15:00:10 <mlavalle> #startmeeting neutron_routed_networks 15:00:10 <openstack> Meeting started Tue Mar 15 15:00:10 2016 UTC and is due to finish in 60 minutes. The chair is mlavalle. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:12 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:14 <openstack> The meeting name has been set to 'neutron_routed_networks' 15:00:20 <mlavalle> hi 15:00:23 <xiaohhui> hi 15:00:23 <reedip__> hi 15:01:05 <mlavalle> #topic Announcements 15:01:32 <mlavalle> As you know, Neutron RC1 is over or very close to over 15:01:46 <mlavalle> any other annoucements from the team? 15:02:01 <reedip__> mlavalle : There are some bugs still left, but armax said it woud be over by tomorrow 15:02:35 <mlavalle> yeah, I knew it was around today, but not the exact date. Thanks for the clarification :-) 15:02:49 <mlavalle> any other annoucements? 15:03:01 <reedip__> not from my end... anyone else? 15:03:16 <mlavalle> ok, let's move on 15:03:29 <mlavalle> #topic Progress 15:04:00 <mlavalle> reedip__: I think you are working on the client, right? Any progress since last week? 15:04:22 <rtheis> mlavalle: I have a couple patches for OSC client 15:04:27 <reedip__> mlavalle : I have put up a WIP fo neutronclient 15:04:33 <rtheis> https://review.openstack.org/#/c/290054/ 15:04:35 <reedip__> and rthies put up some for OSC 15:04:36 <rtheis> https://review.openstack.org/#/c/290138/ 15:05:07 <reedip__> for Neutronclient: https://review.openstack.org/#/c/290267/ 15:05:30 <mlavalle> I assume reviews are requested for those patchsets, correct? 15:05:44 <reedip__> I was thinking of adding the name/description to the segment CRUD : https://review.openstack.org/#/c/284440/6 15:05:49 <rtheis> mlavalle: reviews are welcome 15:06:03 <reedip__> Sorry, wrong link : http://paste.openstack.org/show/490455/ 15:06:13 <rtheis> But maybe after reedip__ adds name/description to segment CRUD 15:06:21 <reedip__> rtheis : ^^ something to check 15:07:07 <reedip__> I have the PS ready but just a tad bit cautious of changing the provider_d b 15:07:10 <rtheis> reedip__: thanks 15:07:34 <mlavalle> reedip__: is that a change to Carl's patchset? 15:07:43 <reedip__> mlavalle: yup 15:08:27 <reedip__> adding the name/desc parameter to the segment CRUD 15:08:32 <mlavalle> I think he made a comment allowing for changes to low hanging fruit, right? 15:09:13 <reedip__> mlavalle: yup , so I will publsh the PS tomorrow. I also have to create a Migration script 15:09:17 <reedip__> 9 15:09:20 <reedip__> 9 15:09:29 <reedip__> ( attached in the paste link above) 15:09:35 <rtheis> thank you reedip__ 15:10:16 <reedip__> as the provider_db has the NetworkSegment class and it needs two new columns for Name/Description ( sorry , the keyboard got stuck... ) 15:10:23 <rtheis> Looks like line 79 in paste link has a copy/paste error (Add ip_version to AddressScope) 15:11:08 <reedip__> Yes, it was an already created file, which I just modified for devstack to work 15:11:21 <rtheis> ok 15:11:30 <reedip__> The migration script was not being generated ( some problem with neutron-db-manage) 15:11:44 <reedip__> I will take a new checkout of neutron tomorrow and fix that 15:11:56 <reedip__> anything else for progress? 15:12:02 <mlavalle> on my side, I have been working on rebasing https://review.openstack.org/#/c/285548/ 15:12:40 <mlavalle> it is not such a simple rebase because modules were deleted from the base patchset 15:13:03 <mlavalle> so I am moving code around to the new module 15:13:14 <mlavalle> I expect to push the revision later today 15:13:24 <reedip__> ok , great ... 15:13:42 <mlavalle> I will keep on eye on reedip__ changes becuase I think those will impact my patchsetr 15:13:56 <mlavalle> I also had a good follow up with the nova team 15:14:46 <reedip__> anything important to share with us mlavalle? 15:14:47 <mlavalle> specifically, I had a conversation with edleafe about nova resource pools https://review.openstack.org/#/c/253187 15:15:30 <mlavalle> and how it impacted by the roued networks spec on the nova side: https://review.openstack.org/#/c/263898 15:15:54 <mlavalle> so he made already comments to both specs based on this conversation 15:16:22 <mlavalle> I will also review both specs between today and tomorrow and will do my best to keep them in synch 15:16:46 <mlavalle> I also intend to attend the weekly nova meeting this week to draw attention to all this 15:17:02 <mlavalle> that's all I have as far as progress this week 15:17:08 <mlavalle> anyone else? 15:17:11 <xiaohhui> For the action in last meeting "Work on plugin interface for host mapping", I have some intial thoughts and send out a mail for discussion. But it is too early to make patch for it now. 15:17:22 * neiljerram arrives late - sorry all! 15:17:50 <mlavalle> xiaohhui: yeah, I saw the email this morning. Thanks. I gave it a first read and looks good to me 15:18:02 <xiaohhui> thanks 15:18:25 <rtheis> I have also been investigating the steps to allow neutronclient to support OSC plugin since the home for the CLI may move from OSC core to OSC plugin 15:18:47 <rtheis> Which is related to my previous WIP patch sets for the CLI 15:19:25 <mlavalle> rtheis: thanks! 15:19:43 <mlavalle> any thing else as far as progress? 15:19:54 <reedip__> rtheis: any documentation for OSC plugin? 15:20:02 <rtheis> yes... 15:20:48 <rtheis> http://docs.openstack.org/developer/python-openstackclient/plugins.html 15:20:55 <reedip__> rtheis : thanks :) 15:21:16 <rtheis> I've also pushed a patch to clarify a few things on OSC side: https://review.openstack.org/#/c/292541/ 15:22:08 <mlavalle> rtheis: this is not specific to routed networks, right? 15:22:16 <rtheis> correct 15:22:30 <mlavalle> ok, anything else as far as progress? 15:22:49 <rtheis> mlavalle: I think routed networks will just be the first to tackle OSC plugin for neutron 15:23:07 <mlavalle> rtheis: ah, ok, that makes sense :-) 15:23:28 <mlavalle> so, let's move on... 15:23:44 <mlavalle> #topic OVN and routed networks 15:24:20 <mlavalle> xiaohhui: I added a topic for OVN to the agenda. Is there anything that you want to add here? 15:24:42 <xiaohhui> With my investigation, I think OVN needs some code change for routed network. I add them in the mail I sent out a litter earlier 15:24:46 <xiaohhui> nothing else 15:25:49 <mlavalle> xiaohhui: thank you. If I understood correctly, you sent that email to Carl, Russel and me. If I am correct, I suggest sending it to the ML, so others can chime in... 15:26:20 <rtheis> yes, that would be good to see 15:26:21 <rtheis> thanks 15:26:31 <mlavalle> IMO it always helps to get feedback from the community 15:26:33 <xiaohhui> OK, I will send in the mail list 15:26:44 <xiaohhui> thanks for the suggestion 15:27:05 <reedip__> m empty ... I agree 15:27:10 <mlavalle> #topic Open Discussion 15:27:36 <mlavalle> Anything else we should discuss with the team today? 15:27:55 <carl_baldwin> Hi 15:28:06 <reedip__> oh hey... he is back 15:28:06 <mlavalle> carl_baldwin: hey 15:28:13 <carl_baldwin> Just getting off of D.C. Metro 15:28:23 <reedip__> Arnold Shwarzennegger of Routed Networks... 15:28:52 <rtheis> I have a couple questions on user interface which may impact CLI and/or API 15:29:01 <carl_baldwin> reedip__: wondering if you're creating a new patch set or updating mine. 15:29:12 <reedip__> carl_baldwin : updating yours 15:29:35 <carl_baldwin> reedip__: would you mind creating a follow on patch set? 15:29:59 <rtheis> Will a segment always be tied to a single network and not standalone or scoped to another resource? 15:30:12 <reedip__> carl_baldwin : yeah, I can, to keep your work and mine independent (??) 15:30:30 <carl_baldwin> rtheis: yes 15:30:44 <rtheis> Is it possible to update any aspects of a segment after created (besides possibly name and description)? 15:30:51 <rtheis> carl_baldwin: thank you 15:30:57 <carl_baldwin> reedip__: minimize scope of each, mostly. 15:31:32 <reedip__> carl_baldwin : okay, will put a depends-on , 15:31:41 <carl_baldwin> reedip__: that way, they ought to merge earlier. 15:32:06 <carl_baldwin> reedip__: just make mine the commit's parent. 15:32:42 <reedip__> Yeah, will check out your code ( PS#6) , make changes and commit as a separate patch ( that would make your PS#6 the parent ) 15:33:12 <carl_baldwin> mlavalle: thanks for running the meeting. I'll read it all when I can. 15:33:23 <carl_baldwin> reedip__: cool, thanks 15:33:30 <mlavalle> carl_baldwin: np, enjoy your vacation :-) 15:33:31 <reedip__> carl_baldwin: np 15:34:13 <mlavalle> carl_baldwin: rtheis also asked Is it possible to update any aspects of a segment after created (besides possibly name and description)? 15:34:54 <rtheis> And finally, when creating a segment via CLI, are there any suggestions for default values? Right now, all options are required. Please comment in CLI patch sets if you believe a default makes sense. 15:35:49 <carl_baldwin> rtheis: no, no defaults 15:36:03 <rtheis> carl_baldwin: thank you 15:36:09 <carl_baldwin> rtheis: unless maybe for name when that's a thing 15:36:16 <rtheis> sure 15:37:12 <reedip__> name as default ? or Update ? 15:37:29 <reedip__> I think its for Update 15:38:04 <carl_baldwin> reedip__: both, default could be none and it could be updated. 15:38:34 <reedip__> carl_baldwin : Oh, ok ... but I think the API would be enough to handle it ? 15:38:56 <carl_baldwin> reedip__: you're right 15:38:58 <reedip__> I mean no need to define defaults in CLI if the API can handle it for name and description 15:39:23 * carl_baldwin on vacation. 15:40:10 <mlavalle> ok anything else? 15:40:16 <rtheis> mlavalle: I don't have anything else 15:41:18 <mlavalle> All right, thank you for attending and for your updates. Will see you next week (1 hour earlier) 15:41:32 <mlavalle> #endmeeting