08:00:24 <gongysh> #startmeeting tacker 08:00:25 <openstack> Meeting started Tue Apr 3 08:00:24 2018 UTC and is due to finish in 60 minutes. The chair is gongysh. Information about MeetBot at http://wiki.debian.org/MeetBot. 08:00:26 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 08:00:28 <openstack> The meeting name has been set to 'tacker' 08:01:08 <gongysh> #topic roll call 08:01:14 <longkb__> o/ 08:01:25 <longkb__> Hi everyone 08:01:31 <gongysh> longkb__, hi 08:01:45 <longkb__> hi gongysh 08:01:57 <phuoc> hi 08:02:05 <keiko> o/ 08:02:06 <joxyuki> hi team 08:02:13 <gongysh> helo, phuoc joxyuki 08:02:15 <longkb__> hi phuoc 08:03:52 <phuoc> o/ 08:04:59 <gongysh> #topic bp 08:05:16 <gongysh> phuoc, your kuryr k8s bp 08:05:22 <phuoc> yes 08:05:28 <haint_> o/ 08:05:49 <phuoc> I will revise it and upload a patch set in tacker in this week 08:06:01 <gongysh> the difference and change between pure k8s and kuryr k8s is yet to made clear. 08:06:24 <phuoc> I saw your comments about it 08:06:42 <phuoc> thanks a lot 08:07:26 <gongysh> haint_. the mistral action patches 08:07:41 <gongysh> https://review.openstack.org/#/c/486924/ 08:08:10 <gongysh> you have said you are looking into the action part, that's fine, thanks. 08:08:32 <haint_> yes, I rewrite the update status of VNF in call action: http://paste.openstack.org/show/718234/ 08:08:52 <haint_> it is shorter, now, find the way to call the action part 08:08:54 <gongysh> trin is on sick leave. so we will skip the web studio bp part. 08:09:35 <haint_> are we now only working on the call_vnf_policy_action(), gongysh? 08:09:52 <haint_> is the other part ok? 08:10:28 <gongysh> haint_, policy is divided into two parts: one is monitoring and another is action to trigger. 08:10:44 <gongysh> monitoring is in mistral, action part is in conductor. 08:11:01 <gongysh> we still have httping monitoring. 08:11:27 <haint_> if the VNF status is DEAD, it will call the action part, is is right? 08:11:29 <gongysh> for the action, we have respawn, log, etc. 08:12:03 <hyunsikyang> Hello!:) i am late today. 08:12:34 <haint_> ah, still have more works. 08:12:35 <gongysh> haint_, that depends on the policy: if the policy is: failure: respawn, that is the way. 08:12:48 <haint_> I understand now. 08:13:21 <longkb__> hi hyunsikyang 08:13:23 <haint_> if you have time, please also help me to finish this 08:14:41 <gongysh> we can wait. but hope you can finish it in p1 with :ping monitor and respawn action. 08:14:59 <gongysh> we can refactor it into other monitor: such as httpping 08:15:18 <haint_> yes, i will do that first. 08:15:27 <gongysh> keiko, joxyuki , hi 08:15:39 <keiko> hi 08:15:40 <joxyuki> hi 08:15:49 <longkb__> gongysh, please keep reviewing my cluster patch set. There are still some nits, I will revise my patch set now. https://review.openstack.org/#/c/529808/ 08:15:53 <gongysh> do you have any update about your stuff? 08:16:46 <joxyuki> gongysh, now we are still working to write our specs, reservatin feature and affinity policy feature 08:17:01 <gongysh> joxyuki. ok 08:17:28 <gongysh> joxyuki, for reservation, please refer to etsi related parts. 08:17:36 <gongysh> longkb__, hi 08:17:46 <longkb__> hi gongysh 08:18:13 <joxyuki> gongysh, yes 08:18:27 <gongysh> I have commented on cluster spec 08:18:52 <gongysh> we need to make the feature multiple vim compatible 08:18:58 <joxyuki> I will submit my spec of reservation next week, please review it. 08:19:08 <gongysh> joxyuki, sure. 08:20:28 <gongysh> I think those are all for our bp parts. 08:20:39 <gongysh> #topic open discussion. 08:21:05 <gongysh> anything to talk? 08:21:06 <longkb__> Oh, I missed your comment. But I think I should finish cluster feature for single VIM firstly 08:21:16 <hyunsikyang> can i re-submit the port monitoring driver ? 08:21:28 <haint_> mistral spec: https://review.openstack.org/#/c/557376/, just modify a little bit, please review it 08:21:47 <hyunsikyang> last time, i submitted, but at that time, tacker didn't support VNFC(application) 08:21:55 <hyunsikyang> so i will update it! 08:21:55 <haint_> bug 1751433, https://review.openstack.org/#/c/557728/, about 08:21:57 <openstack> bug 1751433 in tacker "vnffgd-create cannot load the 'description'" [Medium,In progress] https://launchpad.net/bugs/1751433 - Assigned to Nguyen Hai (nguyentrihai93) 08:21:57 <haint_> Fix create_vnffgd cannot get description from TOSCA file 08:23:10 <haint_> about vnffgd, there is no reason why we should have 2 description lines in the template, so the patch below fixed it. 08:25:49 <haint_> https://docs.openstack.org/tacker/latest/contributor/vnffgd_template_description.html, also in the vnffgd_template_description, it didn't show the 'description' under 'topology_template' 08:25:59 <gongysh> haint_, so you mean description should not under topology_template? 08:26:10 <haint_> yes it is 08:26:57 <haint_> the patch fixed two problems: one is cannot load description, two is delete redundant 'description' line under topology_template 08:27:42 <gongysh> http://docs.oasis-open.org/tosca/tosca-nfv/v1.0/tosca-nfv-v1.0.html 08:27:45 <mardim> hello guys 08:27:53 <mardim> soory I was afk 08:27:54 <gongysh> The service template example of the above VNFD is showing as follow 08:28:07 <gongysh> please look at that sample. 08:28:44 <gongysh> the above is under sample. 08:28:49 <gongysh> The service template example for VDU C is showing as follow. 08:29:06 <gongysh> it is an above sample 08:29:20 <haint_> yes, but it is for VDU only 08:29:51 <haint_> in case of VNFFGD, i cannot see any template with description under the topology_template 08:30:18 <haint_> and I don't understand why we should have two 'description' lines. 08:31:34 <gongysh> do you have any tosca vnffgd type definition in tosca standard community to read the description is above topology_template? 08:32:35 <haint_> the 08:32:36 <haint_> tosca_definitions_version of VNFD is tosca_simple_yaml_1_0 08:32:49 <haint_> VNFFD is tosca_definitions _version: tosca_simple_profile_for_nfv_1_0 08:33:10 <haint_> I mentioned on the link http://docs.oasis-open.org/tosca/tosca-nfv/v1.0/csd04/tosca-nfv-v1.0-csd04.html 08:35:36 <gongysh> haint_, I cannot see any vnffgd at that doc. 08:36:06 <haint_> the final example in that link 08:36:18 <haint_> below "Figure 6.1-2 example of VDU substitution mappings" 08:37:11 <haint_> I will find out more and comment on the patch. 08:37:34 <gongysh> haint_, if you have tosca type definition, it will help more. 08:38:02 <haint_> however, https://docs.openstack.org/tacker/latest/contributor/vnffgd_template_description.html 08:38:21 <haint_> in our docs, it doesn't show the 'description' under 'topology_template' 08:39:49 <gongysh> haint_, ok, we can have this change 08:40:09 <gongysh> anything else, guy? 08:40:16 <gongysh> anybody? 08:40:29 <longkb__> gongysh, I think I should finish cluster feature for single VIM firstly 08:40:46 <longkb__> your comment mentioned about multiple VNFD and multiple VIMs 08:41:28 <gongysh> longkb__, there were comments about lb's position 08:41:33 <gongysh> before. 08:42:01 <gongysh> I think we should make lb to sit on a given vim. 08:42:30 <longkb__> ah, OK. I will update my code to specify lb location 08:42:49 <gongysh> also, for vnfd on each, vim, our vnfd is different for different vims. 08:43:34 <gongysh> For a same feature, the vnfd on openstack vim and on k8s vim are different. 08:44:05 <gongysh> so we need to fill in vnfd name on each vim to cope with multiple vim. 08:44:25 <longkb__> But in this spec, I just make it to work with single VIM 08:44:50 <longkb__> A multiple VNFD and VIMs could be supported in other spec 08:44:59 <gongysh> longkb__, we should think furture to avoid the big code change. 08:46:30 <gongysh> k8s spec is the one we maybe have to change the previous design. 08:47:27 <longkb__> Ah, Ok. I got your point. I will update my code asap 08:47:53 <gongysh> hyunsikyang, hi 08:49:29 <gongysh> helo, anything else? 08:49:53 <gongysh> do you guys have finished the summit ppt? 08:50:00 <gongysh> phuoc. mardim ? 08:50:03 <gongysh> mardim. hi 08:50:08 <mardim> gongysh, hello 08:50:18 <mardim> I am at the begginign 08:50:28 <mardim> I didn't have the time to do much 08:50:41 <mardim> because I had a demo in ONS summit 08:50:41 <gongysh> have you started our fancy session slides? 08:50:43 <mardim> last week 08:50:59 <gongysh> mardim, what demo at ONS summit? 08:51:00 <mardim> I ve written only two slides 08:51:07 <mardim> gongysh, for SFC-XCI 08:51:14 <mardim> I was in LA last week 08:51:18 <mardim> I had a demo there 08:51:33 <gongysh> mardim, one picture worths thousands of words. 08:51:53 <gongysh> SFC-XC? 08:51:56 <gongysh> what is XC? 08:52:20 <mardim> gongysh, I will send you picture from the summit in tacker channel after the meeting 08:52:33 <gongysh> ok, thanks 08:52:36 <mardim> gongysh, Regarding the Openstack PPT 08:52:45 <mardim> gongysh, There is no deadline right ? 08:52:59 <mardim> gongysh, And another thing 08:53:06 <mardim> gongysh, Are we gonna do a demo ? 08:53:10 <mardim> with the slides 08:53:11 <mardim> ? 08:53:17 <gongysh> mardim, not yet. but we maybe need to do some demo setup 08:53:24 <gongysh> I think we will have to. 08:53:39 <mardim> gongysh, ok then if there is no deadline for the PPT 08:53:52 <mardim> gongysh, I will try to spinup the things regarding PPT 08:53:54 <gongysh> mardim, maybe you can start writing at google doc 08:53:57 <mardim> the next weeks 08:54:08 <gongysh> so we can co-work on it. 08:54:09 <mardim> gongysh, a google doc about what ? 08:54:28 <phuoc> gongysh, hi 08:54:38 <gongysh> about the openstack ppt 08:54:47 <gongysh> phuoc, hi 08:54:50 <mardim> ah I ve already have a google doc 08:54:54 <mardim> let me send you 08:56:06 <mardim> gongysh, Did you get the URL ? 08:56:27 <gongysh> got it 08:56:32 <mardim> gongysh, nice 08:56:38 <mardim> yes we can cowork it 08:56:43 <mardim> that will be nice !! 08:56:49 <gongysh> phuoc, I will start a project update slides 08:56:55 <gongysh> at google doc 08:57:13 <gongysh> so that we can co-work on it. 08:57:21 <phuoc> gongysh, okie 08:57:35 <phuoc> I will update on that doc too 08:57:45 <gongysh> anything else? 08:57:52 <haint_> https://bugs.launchpad.net/tacker/+bug/1760256 08:57:53 <openstack> Launchpad bug 1760256 in tacker "functional tests failed in creation ns and vnf_multiple_vdus" [Undecided,New] 08:57:53 <mardim> gongysh, no I am god 08:57:57 <mardim> good* 08:58:04 <haint_> we have some fails on functional tests 08:58:10 <haint_> should we fix it also 08:59:04 <gongysh> haint_, yes. of course we should. but it does not failed definitely. 08:59:19 <gongysh> so it is not easy to find a solution. 08:59:33 <gongysh> it seems it is due to the target vim's resource. 08:59:46 <haint_> and 1 more thing, http://paste.openstack.org/show/718239/ 08:59:48 <phuoc> sometime it passes I see 08:59:57 <haint_> how about this error? 09:00:05 <haint_> it from this patch set: https://review.openstack.org/#/c/556257/ 09:00:08 <gongysh> ok. time is up 09:00:15 <gongysh> #endmeeting