| *** shashank_hegde has joined #openstack-kuryr | 00:01 | |
| *** _oanson has quit IRC | 00:20 | |
| *** limao has joined #openstack-kuryr | 00:39 | |
| openstackgerrit | Liping Mao proposed openstack/kuryr-libnetwork: The kuryr default port has changed from 2377 to 23750, we also need to change it in docker run_kuryr.sh script. https://review.openstack.org/340627 | 00:40 |
|---|---|---|
| *** banix has joined #openstack-kuryr | 00:41 | |
| *** shashank_hegde has quit IRC | 00:45 | |
| *** shashank_hegde has joined #openstack-kuryr | 00:48 | |
| *** bvandewa has joined #openstack-kuryr | 00:50 | |
| *** bvandewa has quit IRC | 00:54 | |
| *** bvandewa has joined #openstack-kuryr | 00:56 | |
| *** bvandewa has quit IRC | 00:57 | |
| *** bvandewa has joined #openstack-kuryr | 00:58 | |
| *** shashank_hegde has quit IRC | 01:08 | |
| *** banix has quit IRC | 01:10 | |
| *** limao_ has joined #openstack-kuryr | 01:30 | |
| *** limao has quit IRC | 01:33 | |
| *** prithiv has quit IRC | 01:40 | |
| *** bvandewa has quit IRC | 02:54 | |
| *** vikasc has joined #openstack-kuryr | 03:01 | |
| *** huikang has joined #openstack-kuryr | 03:12 | |
| *** yamamoto has quit IRC | 03:27 | |
| *** yamamoto has joined #openstack-kuryr | 03:27 | |
| *** shashank_hegde has joined #openstack-kuryr | 03:27 | |
| *** limao_ has quit IRC | 03:36 | |
| *** shashank_hegde has quit IRC | 03:37 | |
| *** shashank_hegde has joined #openstack-kuryr | 03:37 | |
| *** diga has joined #openstack-kuryr | 03:50 | |
| *** shashank_hegde has quit IRC | 03:50 | |
| *** huikang has quit IRC | 03:51 | |
| *** shashank_hegde has joined #openstack-kuryr | 04:00 | |
| *** limao has joined #openstack-kuryr | 04:10 | |
| *** shashank_hegde has quit IRC | 04:16 | |
| *** bvandewa has joined #openstack-kuryr | 04:32 | |
| *** shashank_hegde has joined #openstack-kuryr | 04:59 | |
| *** fawadkhaliq has joined #openstack-kuryr | 05:01 | |
| *** shashank_hegde has quit IRC | 05:04 | |
| *** irenab has joined #openstack-kuryr | 05:07 | |
| *** bvandewa has quit IRC | 05:08 | |
| vikasc | irenab, ping | 05:10 |
| irenab | vikasc: hi | 05:10 |
| vikasc | irenab, your point makes sense to me. | 05:11 |
| irenab | vikasc: regarding REST vs RPC? | 05:11 |
| vikasc | irenab, yes | 05:11 |
| vikasc | irenab, having it configurable | 05:11 |
| *** shashank_hegde has joined #openstack-kuryr | 05:12 | |
| vikasc | irenab, i am thinking of keeping original controllers.py with rest calls as it is | 05:12 |
| vikasc | irenab, and renaming rpc controller.py to something like cotroller_rpc.py | 05:12 |
| irenab | vikasc: alternatively, you can have driver layer for the Controller | 05:12 |
| irenab | since most of the code would be generic, and only the way to call neutron will differ | 05:13 |
| vikasc | irenab, do you mean like having fucntions like 'create_neutron_network' etc in kuryr/lib | 05:14 |
| vikasc | irenab, and then based on configuration framing the rpc call or neutron call | 05:15 |
| vikasc | irenab, in the function definition? | 05:15 |
| vikasc | adding these wrapper functions in kuryr/lib/utils | 05:16 |
| vikasc | irenab, kuryr/lib/utils.py i mean | 05:16 |
| irenab | the analogy would be the neutron services plugins that have drivers that you specify in config file | 05:16 |
| irenab | driver has base API implemented acccordingly (REST calls/ RPC calls) | 05:17 |
| vikasc | irenab, means all these methods like "create_neutron_network" will be defined in each driver file | 05:18 |
| *** limao_ has joined #openstack-kuryr | 05:19 | |
| vikasc | irenab, and from controllers.py will be invoked as app.driver.create_neutron_network | 05:19 |
| irenab | yes | 05:19 |
| irenab | you basically has base driver that defines methods and specific one that implement them | 05:20 |
| vikasc | irenab, thanks , will go through neutron service plugins also for reference. | 05:20 |
| vikasc | irenab, understood | 05:20 |
| vikasc | irenab, importing the base driver | 05:20 |
| irenab | vikasc: check this one: https://github.com/openstack/neutron/blob/master/neutron/services/qos/qos_plugin.py | 05:21 |
| vikasc | irenab, i think i understood your idea | 05:21 |
| vikasc | irenab, thanks a lot for guidance | 05:22 |
| *** limao has quit IRC | 05:22 | |
| vikasc | irenab, one more doubt | 05:22 |
| vikasc | irenab, to be honest i could not understand absoulte clearly why in case of vms we want rpc | 05:23 |
| vikasc | irenab, when i asked in meeting i got impression that everybody is certain that rpc is needed but due to lack of time i could not ask in detail. | 05:24 |
| vikasc | irenab, i decided to understand in future real pros of having rpc with vms | 05:24 |
| irenab | vikasc: I need to revisit the nested spec to make sure I remember the details. Lets try to chat about it here a bit later once apuimedo is online. I think its was about security and credentials distribution and related to Magnum being deploying the cluster | 05:25 |
| vikasc | irenab, sure. will keep looking for apuimedo | 05:26 |
| irenab | for Bare Metal I would love to hear banix, I think RPC is not required | 05:26 |
| vikasc | irenab, i am not able to clearly understand why rpc is needed even in vms case :D | 05:27 |
| vikasc | irenab, credentials anyways we have to store at vms to create rpc context | 05:28 |
| irenab | vikasc: but I think its not keystone | 05:28 |
| vikasc | irenab, keystone admin endpoint we will be able to skip with implementing rpc | 05:29 |
| vikasc | irenab, i just want to confirm is this the only purpose of rpc | 05:29 |
| irenab | vikasc: exactly, I agree | 05:29 |
| vikasc | irenab, otherwise we can move all neutron calling functionality to common kuryr/lib/utils.py | 05:30 |
| vikasc | irenab, with a little bit refactoring | 05:30 |
| vikasc | irenab, and then all k8s or any other coe can reuse it | 05:30 |
| *** limao_ has quit IRC | 05:32 | |
| irenab | vikasc: this can be great, and I think we can do it anyway regardless split to drivers | 05:32 |
| *** limao has joined #openstack-kuryr | 05:32 | |
| irenab | to avoid code duplication | 05:33 |
| vikasc | irenab, will be doing it this way only even with drivers | 05:33 |
| vikasc | irenab, i mean agree.. will be doing it in any case | 05:33 |
| vikasc | irenab, but if rpc is used functionality of these methods will be there in handlers on kuryr rpc server | 05:34 |
| *** obondarev has joined #openstack-kuryr | 05:35 | |
| vikasc | irenab, if we do rpc and decide to call neutron rest apis in case of baremetal as you suggested, then in that case methods from kuryr/lib/utils.py will be used | 05:36 |
| vikasc | irenab, ahh.. sorry rpc_handlers can also use the same util methods | 05:37 |
| vikasc | irenab, in all cases, its making sense to keep these at kuryr/lib/utils | 05:38 |
| irenab | vikasc: or maybe kuryr/lib/drivers | 05:40 |
| vikasc | irenab, even if we decide not to use rpc? | 05:41 |
| irenab | but in general I like the direction. There is nothing COE specific, so makes sense to keep in in kuryr/lib | 05:41 |
| irenab | vikasc: utils makes more sense to have more auxiliry methods. ‘drivers’ maybe named better :-) | 05:42 |
| vikasc | irenab, i mean introducing "drivers" term, even in case if decide not tu use rpc | 05:42 |
| vikasc | irenab, ? | 05:43 |
| vikasc | irenab, i think are almost on the same page now | 05:44 |
| vikasc | irenab, :) | 05:44 |
| irenab | vikasc: named not drivers, but something like ‘neutron’ … Not sure | 05:44 |
| *** obondarev has quit IRC | 05:44 | |
| vikasc | irenab, next is to discuss pros of rpc | 05:44 |
| irenab | vikasc: great. Maybe worth to have it on email thread | 05:45 |
| vikasc | irenab, lets discuss with apuimedo atleat first | 05:46 |
| irenab | vikasc: sounds like a plan | 05:46 |
| vikasc | irenab, plan? sorry i did not get you :) | 05:47 |
| vikasc | irenab, it takes a lot of time to get reply on ml :P | 05:48 |
| irenab | I mean, sounds great. Lets do it | 05:48 |
| vikasc | irenab, thanks | 05:48 |
| irenab | first discuss with apuimedo and then see if need to communicate more over email | 05:48 |
| irenab | or just follow up with the patch | 05:49 |
| vikasc | irenab, exactly .. bye till we have apuimedo on irc | 05:49 |
| irenab | vikasc: have a great day | 05:49 |
| *** fawadkhaliq has quit IRC | 05:58 | |
| *** obondarev has joined #openstack-kuryr | 05:59 | |
| *** bvandewa has joined #openstack-kuryr | 06:08 | |
| *** bvandewa_ has joined #openstack-kuryr | 06:12 | |
| *** bvandewa has quit IRC | 06:12 | |
| vikasc | apuimedo, ping | 06:16 |
| *** obondarev has quit IRC | 06:18 | |
| *** obondarev has joined #openstack-kuryr | 06:25 | |
| openstackgerrit | janonymous proposed openstack/kuryr: Usage is /Plugin.Active --> /Plugin.Activate https://review.openstack.org/340702 | 06:28 |
| *** obondarev has quit IRC | 06:28 | |
| *** fawadkhaliq has joined #openstack-kuryr | 06:56 | |
| *** fawadkhaliq has quit IRC | 06:59 | |
| *** fawadkhaliq has joined #openstack-kuryr | 06:59 | |
| *** oshidoshi has joined #openstack-kuryr | 07:01 | |
| *** oshidoshi has quit IRC | 07:13 | |
| vikasc | apuimedo, ping | 07:16 |
| *** obondarev has joined #openstack-kuryr | 07:26 | |
| *** oshidoshi has joined #openstack-kuryr | 07:39 | |
| *** garyloug has joined #openstack-kuryr | 07:59 | |
| *** fawadkhaliq has quit IRC | 07:59 | |
| *** fawadkhaliq has joined #openstack-kuryr | 08:01 | |
| *** fawadkhaliq has quit IRC | 08:05 | |
| *** oshidoshi has quit IRC | 08:23 | |
| *** limao has quit IRC | 08:27 | |
| *** shashank_hegde has quit IRC | 08:27 | |
| *** oshidoshi has joined #openstack-kuryr | 08:28 | |
| *** devvesa has joined #openstack-kuryr | 08:35 | |
| *** devvesa has quit IRC | 08:38 | |
| *** devvesa has joined #openstack-kuryr | 08:39 | |
| *** obondarev has quit IRC | 08:39 | |
| *** obondarev has joined #openstack-kuryr | 08:41 | |
| *** limao has joined #openstack-kuryr | 08:44 | |
| *** obondarev has quit IRC | 08:56 | |
| *** fawadkhaliq has joined #openstack-kuryr | 09:02 | |
| *** fawadkhaliq has quit IRC | 09:06 | |
| *** neiljerram has joined #openstack-kuryr | 09:19 | |
| *** fawadkhaliq has joined #openstack-kuryr | 09:24 | |
| *** obondarev has joined #openstack-kuryr | 10:05 | |
| *** limao has quit IRC | 10:17 | |
| apuimedo | ping neiljerram | 10:28 |
| apuimedo | vikasc: hey | 10:28 |
| apuimedo | Sorry I didn't see it untiil a bit late :/ | 10:28 |
| vikasc | apuimedo, hi | 10:28 |
| vikasc | apuimedo, np | 10:28 |
| apuimedo | glad you are still around! | 10:28 |
| vikasc | :) | 10:29 |
| apuimedo | I thought I may have missed you | 10:29 |
| vikasc | apuimedo, lets see if irenab is also around | 10:29 |
| neiljerram | apuimedo, hi | 10:29 |
| vikasc | apuimedo, on my rpc support patch she raised a concern that why we need rpc | 10:30 |
| irenab | vikasc: it was more about replacment of the REST API versus have two drivers to choose based on deployment option | 10:30 |
| vikasc | apuimedo, only pros which i could share to her was that keystone and neutron endpoints can be avoided on vms | 10:31 |
| apuimedo | from where I stand | 10:31 |
| apuimedo | I would only use RPC in the nested case | 10:32 |
| vikasc | apuimedo, can you please share reasons for your preference | 10:32 |
| apuimedo | vikasc: I think on non-nested case adding another level of indirection would introduce too much latency | 10:34 |
| vikasc | apuimedo, skipping service endpoints is the one | 10:34 |
| vikasc | apuimedo, agree | 10:34 |
| vikasc | apuimedo, on that part i agree that introducing a driver layer makes sense | 10:35 |
| vikasc | irenab, ping | 10:35 |
| apuimedo | so, for me, I'd have kuryr-common to have an optional rpc server | 10:36 |
| vikasc | apuimedo, was wondering about pros of using rpc instead of calling neutron rest api directly | 10:36 |
| apuimedo | that's a good question | 10:37 |
| apuimedo | I would make the RPC probably use a different auth | 10:37 |
| vikasc | apuimedo, means other than keystone | 10:38 |
| vikasc | apuimedo, ? | 10:38 |
| apuimedo | yes, I was thinking that it could just be an ssl connection between VM and kuryr server | 10:39 |
| irenab | vikasc: sorry in the meeting. Will check the log in 40 mins | 10:40 |
| vikasc | irenab, sure | 10:40 |
| vikasc | apuimedo, and how certificates will be exchanged? | 10:41 |
| vikasc | and stored? | 10:41 |
| vikasc | apuimedo, can you please give some reference example? | 10:42 |
| vikasc | apuimedo, is it like using barbican | 10:43 |
| vikasc | apuimedo, do you mean like this, https://github.com/openstack/magnum/blob/master/specs/tls-support-magnum.rst | 10:44 |
| vikasc | apuimedo, tls support | 10:45 |
| apuimedo | I would need to talk with Magnum people about that :P | 10:45 |
| apuimedo | I haven't really gone into detail | 10:45 |
| apuimedo | but I guess that hongbin would be able to tell us | 10:45 |
| vikasc | apuimedo, okay on high level its all about security | 10:45 |
| vikasc | apuimedo, right? | 10:46 |
| vikasc | apuimedo, aim is to try to avoid user credentials and endpoints on vms | 10:46 |
| apuimedo | right | 10:48 |
| vikasc | apuimedo, rpc support is working at the moment and has been pushed for review on both repos | 10:49 |
| apuimedo | vikasc: I'll take a look this week | 10:50 |
| apuimedo | I still don't know exactly when | 10:50 |
| vikasc | apuimedo, what is your opinion on introducing a driver layer as Irena and I discussed? | 10:50 |
| apuimedo | where can I read more about it? | 10:50 |
| vikasc | apuimedo, and what is the plan to accept patches in kuryr repo | 10:50 |
| vikasc | apuimedo, at the moment requirements is being met from my github repo | 10:51 |
| apuimedo | vikasc: I'd like to be able to make a kuryr release by the end of next week | 10:51 |
| vikasc | apuimedo, i have to sync the patches in kuryr repo with my account each time i make a change | 10:51 |
| apuimedo | which should unblock us in kuryr-libnetwork | 10:52 |
| apuimedo | sorry about that :( | 10:52 |
| vikasc | apuimedo, no problem. Request you to just review kuryr repo patches | 10:53 |
| apuimedo | vikasc: will do! | 10:53 |
| apuimedo | thanks! | 10:53 |
| vikasc | apuimedo, thanks !! | 10:53 |
| *** apuimedo is now known as apuimedo|lunch | 10:57 | |
| *** obondarev has quit IRC | 11:00 | |
| *** diga has quit IRC | 11:04 | |
| *** bvandewa_ has quit IRC | 11:20 | |
| janonymous | Hi guys, This reference guide: http://docs.openstack.org/developer/kuryr/devref/index.html, contains milestones for mitaka where could i found milestones for this release ? | 11:20 |
| *** obondarev has joined #openstack-kuryr | 11:41 | |
| *** irenab has quit IRC | 11:42 | |
| *** irenab has joined #openstack-kuryr | 11:42 | |
| *** fawadkhaliq has quit IRC | 11:50 | |
| *** fawadkhaliq has joined #openstack-kuryr | 12:17 | |
| *** garyloug has quit IRC | 12:18 | |
| *** fawadkhaliq has quit IRC | 12:20 | |
| apuimedo|lunch | janonymous: I'm not sure we set any | 12:31 |
| *** obondarev has quit IRC | 12:31 | |
| apuimedo|lunch | I'll try to get around to doing some | 12:32 |
| *** apuimedo|lunch is now known as apuimedo | 12:32 | |
| *** obondarev has joined #openstack-kuryr | 12:38 | |
| *** garyloug has joined #openstack-kuryr | 13:07 | |
| *** obondarev has quit IRC | 13:13 | |
| *** banix has joined #openstack-kuryr | 13:14 | |
| *** neiljerram has quit IRC | 13:17 | |
| *** vikasc has quit IRC | 13:18 | |
| *** neiljerram has joined #openstack-kuryr | 13:23 | |
| *** obondarev has joined #openstack-kuryr | 13:29 | |
| *** banix has quit IRC | 13:57 | |
| *** yamamoto has quit IRC | 13:59 | |
| *** limao has joined #openstack-kuryr | 13:59 | |
| *** banix has joined #openstack-kuryr | 14:00 | |
| *** limao has quit IRC | 14:01 | |
| *** banix has quit IRC | 14:01 | |
| *** huikang has joined #openstack-kuryr | 14:02 | |
| *** huikang has quit IRC | 14:09 | |
| *** huikang has joined #openstack-kuryr | 14:17 | |
| *** _oanson has joined #openstack-kuryr | 14:29 | |
| *** huikang has quit IRC | 14:34 | |
| *** oshidoshi has quit IRC | 14:41 | |
| *** yamamoto has joined #openstack-kuryr | 14:59 | |
| *** huikang has joined #openstack-kuryr | 15:05 | |
| *** yamamoto has quit IRC | 15:07 | |
| *** vikasc has joined #openstack-kuryr | 15:20 | |
| *** huikang has quit IRC | 15:40 | |
| *** banix has joined #openstack-kuryr | 15:45 | |
| *** mestery has quit IRC | 15:49 | |
| *** mestery has joined #openstack-kuryr | 15:50 | |
| *** obondarev has quit IRC | 15:56 | |
| *** diogogmt has joined #openstack-kuryr | 15:56 | |
| *** _oanson has quit IRC | 16:00 | |
| *** obondarev has joined #openstack-kuryr | 16:02 | |
| *** obondarev has quit IRC | 16:16 | |
| *** huikang has joined #openstack-kuryr | 16:21 | |
| *** shashank_hegde has joined #openstack-kuryr | 16:22 | |
| *** devvesa has quit IRC | 16:22 | |
| *** vikasc has quit IRC | 16:25 | |
| *** obondarev has joined #openstack-kuryr | 16:26 | |
| *** obondarev has quit IRC | 16:32 | |
| *** diga has joined #openstack-kuryr | 16:38 | |
| *** huikang has quit IRC | 16:40 | |
| *** shashank_hegde has quit IRC | 16:40 | |
| *** huikang has joined #openstack-kuryr | 16:49 | |
| banix | apuimedo: hi toni; are you working on the kuryr gate issues? need to get vikas’ patch for removing stuff from kuryr-lib in. | 16:53 |
| apuimedo | I'm without broadband at home. So I can't tackle it until the morning :( | 16:55 |
| banix | that should be good enigh :) | 17:12 |
| *** bvandewa has joined #openstack-kuryr | 17:18 | |
| *** banix has quit IRC | 17:25 | |
| *** shashank_hegde has joined #openstack-kuryr | 17:38 | |
| *** huikang has quit IRC | 17:42 | |
| *** bvandewa has quit IRC | 17:44 | |
| *** huikang has joined #openstack-kuryr | 17:51 | |
| *** fawadkhaliq has joined #openstack-kuryr | 17:57 | |
| *** garyloug has quit IRC | 17:59 | |
| *** janonymous has quit IRC | 18:06 | |
| *** bvandewa has joined #openstack-kuryr | 18:14 | |
| *** banix has joined #openstack-kuryr | 18:15 | |
| *** huikang has quit IRC | 18:16 | |
| *** bvandewa has quit IRC | 18:19 | |
| *** bvandewa has joined #openstack-kuryr | 18:40 | |
| *** bvandewa has quit IRC | 18:41 | |
| *** bvandewa has joined #openstack-kuryr | 18:41 | |
| *** huikang has joined #openstack-kuryr | 19:17 | |
| *** obondarev has joined #openstack-kuryr | 19:18 | |
| *** huikang has quit IRC | 19:21 | |
| *** fawadkhaliq has quit IRC | 19:53 | |
| *** banix has quit IRC | 20:00 | |
| *** banix has joined #openstack-kuryr | 20:04 | |
| *** huikang has joined #openstack-kuryr | 20:18 | |
| *** huikang has quit IRC | 20:24 | |
| *** obondarev has quit IRC | 20:24 | |
| *** irenab has quit IRC | 20:40 | |
| *** shashank_hegde has quit IRC | 20:57 | |
| *** shashank_hegde has joined #openstack-kuryr | 21:01 | |
| *** banix has quit IRC | 21:08 | |
| *** shashank_hegde has quit IRC | 21:09 | |
| *** banix has joined #openstack-kuryr | 21:43 | |
| *** shashank_hegde has joined #openstack-kuryr | 22:21 | |
| *** diogogmt has quit IRC | 22:34 | |
| *** diogogmt has joined #openstack-kuryr | 22:38 | |
| *** shashank_hegde has quit IRC | 23:03 | |
| *** shashank_hegde has joined #openstack-kuryr | 23:03 | |
| *** shashank_hegde has quit IRC | 23:19 | |
| *** banix has quit IRC | 23:29 | |
| *** diogogmt has quit IRC | 23:39 | |
| *** shashank_hegde has joined #openstack-kuryr | 23:53 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!