to name things differently, e.g. firewall rule create 14:14:27 <xgerman> so not sure how much work that will end up being 14:14:28 <reedip> xgerman : I know about the octavia client :) 14:14:34 <reedip> I am working on that with Ankur :) 14:14:57 <xgerman> cool, I am just afraid the osc work is more then we anticipate right now 14:15:11 <reedip> xgerman : let me look into the OSC side, but it will take some time, so if someone else wants to take it up meanwhile, loop me in :) 14:17:20 <xgerman> ok, let’s keep an eye on that; yushiro let us know if there is anything we can help you with… 14:17:30 <yushiro> xgerman, Sure. 14:17:32 <reedip> +1 14:17:39 <xgerman> next… L2 support 14:18:07 <chandanc> xgerman, driver or agent ? 14:18:13 <xgerman> both ;-) 14:18:16 <chandanc> :) 14:18:24 <reedip> :P 14:18:38 <chandanc> on the driver side i need to make the co-existance work now 14:19:02 <chandanc> i have started some experiment but will take a little more time 14:19:37 <chandanc> the fwg delete api in driver is dependent on the default FWG implementation 14:19:49 <chandanc> on no progress on the delete part 14:19:59 <xgerman> no worries… I don’t think we need to release soemthing for 14:20:02 <xgerman> P-1 14:20:09 <xgerman> SridarK_? 14:20:23 <reedip> chandanc : any reason for its dependency in the Default FWG implementation ? Just asking 14:20:24 <chandanc> yes, but any one helping with the agent side will be great 14:20:25 <SridarK_> xgerman: yes agree - we have a few things we need to clear up here 14:20:47 <chandanc> reedip, yes, on delet ports need to be assigned to default FWG 14:20:57 <SridarK_> reedip: yes ^^ 14:21:06 <yushiro> chandanc, yes 14:21:20 <reedip> interesting use case 14:21:22 <SridarK_> chandanc: i think we can work thru this while we settle down on this 14:21:29 <reedip> if we do not have default FWG, then ? 14:21:47 <chandanc> SridarK_, please let me know your feedback on the plugin change 14:21:51 <chandanc> SridarK_, sure 14:22:16 <SridarK_> chandanc: will do lets sync on this - sorry i am really swamped at work too 14:22:23 <chandanc> the plugin changes are part of the l2 ext agent patch 14:22:29 <SridarK_> but lets close this either tomorrow or day after 14:22:32 <chandanc> SridarK_, no worries 14:22:34 <reedip> chandanc, SridarK_ :I mean is it necessary to send it back to default FWG ? 14:22:38 <yushiro> reedip, If a user deletes default fwg, new default fwg will be generated automatically 14:22:58 <chandanc> SridarK_, sure 14:22:59 <xgerman> unless we switch it off (as discussed last week) 14:23:17 <SridarK_> reedip: yes the default fwg will be the default if we remove something else configured 14:23:22 <chandanc> xgerman, i have some comments on reedip's mail 14:23:22 <reedip> xgerman : exactly 14:23:27 <SridarK_> xgerman: yes pending that 14:23:36 <chandanc> but i thing i got a bounce from your mail id 14:23:46 <SridarK_> but we need a model when it is not turned off too 14:23:51 <chandanc> s/have/had/ 14:23:51 <reedip> chandanc : I saw your email, but my intention was not to get rid of default FWG but keep it optional for the user 14:24:12 <reedip> IIUC :) 14:24:40 <chandanc> reedip, i agree with your intentions :) only the implementation i have some concerns 14:24:58 <xgerman> yeah, and I can see us switching off = applying a default which allows or denies all 14:25:51 <reedip> lets continue that in the Email itself .. SridarK_ , xgerman , yushiro , @all : your comments are pending :) 14:25:58 <xgerman> ok 14:26:08 <SridarK_> reedip: yes i am guilty - sorry just not had a chance 14:26:16 <SridarK_> but +1 lets discuss on email 14:26:17 <chandanc> sure reedip , i think xgerman mail id was wrong 14:26:17 <yushiro> aha, I'll reply it. sorry for slow 14:26:35 <xgerman> chandanc german.eichberger @ rackspace . com 14:27:00 <reedip> xgerman : silly me :D 14:27:01 <chandanc> xgerman, will forward the thread 14:27:07 <xgerman> thanks 14:27:09 <SarathMekala> xgerman, you may need to update your email id on the etherpad 14:27:33 <xgerman> yep 14:27:42 <SarathMekala> a few of my previous mails bounced off 14:27:47 <SarathMekala> xgerman, thanks 14:28:21 <xgerman> sorry, forgot that update 14:28:30 <reedip> xgerman : did u get it? 14:28:52 <xgerman> yep 14:29:30 <reedip> ok, so where were we ? 14:29:44 <xgerman> well, next would be Horizon & Tempest 14:29:54 <xgerman> any updates? 14:29:59 <SridarK_> lets close the default fwg on email 14:30:04 <xgerman> +1 14:30:07 <chandanc> +1 14:30:12 <SridarK_> and on L2 chandanc lets see what else is pending 14:30:21 <SarathMekala> I am sorry, I was down sick for the last 6 days and could not make much progress 14:30:28 <SarathMekala> Did some catch up today: 14:30:29 <chandanc> SridarK_, sure 14:30:32 <xgerman> hope you feel better! 14:30:38 <SridarK_> +1 14:30:48 <SarathMekala> thanks...am a bit better now 14:30:55 <SarathMekala> 1. Created a project https://launchpad.net/neutron-fwaas-v2-dashboard on launchpad 14:31:04 <SarathMekala> 2. Will also create a docs.openstack.org site for it 14:31:04 <reedip> ohh 14:31:14 <SarathMekala> 3. Generated the cookiecutter code for fwaas and compared it aginst lbaas 14:31:19 <reedip> so we have a new project now ... 14:31:27 <xgerman> nice! 14:31:36 <SarathMekala> yeah.. its an individual project 14:31:49 <SarathMekala> need to support devstack as well 14:32:02 <reedip> okk.... and it would be a part of neutron-dashboard in the future ???? 14:32:09 <SarathMekala> no 14:32:31 <SarathMekala> this is the new path set for all the neutron plugins 14:32:38 <reedip> hmm .. ok... not sure whether its required for the stadium criteria, so we may need to clear that doubt up 14:32:40 <SarathMekala> such as fwaas, lbaas, vpnaas e.t.c 14:32:46 <reedip> okk.... 14:33:16 <SarathMekala> I found some differneces against lbaas and two generaion options were not clear 14:33:31 <SarathMekala> I have a feeling that I can hack around and get the stuff working but not sure if its the right way to go 14:33:34 <SridarK_> SarathMekala: one thought is that we can balance out the repo structure etc along with functionality - so worst case we do have something to demo (even if the code has not merged) 14:33:46 <SridarK_> by Summit time 14:34:00 <SarathMekala> yeah SridarK_ , I can get that done 14:34:02 <SridarK_> but lets do get the repo model right 14:34:05 <xgerman> SarathMekala you cna probably ask the horizon people what tey prefer 14:34:16 <SarathMekala> yes xgerman 14:34:18 <SarathMekala> Had a brief chat with Rob Cresswell.. will send across a detailed mail to him tonight and get things clarified 14:34:30 <xgerman> awesome — he’s the right guy 14:34:37 <SridarK_> and Plan B is that u will be able to hack something that shows users what to expect 14:34:38 <reedip> +1 14:35:14 <SarathMekala> as plan B we can go with the old structure and show case the usage.. it will be fast 14:35:29 <SarathMekala> Plan A is to hack the new model and get it working 14:35:49 <reedip> regarding tempest, I didnt get the time, though I am looking into the fullstack now. I respun the neutron-lib patch of njohnston , and found some issues, have pushed the latest iteration 14:36:30 <xgerman> yeah, from what I have heard tempest is also a moving target since they actively move stuff into te,pest-lib 14:36:39 <SridarK_> reedip: yes this seemed to run into some issues 14:36:49 <reedip> I am targetting completing that patch so that API related changes then remain in neutron-lib itself 14:37:34 <reedip> SridarK_ : right now,the duplicate extensions are an issue 14:38:10 <reedip> but also , this patch is needed : https://review.openstack.org/453110 14:39:12 <xgerman> k, looks sane 14:41:24 <xgerman> moving on: 14:41:25 <xgerman> #topic performance improvement for v2 14:41:54 <cuongnv> I'm still working on that 14:42:32 <xgerman> nice. 14:42:35 <cuongnv> for now I've pushed some updates based on reviews of Cedric, but got a problem with ctypes 14:42:57 <cuongnv> had a short discussion with Cedric too, but we couldn't figure out what happened 14:43:09 <reedip> thats the same issue wrt to the gate's libraries, right ? 14:43:44 <cuongnv> reedip, not really 14:43:53 <yushiro> cuongnv and Cedric Thank you so much. I'm sorry I couldn't review so much in these month.. 14:43:59 <reedip> oh ok, thats a separate issue ?? :D 14:44:18 <cuongnv> yushiro, yeah np 14:44:38 <cuongnv> reedip, i think so, haven't took a deeper look yet 14:44:49 <reedip> ok ... no issues 14:45:09 <cuongnv> so, if anyone familiar with ctypes it'd be great for a help :) 14:45:55 <cuongnv> my next action would be compare it with previous uploaded ps and discuss with Cedric to see what'd be the problem 14:46:40 <cuongnv> that's all for performance improvement atm 14:46:42 <annp_> cuongnv: I have same problem with ctypes, too.:) 14:47:13 <cuongnv> yeah annp_ 14:48:02 <cuongnv> if no one has any comments or questions, we can move on.. 14:48:06 <xgerman> blallau had some patches to speed up the RPC calls and DB queries 14:48:15 <xgerman> but I haven’t seen updates in a while… 14:49:16 <cuongnv> xgerman, was there any problem with ctypes on that too? 14:49:21 <reedip> let me spin up this one https://review.openstack.org/#/c/436978/ 14:49:32 <reedip> remove the merge conflict and lets see where it takes us 14:49:52 <xgerman> cuongnv don’t think so… 14:50:00 <cuongnv> xgerman, ok 14:50:23 <cuongnv> thought you're continue with my problem :) 14:51:07 <SridarK_> xgerman: i think blallau finished his work which should get us some good performance - but lets take a look to see something is pending 14:52:11 <xgerman> well, some of it we can’t merge without improvements — so wondering if we need to adopt the patches or if he will finish 14:52:33 <xgerman> e.g. #link https://review.openstack.org/#/c/428034/ 14:53:50 <reedip> respun this https://review.openstack.org/436978 xgerman 14:53:54 <SridarK_> xgerman: ok 14:53:56 <xgerman> thanks 14:54:31 <xgerman> ok, 6 minutes left… 14:54:49 <xgerman> #topic bugs 14:54:56 <xgerman> we did some cleanup… 14:55:01 <reedip> kk 14:55:37 <cuongnv> I'd like to have your guys reviews on this https://review.openstack.org/436978 14:55:50 <cuongnv> xgerman, thanks for your review and +2 on that 14:56:21 <reedip> cuongnv : whats the issue with 436978 ? 14:56:42 <cuongnv> reedip, it's a fix for a port range bug 14:56:53 <hoangcx> cuongnv wrong link? 14:56:59 <xgerman> yep 14:57:04 <cuongnv> oh my 14:57:06 <reedip> yeah, it is hoangcx :D 14:57:07 <cuongnv> sorry guys 14:57:09 <cuongnv> this one https://review.openstack.org/#/c/443385/ 14:57:10 <reedip> thats what I thought :P 14:57:28 <cuongnv> something wrong with my ctrl+c :D 14:58:05 <xgerman> yeah, we need to do more reviews… 14:58:22 <cuongnv> okay, as it's been there for weeks, could you guys help to take a look? 14:58:24 <SridarK_> 2 min warning :-) 14:58:37 <hoangcx> Yeah. Review velocity so slow, i guess 14:58:58 <yushiro> reedip, I've just added https://review.openstack.org/#/q/topic:bug/1676922 14:59:11 <yushiro> Pls review it. 14:59:15 <hoangcx> Especially from fwaas cores 14:59:31 <SridarK_> hoangcx: will look - just been too swamped 14:59:31 <xgerman> hoangcx we know :-( 14:59:45 <yushiro> hoangcx, indeeeeeeeeed ;( 14:59:53 <hoangcx> :-) 15:00:20 <reedip> over and out 15:00:20 <xgerman> ok, time 15:00:25 <xgerman> #endmeeting