04:31:04 <gongysh> #startmeeting tacker 04:31:05 <openstack> Meeting started Wed Dec 20 04:31:04 2017 UTC and is due to finish in 60 minutes. The chair is gongysh. Information about MeetBot at http://wiki.debian.org/MeetBot. 04:31:06 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 04:31:09 <openstack> The meeting name has been set to 'tacker' 04:31:21 <dkushwaha> o/ 04:31:24 <gongysh> dkushwaha, hi 04:31:37 <phuoc_> o/ 04:31:38 <gongysh> #topic roll call 04:31:43 <gongysh> phuoc_, hi 04:32:04 <phuoc_> hi mr.Gongysh 04:32:20 <joxyuki> hi 04:32:33 <phuoc_> hi joxyuki 04:32:40 <diga> o/ 04:34:18 <gongysh> tbh, YanXing_an , hi 04:34:29 <YanXing_an> hi 04:34:35 <gongysh> #topic bp 04:34:46 <gongysh> phuoc_, do you have any update for your bps? 04:35:07 <phuoc_> gongysh, I am starting in implementation 04:35:08 <phuoc_> https://review.openstack.org/#/c/529116/ 04:35:50 <phuoc_> if you guys have time, please give me feed back in https://review.openstack.org/#/c/529116/1/tacker/vnfm/infra_drivers/kubernetes/k8s/translate_inputs.py 04:36:05 <phuoc_> and https://review.openstack.org/#/c/529116/1/tacker/tosca/lib/tacker_defs.yaml 04:36:22 <phuoc_> I will update the other, and document for this patch soon 04:37:43 <gongysh> phuoc_, nice progress 04:38:20 <phuoc_> thanks 04:38:20 <gongysh> https://review.openstack.org/#/c/529116/1/samples/tosca-templates/vnfd/tosca-vnfd-containerized.yaml 04:38:52 <gongysh> it seems there are difference between vnfds for openstack and k8s 04:39:58 <phuoc_> gongysh, I follow http://docs.oasis-open.org/tosca/tosca-nfv/v1.0/csd04/tosca-nfv-v1.0-csd04.html#_Toc482896065 04:40:21 <phuoc_> they define new type: tosca.datatypes.nfv.VnfcConfigurableProperties 04:41:18 <phuoc_> so, I try to match to new definition in TOSCA 04:41:57 <gongysh> tbh, hi 04:43:48 <gongysh> ok, I will have a look at. 04:44:29 <gongysh> dkushwaha, could you have a look at it too? 04:44:43 <dkushwaha> gongysh, sure. 04:44:54 <gongysh> thanks 04:45:12 <gongysh> phuoc_, do you have other bps during the cycle? 04:45:38 <phuoc_> no, I don't 04:45:47 <gongysh> ok 04:45:58 <gongysh> diga, could you please talk about the API bp? 04:46:02 <phuoc_> I and mardim will implement vnffg update 04:46:17 <gongysh> phuoc_, yes, thanks 04:47:01 <phuoc_> thank gongysh and dkushwaha 04:47:12 <diga> gongysh: okay 04:47:22 <gongysh> ok, please 04:47:43 <diga> gongysh: https://review.openstack.org/#/c/520495/ 04:47:48 <diga> hello everyone 04:48:07 <diga> current approach I have taken using the falcon framework 04:48:14 <phuoc_> hi diga :) 04:48:32 <diga> I see we need only one dependancy now which falcon module 04:48:40 <diga> which is also part of openstack requirement 04:48:59 <diga> so we don't have to ask for approval for openstack requirement team 04:49:06 <diga> which is already exist there 04:49:22 <gongysh> Which frameworks are in use in Mitaka: 04:49:22 <gongysh> Falcon: 4 projects 04:49:24 <gongysh> Custom + routes: 12 projects 04:49:24 <gongysh> Pecan: 12 projects 04:49:26 <gongysh> Flask: 2 projects 04:49:26 <gongysh> web.py: 1 project 04:49:29 <diga> AS this is first patch as part of this developmenet 04:49:43 <gongysh> this is from https://www.openstack.org/blog/2016/06/openstack-developer-mailing-list-digest-20160624/ 04:49:45 <diga> yes gongysh 04:49:53 <gongysh> so we will use falcon anyway. 04:49:59 <gongysh> that's fine. 04:50:00 <diga> gongysh: yeah 04:50:44 <diga> currently zuul is failing because falcon feature branch is too old, so I am working with infra team to rebase the feature branch with master branch 04:50:59 <diga> I hope this issue will get resolved by today/tomorrow 04:51:12 <phuoc_> diga, that's good 04:51:22 <gongysh> diga, we will wait for you and let the API improvement go smoothly and fast. 04:51:46 <diga> once we merge the first patch, my next target is to implement one after one api of tacker into falcon framework 04:51:57 <diga> gongysh: yes, sure 04:52:24 <gongysh> diga, thanks 04:52:33 <diga> gongysh: I understand that this is long term pending thing but anyway I am targetting this for queens release 04:52:56 <gongysh> diga, I remember that your team help with tacker openstack command 04:52:59 <diga> lets see how far we go on this if nothing major comes on the road 04:53:24 <diga> gongysh: yes, but I have now only one person working on tacker 04:53:38 <diga> currently kapil_ is working on solving bugs 04:53:48 <gongysh> ok 04:54:12 <gongysh> dkushwaha, do you have any updates on bps? 04:54:14 <diga> I am giving him enough time to solve the bugs first, then I am thinking to involve him in the tacker api framework 04:54:35 <diga> because i see so much work involved in this feature 04:54:41 <dkushwaha> gongysh, not for now. 04:54:48 <gongysh> ok 04:55:08 <gongysh> YanXingAn, do you have any to say for your bps? 04:56:21 <gongysh> ok, seems YanXingAn is still at their company meeting. 04:56:28 <gongysh> #topic bug 04:56:42 <gongysh> first bug is about CI system 04:56:57 <gongysh> https://bugs.launchpad.net/tacker/+bug/1739342 04:56:58 <openstack> Launchpad bug 1739342 in tacker "tacker ci does not support tls_proxy service enabled" [High,In progress] - Assigned to yong sheng gong (gongysh) 04:57:17 <gongysh> it seems our tacker CI job does not support tls proxy 04:57:31 <gongysh> I have an quick fix to disable the tls proxy first. 04:57:49 <gongysh> and need some time to investigate how we support tls proxy. 04:57:57 <phuoc_> https://bugs.launchpad.net/tacker/+bug/1719841 04:57:57 <openstack> Launchpad bug 1719841 in tacker "vim configuration does not allow the INSECURE property" [Medium,Triaged] - Assigned to Taseer Ahmed (taseer94) 04:58:05 <phuoc_> I see many people meet that bug 04:59:27 <gongysh> ok, let fix it in that bug 04:59:46 <gongysh> https://launchpad.net/~taseer94 05:00:02 <gongysh> he is not active, then I will try to fix it. 05:00:36 <gongysh> who knows Taseer Ahmed (taseer94) ? 05:01:12 <phuoc_> he is from opnfv-sfc project 05:01:30 <gongysh> ok 05:02:12 <gongysh> dkushwaha, do you have any other bug to talk? 05:02:19 <gongysh> to discuss 05:02:57 <dkushwaha> gongysh, not worked in that part in last week. 05:03:27 <gongysh> ok, if you have time, please keep an eye on the bug reports. 05:03:38 <gongysh> #topic open dicusssion 05:03:45 <dkushwaha> gongysh, sure, will do. 05:04:20 <dkushwaha> gongysh, i have some points to discuss regarding heat stack 05:04:33 <gongysh> dkushwaha, sure, please 05:05:11 <dkushwaha> gongysh, we are doing some PoC regarding resource reservation in Tacker 05:05:26 <gongysh> en 05:05:34 <dkushwaha> gongysh, so we are integrating Tacker-Blazor for that 05:06:13 <phuoc_> dkushwaha, that sounds good 05:06:40 <dkushwaha> gongysh, in some cases we needs to update the heat stack, but as of now we can not update heat stack in vnf-update operation 05:07:23 <dkushwaha> gongysh, i want to know the approach how we can update it? 05:08:22 <gongysh> dkushwaha, what do you mean the resource reservation related to heat stack. 05:08:52 <gongysh> do you have devise a way to do it in pure heat environment? 05:09:31 <gongysh> can you paste tacker-blazor url here too? 05:10:07 <phuoc_> dkushawaha, I think we can call e.g. heat stack update to update vnf 05:10:59 <dkushwaha> gongysh, i am not sure, i can paste link here, as it private for now, but my team started drafting spec for that 05:12:56 <gongysh> dkushwaha, to do it, the steps are: first devise a way to do in pure heat way, and then do integration on tacker. maybe we need to update tacker codes to do it. 05:14:06 <joxyuki> hi all, I am working tacker-blazar with dkushwaha. 05:14:06 <dkushwaha> gongysh, once a vnf deployed, we create reservation in blazor side, and then needs to update vnf with that reservation data. which requires update 05:14:56 <dkushwaha> joxyuki, could you please explain 05:15:06 <joxyuki> yes, to create a VM using reserved resource, heat stack need to know reservation-id in HOT. 05:16:18 <gongysh> joxyuki, that's good. 05:16:23 <joxyuki> gongysh, what do you mean "device" exactly? 05:16:58 <gongysh> device is old term for vnf? 05:17:16 <gongysh> device is old term for vnf. 05:17:24 <joxyuki> i see, thax 05:17:27 <joxyuki> thanks 05:17:44 <gongysh> previously, tacker is service vm manager. 05:18:32 <gongysh> joxyuki, so you seq is: vnfm -> nfvo: reserve resource 05:18:54 <gongysh> vnfm -> instantiate vnf with reservation id. 05:19:47 <joxyuki> gyngysh, no. 05:20:37 <dkushwaha> joxyuki, sorry for interrupting, but i think you interpreting "devise" as "device" 05:21:18 <joxyuki> 1) user-> tacker: vnf-create, 2) user -> blazar: reservation, 3) user->tacker: input reservation-id 05:21:39 <joxyuki> yes, devise -> device 05:22:26 <gongysh> joxyuki, 1) user -> tacker: vnf-create will create vms on target openstack, why do you still need to do 2)? 05:23:28 <gongysh> I think the sequence is to be: a) user -> blazar: reservation b) user-> tacker: vnf-create with reservation-id 05:24:25 <joxyuki> gongysh, could you see https://youtu.be/a1loFExpjb0 ? this is our use case. 05:25:19 <gongysh> joxyuki, ok, I will have a look 05:25:28 <joxyuki> gongysh, this is a use case need reservation for short term events such as sports match. 05:25:54 <gongysh> joxyuki, glad to know your team is trying to use tacker. 05:27:24 <gongysh> joxyuki, keep in touch at tacker channel 05:27:29 <gongysh> thanks everyone 05:27:32 <joxyuki> gongysh, yes, keep in touch. 05:27:42 <gongysh> #endmeeting