========================================= #opnfv-meeting: Colorado planning meeting ========================================= Meeting started by ChrisPriceAB at 15:08:05 UTC. The full logs are available at http://ircbot.wl.linuxfoundation.org/meetings/opnfv-meeting/2016/opnfv-meeting.2016-03-22-15.08.log.html . Meeting summary --------------- * rprakash will help (rprakash, 15:09:14) * Roll Call (ChrisPriceAB, 15:09:15) * Bryan Sullivan (bryan_att, 15:09:22) * Maryam Tahhan (VSPERF, SFQM) (mtahhan, 15:09:22) * Bin Hu (bin_, 15:09:24) * Frank Brockners (frankbrockners, 15:09:28) * Chris Price (ChrisPriceAB, 15:09:28) * Ana Cunha (Yardstick) (anac1, 15:09:35) * Morgan Richomme (morgan_orange, 15:09:36) * Fatih Degirmenci (Releng) (fdegir, 15:09:41) * Tim Irnich (SDN VPN) (timirnich, 15:09:41) * Uli Kleber, Octopus (uli-k_, 15:09:45) * Bryan Sullivan (Copper, Models) (bryan_att, 15:09:46) * RayNugent (RayNugent_, 15:09:54) * Ashlee Young Networkfw (rprakash, 15:10:34) * Tim Rozet (Apex) (trozet, 15:10:36) * Mark Beierl (StorPerf) (mbeierl, 15:10:49) * Colorado Planning (ChrisPriceAB, 15:11:36) * LINK: https://wiki.opnfv.org/releases/colorado current approved Colorado plan (ChrisPriceAB, 15:12:18) * Should we include the new release manager in these planning meetings? (RayNugent_, 15:14:58) * LINK: https://wiki.opnfv.org/releases/colorado/projects list of participating projects for colorado (ChrisPriceAB, 15:15:02) * Hack-Fest retrospective follow up (ChrisPriceAB, 15:18:30) * David outlines the feedback that the test development and framework development happened in parallel with test cases in Brahmaputra. (ChrisPriceAB, 15:19:33) * David raises the questions on if we would be able to close on the test infrastructure during Colorado. (ChrisPriceAB, 15:20:36) * Capbity ofr feature tesing should be listed early like MS1 rther than Ms2 (rprakash_, 15:23:34) * Capbity for feature tesing should be listed early like MS1 rther than Ms2 (rprakash_, 15:24:03) * Framk says that we now have a test framework like Yardstick, Functest... (rprakash_, 15:24:36) * Frank says that we now have a test framework like Yardstick, Functest... and even Installers are ready (rprakash_, 15:25:11) * Morgan says create a stable scenerio with features declaraton and tickets for steering to correct testing project (rprakash_, 15:27:12) * Selecting test cases accordingly and testing what is proven to work is paramount (fdegir, 15:28:09) * Challenging but need to define CI test objective as well Feature test Objectives seperated and also frequenecy like weekly monthly tests etc. (rprakash_, 15:28:48) * There is no point in testing everything on Baremetal is not necessary (rprakash_, 15:29:19) * Tests Strategy must agree on mimimum vs exhaustive testing and schedule accordingly (rprakash_, 15:30:05) * Relaying too much on Installer development is a sore point for some projects (rprakash_, 15:30:40) * Relying too much on Installer development is a sore point for some projects (rprakash_, 15:31:04) * Creating environment in CI so that Developers can test independent for minimal testing (rprakash_, 15:31:52) * We need an OPNFV as a service project for testing from Cloud needs to be introduced and Bryan suggests Ravello as a showcase to see its viability (rprakash_, 15:33:44) * Fathi to review if OPNFV virtual environment using BP can be established uisng Ravello for VNF (rprakash_, 15:35:37) * Ravello is running on Mutiple servers and has emulated IPMI as per Bryan and Frank mentions that its a paid service and verify nested virtualization in LF (rprakash_, 15:37:01) * Ravello is just a trial and if we can do it in LF then we will be in control (rprakash_, 15:37:39) * TSC call to adress broad community colaboration issue (rprakash_, 15:42:20) * Bryan can demo Ravello session in next 2 weeks (rprakash_, 15:43:42) * Weekly test co-ordination call should address test co-ordination plan - Trevor can be assigned th Action Item (rprakash_, 15:45:08) * need to define detail to seperate Dvelopment and Production testing (rprakash_, 15:46:03) * nextweeks call we can discuss How do we allow individual Scenerios to allow testing like immediate dependencies versus actual targgetted (rprakash_, 15:48:59) * eg starting with Liberty but ending with Mitaka etc. (rprakash_, 15:49:24) * to cover upstream feedback etc so we should start discussions at next meeting (rprakash_, 15:50:14) * Frank is documenting and will bring it to next meeting (rprakash_, 15:50:38) * am going to end the meeting (rprakash_, 15:51:00) Meeting ended at 15:51:24 UTC. People present (lines said) --------------------------- * rprakash_ (29) * ChrisPriceAB (17) * fdegir (8) * collabot (8) * bryan_att (7) * frankbrockners (3) * rprakash (3) * RayNugent_ (3) * morgan_orange (2) * uli-k_ (2) * trozet (2) * mtahhan (1) * anac1 (1) * timirnich (1) * bin_ (1) * mbeierl (1) * davidmcbride (0) * HKirksey (0) Generated by `MeetBot`_ 0.1.4