14:01:44 <bh526r> #startmeeting Weekly Technical Discussion 14:01:44 <collabot`> Meeting started Thu Mar 9 14:01:44 2017 UTC. The chair is bh526r. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:44 <collabot`> Useful Commands: #action #agreed #help #info #idea #link #topic. 14:01:44 <collabot`> The meeting name has been set to 'weekly_technical_discussion' 14:02:01 <bh526r> #topic Roll Call 14:02:06 <bh526r> #info Bin Hu 14:02:27 <radez> #info radez 14:02:29 <uli-k> #info Uli 14:02:34 <jmorgan1> #info Jack Morgan 14:02:35 <RandyLevensalor> #info Randy Levensalor 14:04:17 <bh526r> #topic New Project Proposal "Sample VNF" 14:04:35 <bh526r> #link https://wiki.opnfv.org/display/PROJ/Project+Proposal%3A+samplevnf 14:05:00 <bh526r> #info Kannan introduced the project proposal 14:09:15 <bh526r> #info The key is that not only the VNF catalog and VNF, but also the source code of VNFs 14:09:39 <bh526r> #info Uli said that it is very important to have VNFs for our testing 14:10:10 <bh526r> #info What is the difference between commercial VNF and sample VNF/testing VNF 14:11:00 <bh526r> #info Kannan responded that commercial VNF has production quality, but sample/testing VNF has feature-complete, but may not be production quality 14:11:50 <bh526r> #info So we focus on feature-complete, not on production quality 14:14:06 <bh526r> #info Brian gave an example - data plane complete, but may not complete in control plane such as HA 14:14:59 <bh526r> #info Uli suggested to clarify it in the project proposal in wiki, ie. feature complete in data plane, but may not complete in control plane 14:16:04 <bh526r> #info Kannan will add a sentence in wiki "VNF may not be deployable in real network environment" 14:17:00 <uli-k> I think we can be feature complete on data plane, have a good coverage even on user plane, but for control plane and management plane, we will implement only a minimum of features. 14:17:25 <bh526r> #info I think we can be feature complete on data plane, have a good coverage even on user plane, but for control plane and management plane, we will implement only a minimum of features. 14:18:54 <bjskerry> #info plan is for the license to be Apache or BSD 14:19:44 <bjskerry> #info will need to comply with OPNFV guidance 14:20:13 <bjskerry> #info currently the VNFs are Apache 2.0 14:20:38 <bjskerry> #info will be hosting complete source code 14:22:27 <Julien-z_> #info Julien 14:26:22 <bh526r> #info Bryan wanted to clarify that we don't intend to test the VNF itself for its real conformance, e.g. with any standards. 14:28:16 <bh526r> #info Kannan said performance testing is our of scope. 14:28:30 <bjskerry> #info example of vPE, we won't test conformance to the standard, but could use them for end to end testing, including SFC if applicable 14:30:09 <bh526r> #info Dave said it is very vital project, and key to moving forward to next stage 14:32:02 <bh526r> #info add two sentences (1) no conformance testing (2) no field deployment 14:33:40 <Julien-z_> is this project focus only on Data Plane VNF? 14:34:01 <Julien-z_> shall we can the rely on DPDK? 14:35:02 <bh526r> #info Morgan wanted more community support from other companies and other contributors 14:35:30 <bh526r> #info People are encouraged to join the project 14:35:39 <xuanjia> Does this project provide docker container based VNF ? 14:36:25 <Julien-z_> currently vIMS is used in Funtect 14:36:36 <Julien-z_> can this project include vIMS? 14:37:03 <Julien-z_> bh526r, voice of quality is not good enough. I give my response here 14:37:35 <Julien-z_> Parser project is working on deploy VNF with tosca, and which can consume sampleVNFs to deploy 14:39:02 <bh526r> #info Group achieved consensus to recommend for TSC creation review by addressing 3 issues (1) clarify feature-completeness v.s. production quality, and for testing purpose instead of commerical purpose (2) no conformance testing (3) no field deployment 14:40:10 <bh526r> #info Apache License will be used. 14:41:49 <bh526r> #info Kannan clarified that it is not limited to DPDK. Initial implementation is based on DPDK. But other contributions without DPDK is very welcome 14:42:16 <bh526r> #info And of course, it includes vIMS. And contributions are very welcome. 14:45:22 <bjskerry> #info Code for SampleVNFs does not have to be hosted by OPNFV, ie if source code is available elswhere it could be used 14:45:40 <Julien-z_> good suggestion uli-k 14:46:19 <Julien-z_> it can work either with or without DPDK or any other limits 14:46:29 <Julien-z_> then , it will be good 14:48:48 <bh526r> #topic MANO WG Update 14:50:53 <rprakash> #info rprakash 14:51:22 <rprakash> https://wiki.opnfv.org/pages/viewpage.action?pageId=6827111 14:51:38 <bh526r> #link https://wiki.opnfv.org/display/mano/MANO+Group+Home 14:59:47 <bh526r> #info MANO WG has done a great job, performed its operation regularly, and set a series of action items to address scenarios, use cases, etc. 15:01:10 <bh526r> #endmeeting