==================================== #opnfv-sfc: OPNFV SFC weekly meeting ==================================== Meeting started by ebrjohn at 14:03:44 UTC. The full logs are available at http://ircbot.wl.linuxfoundation.org/meetings/opnfv-sfc/2015/opnfv-sfc.2015-09-09-14.03.log.html . Meeting summary --------------- * Tim Rozet (trozet, 14:04:36) * Chris Price (IRC only) (ChrisPriceAB, 14:04:55) * ebrjohn Brady Johnson (ebrjohn, 14:06:03) * rprakash (rprakash, 14:06:18) * Bryan Sullivan (bryan_att, 14:06:27) * pcarver Paul Carver AT&T (pcarver, 14:06:35) * sridhar_ram Sridhar Ramaswamy (Tacker) (sridhar_ram, 14:06:46) * yuyang (yuyang, 14:07:49) * iben rodriguez spirent (iben__, 14:07:52) * Eric Multanen (snackewm, 14:08:07) * will take notes (rprakash, 14:08:35) * LINK: https://www.ietf.org/meeting/94/ IETF 94 hackfest (ebrjohn, 14:09:06) * OPNFV SFC does not need more than normal installation (rprakash, 14:11:32) * except for two dependencies Tacker and specific type of OVS (rprakash, 14:12:12) * Paul Quin not available leaving NSH item open for now (rprakash, 14:12:57) * add to agenda starting aVM with VXLAN (rprakash, 14:14:03) * Ference on SFC Architecture (rprakash, 14:15:03) * Ferenc is showing how OpenStack starts a VM with vxlan (ebrjohn, 14:15:37) * LINK: https://docs.google.com/presentation/d/1MRlvs-49PO2ZrYJdZ1UVAdl5FYPSUCTd-PR2lCkiPtQ/edit#slide=id.gc7f034de0_0_29 How OpenStack starts a VM with vxlan (ebrjohn, 14:17:03) * Control & Data plane with management and overlay plan networks (rprakash, 14:17:35) * currently using IP/MPLS header and when NSH ready OVS is available will swicthn over to the same (rprakash, 14:25:34) * OpenStack uses creates network for tenant isolation using VXLAN that may conflict with ODLwas the question (rprakash, 14:27:59) * ODL uses for vm on host to another vm on to another through br-int through tun and thus uses different VNI from Tenant allocated by Openstack and hus avoiding conflict (rprakash, 14:30:22) * will need to be verified and or live with this limitation for ghis release (rprakash, 14:30:58) * sridhar can you pound in the link? (rprakash, 14:33:21) * LINK: https://docs.google.com/presentation/d/18AGaiysVgHOd_fIHVpObMO7zUkMjJGOQ98CUwkxU1xo/edit?usp=sharing (sridhar_ram, 14:35:01) * function without NSH will need code change what Sam had presented using VRM (rprakash, 14:35:33) * Dan has started simple scripts to bring VM inside Openstack using intercepting service function and send it to ODL which Takcer will us in short term (rprakash, 14:39:01) * This will help flush out the flows to test using Tacker (rprakash, 14:40:12) * slide 5 on Tacker slides shows the SFC API being used through RESTful API (rprakash, 14:42:58) * Few people from Yardstick reached out for testing and wanted to know about Infra to ensure that they get ready for testing (rprakash, 14:44:42) * Basic Pharos setup of 3 or 5 nodes should suffice no need for HA for now (rprakash, 14:47:04) * NSH release being OVS 2,5 is a concern and not Openstack dependencies or changes (rprakash, 14:49:33) * ACTION: Hose to send email to list for starting efforts for testing sfc (rprakash, 14:50:40) * ACTION: Joseto send email to list for starting efforts for testing sfc (rprakash, 14:51:08) * what kid of tests are available now? (rprakash, 14:51:30) * currently only Tempest is used thus only a Platform API level plus some for ODL Robo (rprakash, 14:52:44) * Func_Test uses Robo and thus needs only scenerios like VM creating , chaining etc to help testing (rprakash, 14:54:12) * SF C will be an intresting use case for Func_Test (rprakash, 14:54:56) * where does the test start which lab and transitions to which lab? (rprakash, 14:55:40) * LF lab and Ericsson Lab or any community lab and in end JJB will rn SFC from LF (rprakash, 14:56:56) * Jira Entry needs to be completed for B release and so high level story needs some enry for meeting Milestone (rprakash, 14:58:38) * ACTION: Brady Get started on JIRA. Make it a topic on next week's meeting agenda. (ebrjohn, 14:59:14) * its Iben (rprakash, 15:00:02) * ACTION: an irc meeting or ether pad entry will help close the gap for Jira entry (rprakash, 15:01:14) Meeting ended at 15:01:55 UTC. People present (lines said) --------------------------- * rprakash (34) * ebrjohn (12) * collabot (5) * iben__ (3) * sridhar_ram (3) * ChrisPriceAB (3) * yuyang (2) * ethfci (1) * bryan_att (1) * pcarver (1) * snackewm (1) * trozet (1) Generated by `MeetBot`_ 0.1.4