17:02:22 #startmeeting networking_l2gw 17:02:22 Meeting started Mon Aug 3 17:02:22 2015 UTC and is due to finish in 60 minutes. The chair is Sukhdev. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:02:23 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:02:25 The meeting name has been set to 'networking_l2gw' 17:02:42 #topic: Agenda 17:03:08 #link: https://wiki.openstack.org/wiki/Meetings/L2Gateway 17:03:44 Anybody would like to make any changes to the agenda? 17:03:45 Hello everyone 17:03:50 armax: are you here? 17:04:00 yes 17:04:06 #topic: Announcements 17:04:46 Liberty feature freeze is one month away 17:05:35 Hi 17:07:30 Bhooshan: hello 17:07:40 does anybody have anything to announce? 17:08:20 Sukhdev: I have joined l2gateway team 17:08:33 Bhooshan: welcome to the team 17:09:01 Thank you 17:09:09 #topic: meeting time discussion 17:09:22 I sent out an email last week - did not hear any response 17:09:42 marutikamat armax : are you guys here? 17:09:56 I suppose we are 17:09:57 I think it is better to continue with the current timings unless someone has a problem with it 17:10:23 armax: ah ha looks like I woke somebody up :-):-) 17:10:37 Also, given that it is fortnightly meetings 17:10:54 :-) 17:11:25 what does the rest of the team think? 17:11:35 shall we stay with the current time slot? 17:11:42 agree with maruti 17:12:03 vikas_: cool 17:12:12 shall we consider this issue closed then? 17:12:33 if anybody feels strongly, feel free to reply to the email that I sent out last week 17:12:47 sukhdev: yes, for now we can close this issue 17:12:52 :-) 17:13:01 lets move on - couple of critical issues to discuss 17:13:06 phani: thanks 17:13:15 #topic: CI breakage issue 17:13:26 This has been a blocker for a while - 17:13:35 anybody has any update on this? 17:14:21 no 17:14:42 we’re still working through the alembic changes to support out of tree data schema changes 17:14:50 it’ll be a while 17:15:29 Thanks armax for the update 17:16:14 armax: I have seen patches with your comments between Ihar and Henry 17:16:20 ok 17:17:10 armax: any conclusion to that? 17:17:45 not that I am aware 17:18:01 Sukhdev: last time I checked it was late on Thursday and I was out on Friday 17:18:09 Sukhdev: I’ll poke at it later in the day 17:19:01 armax: cool - this will help 17:19:19 armax: I have been holding off reviewing the code because of this 17:19:20 I wish I could be faster these days, but I think I ran out of steam 17:19:41 armax: :-) 17:19:59 so, that was one critical issue on my list 17:20:09 lets move on to the next ones 17:20:17 if someone wants to take over, please have at it 17:20:22 #topic: Bugs 17:20:42 the mother of all changes is here: https://review.openstack.org/#/c/198524/ 17:20:43 We have a few bugs in new state which are unassigned 17:21:00 #link: https://bugs.launchpad.net/networking-l2gw 17:21:18 Myself and Bhooshan are currently working further on https://review.openstack.org/#/c/202495/ 17:21:35 that is bug 1475251 17:21:35 bug 1475251 in networking-l2gw "L2 gateway agent does not populate mcast_macs_remote table entries" [High,In progress] https://launchpad.net/bugs/1475251 - Assigned to Maruti Kamat (marutik) 17:21:57 m working on https://review.openstack.org/#/c/208524/ 17:22:22 bug https://bugs.launchpad.net/networking-l2gw/+bug/1466302 17:22:23 Launchpad bug 1466302 in networking-l2gw "OVSDB Server connection initiation to be configurable" [Undecided,New] 17:22:37 marutikamat: We tested this one, and it seems to be working fine - I posted the comments on the patch 17:23:20 armax: why is that a blocker - I just moved arista repo to openstack - I did not have to add the alembic stuff to make the CI pass 17:23:28 thanks Sukhdev. I will respond to your comments. I will also need to address other use cases which are not covered as part of that review. 17:23:58 armax: it is my understanding that the alembic stuff is needed if one makes changes to the DB tables - for the initial cut, it is not needed 17:24:00 Sukhdev: that’s because you don’t check run check_migration on pep8 17:24:30 armax: Oh - I was not aware of that angle :-) 17:25:19 Sukhdev: I’d give us a few more days to resolve this aspect 17:25:21 armax: I saw your chat with HenryG last week on this subject - I hope the patch goes through and we are past this issue soon 17:25:37 and if we’re still stuck we can perhaps put a stop-gap in place 17:25:46 Sukhdev: even stable is pretty busted right now 17:25:55 Myself and Bhooshan are also working on bug 1478812 17:25:55 bug 1478812 in networking-l2gw "L2gw API framework does not provide a mechanism to support other south bound implementations" [High,In progress] https://launchpad.net/bugs/1478812 - Assigned to Maruti Kamat (marutik) 17:26:03 Sukhdev: and that is also waiting on an infra change 17:26:20 armax: sounds like a plan - glad you are on top of this 17:26:35 Sukhdev: more like under it 17:26:37 but sure 17:26:48 armax: :) 17:27:32 marutikamat: I looked at your patch for 1478812 - 17:28:23 vikas_: Thanks for working on that patch - I looked at that one as well - 17:28:29 Thanks Sukhdev 17:28:36 vikas_: if you can update it with the bug number that will be good 17:28:41 thanks sukdev 17:28:57 vikas_: let me know if I can help with this one - I can test this one for you as well 17:29:00 yes updated the commit message 17:29:31 vikas_: We have setup a test bed specifically for this testing and can provide the feedback as soon as you give me a green light 17:29:59 yes sure sukdev 17:30:18 Preeti is working on moving neutron-l2gw to neutron (python neutron client). She is in touch with armax 17:30:28 yes 17:30:50 marutikamat Preeti: you stole my thunder - I was about to jump to this one :-):-) 17:31:08 :-) 17:31:18 I would love to see Preeti's patch merged in the Liberty - I think this is a good one as well 17:31:29 i'm trying to load extension using entrypoints. 17:31:58 Preeti: correct 17:32:08 but not able to do so. Need suggestion from armando 17:32:33 Preeti: in python-neutronclient, right? 17:32:37 yes 17:33:23 you are good hands if armax is looking it for you 17:33:31 Sukhdev: i am not 17:33:54 armax: some body is being modest here - I wonder who he is? :-) 17:34:04 Sukhdev: I am honestly not 17:34:22 armax: no kidding!! 17:34:24 Preeti reached out to me, but I have no bandwidth left 17:34:47 I don’t know where the blocker is though, people are using client extensions successfully 17:35:09 armando i have resumed from where i had left it last time 17:35:24 armax: let me ask around - I am not familiar with this stuff either 17:35:44 but facing the same problem.. unable to load entrypoints 17:36:11 Preeti: ok can you determine the root cause? 17:36:39 Preeti: an error traceback should hint you where to look 17:36:48 ok 17:37:26 Preeti: if you can capture the traceback and paste it, perhaps we can look at it collectively and try to figure out the next step 17:37:57 ok i'll do that 17:38:33 marutikamat: there are few unassigned new bugs - 17:39:10 Preeti, Sukhdev: most likely the use of debugger would help 17:39:15 Phani is working on some of them 17:39:47 and rest are new features like supporting DVR, software L2gw, etc. 17:39:48 armax: I rely heavily on the debugger - it is hard to do anything without it 17:40:10 sometimes troubleshooting can be diffult by analyizing a static traceback; one would really need to dive in 17:40:18 Preeti: have you tried that? 17:40:23 I am working on couple of low hanging bugs 17:40:33 no i have not tried it 17:40:56 phani: excellent - you may want to assign them to yourselves 17:41:24 sukhdev: yeah, i will do that and post the patchsets in this week 17:41:39 phani: cool - perfect 17:42:14 armax , sukhdev please give me some pointers on how to use debugger 17:42:31 i have never tried it 17:43:19 Preeti: quick and dirty way is to put "import pdb" and then add prb.set_trace() where you would like to break and run the prgram 17:43:50 Preeti: once it breaks there, then you can inspect the context and step through the program to figure out what is going on 17:44:24 Preeti: https://docs.python.org/2/library/pdb.html 17:44:30 Preeti: ask for someone local to help you out 17:44:36 thanks sukhdev. 17:44:36 Preeti: when you are ready, ping me on irc and I can walk you through it 17:44:41 ok armax 17:45:33 ok sukhdev . thanks 17:45:59 I think we have covered most of the critical ones - that I wanted to cover 17:46:08 I have a few questions on supporting a service driver for ODL (south bound). I will send out an email on that topic. 17:46:34 marutikamat: sure - that is a good one BTW - much needed 17:47:10 Yes Sukhdev 17:47:49 marutikamat: let me know if I can help answer any/some of them 17:48:04 Sukhdev: thanks 17:48:48 #topic: Open Discussion 17:49:03 Anybody wants to bring up anything? 17:49:33 BTW, as I mentioned earlier - We have setup a dedicated setup to test L2 GW stuss 17:49:37 stuff 17:49:45 Ashwin is helping me out with that 17:50:03 Sukhdev: Glad to know that 17:50:14 It will help 17:50:21 Whenever you have a critical patch to test, let us know and we will be happy to test it for you 17:50:49 I tested marutikamat's patch that way and would love to test vikas_'s patch as well - as soon as he is ready 17:51:31 anything else? 17:51:35 * Sukhdev waiting 17:52:16 looks like we are done 17:52:20 Thanks folks 17:52:28 bye 17:52:34 thanks everyone 17:52:37 bye 17:52:38 bye 17:52:41 #endmeeting