Thank you so much with kind comments! :-) I can clarify the concept of scenario.
---------------------------------------------------------- Chunghan Lee Cloud Network Project. Fujitsu Laboratories Ltd. 4-1-1 Kamikodanaka, Nakahara-ku, Kawasaki-shi, Kanagawa 211-8588, Japan E-mail： lee.chung...@jp.fujitsu.com<mailto:lee.chung...@jp.fujitsu.com> ---------------------------------------------------------- From: O Mahony, Billy [mailto:billy.o.mah...@intel.com] Sent: Thursday, July 28, 2016 10:58 PM To: Lee, Chunghan/李 忠翰; opnfv-tech-discuss@lists.opnfv.org Subject: RE: [ovsnfv] Some questions From: Lee, Chunghan [mailto:lee.chung...@jp.fujitsu.com] Sent: Wednesday, July 27, 2016 5:23 AM To: O Mahony, Billy <billy.o.mah...@intel.com<mailto:billy.o.mah...@intel.com>>; opnfv-tech-discuss@lists.opnfv.org<mailto:opnfv-tech-discuss@lists.opnfv.org> Subject: RE: [ovsnfv] Some questions Hi Billy, Thank you for e-mail. > That file is yardstick/tests/opnfv/test_suites/opnfv_ericsson-pod1_daily.yaml. I confirmed the yaml file. Thank you for the location. > The term ‘scenario’ is used to define an OPNFV deployment. Usually something > like os-ha-nosdn-ovs would mean openstack in high availability mode with no > SDN controller and the ‘ovs’ feature >(i.e DPDK enabled OVS) turned on. > Scenarios are also per installer so it’s implied that the scenario is for > Fuel or Apex or Joid etc eventhough that is not mentioned in the scenarios > name. >So CI is run against certain scenario’s (ie. not every single combination of >every installer, mode, controller & feature is tested). Thank you for the detail explanation. Some unclear points are clear. For example, the ovsnfv scenario for Colorado will be same per the installer (Fuel or Apex) although the implementation method for scenario is different per the installer. My understanding is correct ? [[BO'M]] Yes it’s the same scenario and shares (As far as I know) the same scenario files. It’s up to the CI jobs to run the right scripts based on the installer which interpret the scenario.yaml files (for example Fuel as a deploy.sh script that does this). Other installers will have similar scripts but I’m not familiar with them as OVSNFV only dealt with Fuel for Rel. B. > The job is a little complicated as it must run with neutron VLAN segmentation > rather than the usual VxLAN tunneling. Ok. I can understand it. It means that the ovsnfv scenario for Colorado is based on VLAN segmentation. Is it correct ?? [[BO'M]] Yes OVS-DPDK feature of Fuel9 only support VLAN so the scenario will have to ensure that the CI pod is set up with VLAN. Regards, Chunghan Lee ---------------------------------------------------------- Chunghan Lee Cloud Network Project. Fujitsu Laboratories Ltd. 4-1-1 Kamikodanaka, Nakahara-ku, Kawasaki-shi, Kanagawa 211-8588, Japan E-mail： lee.chung...@jp.fujitsu.com<mailto:lee.chung...@jp.fujitsu.com> ---------------------------------------------------------- From: O Mahony, Billy [mailto:billy.o.mah...@intel.com] Sent: Monday, July 25, 2016 6:03 PM To: Lee, Chunghan/李 忠翰; opnfv-tech-discuss@lists.opnfv.org<mailto:opnfv-tech-discuss@lists.opnfv.org> Subject: RE: [ovsnfv] Some questions Hi Chunghan, That file is yardstick/tests/opnfv/test_suites/opnfv_ericsson-pod1_daily.yaml. The term ‘scenario’ is used to define an OPNFV deployment. Usually something like os-ha-nosdn-ovs would mean openstack in high availability mode with no SDN controller and the ‘ovs’ feature (i.e DPDK enabled OVS) turned on. Scenarios are also per installer so it’s implied that the scenario is for Fuel or Apex or Joid etc eventhough that is not mentioned in the scenarios name. So CI is run against certain scenario’s (ie. not every single combination of every installer, mode, controller & feature is tested). I’m sure you’ve seen https://build.opnfv.org/ci/ which details all the job, results etc. We expect a CI job for OVSNFV scenario to land shortly. The job is a little complicated as it must run with neutron VLAN segmentation rather than the usual VxLAN tunneling. Cheers, Billy. From: Lee, Chunghan [mailto:lee.chung...@jp.fujitsu.com] Sent: Monday, July 25, 2016 6:16 AM To: O Mahony, Billy <billy.o.mah...@intel.com<mailto:billy.o.mah...@intel.com>>; opnfv-tech-discuss@lists.opnfv.org<mailto:opnfv-tech-discuss@lists.opnfv.org> Subject: RE: [ovsnfv] Some questions Hi Billy, Thank you for your reply. >I’ve given some summary above but to understand fully what the tests do you’ll >have to read the yardstick code. I checked the yardstick documents and the source codes of yardstick yaml files, and can understand which benchmark programs are executed briefly. > Afaik, those tests were chosen as the standard daily suite for the scenarios > deployed by Fuel. And defined in test suite opnfv_ericsson-pod1_daily.yaml I cannot find the above yaml file… Sorry where is it ? # I searched the follow link (https://git.opnfv.org/cgit/ovsnfv/tree/), but I cannot find it. > The Brahmaputra CI for ovsnfv scenarios has been discontinued - as it’s no > longer under development. The CI for the ovsnfv scenario for Colorado should > >be coming on line soon. In my understanding, the CI is the ovsnfv scenarios that describes the procedures of yardstick testcases. Am I wrong ? If there are the wrong points, please give me comments. Regards, Chunghan Lee ---------------------------------------------------------- Chunghan Lee Cloud Network Project. Fujitsu Laboratories Ltd. 4-1-1 Kamikodanaka, Nakahara-ku, Kawasaki-shi, Kanagawa 211-8588, Japan E-mail： lee.chung...@jp.fujitsu.com<mailto:lee.chung...@jp.fujitsu.com> ---------------------------------------------------------- From: O Mahony, Billy [mailto:billy.o.mah...@intel.com] Sent: Wednesday, July 20, 2016 8:00 PM To: Lee, Chunghan/李 忠翰; opnfv-tech-discuss@lists.opnfv.org<mailto:opnfv-tech-discuss@lists.opnfv.org> Subject: RE: [ovsnfv] Some questions Hi Chunghang, Some answeres in-line below for you. Hope that helps, /Billy. From: opnfv-tech-discuss-boun...@lists.opnfv.org<mailto:opnfv-tech-discuss-boun...@lists.opnfv.org> [mailto:opnfv-tech-discuss-boun...@lists.opnfv.org] On Behalf Of Lee, Chunghan Sent: Wednesday, July 20, 2016 3:44 AM To: opnfv-tech-discuss@lists.opnfv.org<mailto:opnfv-tech-discuss@lists.opnfv.org> Subject: [opnfv-tech-discuss] [ovsnfv] Some questions Hi Mark and billy, I’d like to ask some unclear points in my understanding. Could you give me answer ? 1. I’d like to understand our micro benchmark results clearly. I found the Yardstick document from the follow link. (http://artifacts.opnfv.org/yardstick/brahmaputra/docs/yardstick/03-list-of-tcs.html) Is it correct ? Currently, test cases (TC001,2,8,9,10,12,37.39) are available only ? In my understanding, only 7 tests are passed only. [[BO'M]] There were only ever 7 yardstick test executed for the ovsnfv scenario which all pass TC002-Ping TC005-File i/o TC010-Mem read latency TC011-Packet delay variation TC012-Mem r/w TC014-CPU Speed TC037-pktgen Afaik, those tests were chosen as the standard daily suite for the scenarios deployed by Fuel. And defined in test suite opnfv_ericsson-pod1_daily.yaml I’ve given some summary above but to understand fully what the tests do you’ll have to read the yardstick code. 2. From logs (14:35:54 11th July 2016), there is our scenario. What is our scenario ? Could you give me a link or the information related to the scenario ? Sorry… I cannot understand the scenario. 3. Is it Our CI link (https://build.opnfv.org/ci/view/ovsnfv/) ? [[BO'M]] The Brahmaputra CI for ovsnfv scenarios has been discontinued - as it’s no longer under development. The CI for the ovsnfv scenario for Colorado should be coming on line soon. Regards, Chunghan Lee ---------------------------------------------------------- Chunghan Lee Cloud Network Project. Fujitsu Laboratories Ltd. 4-1-1 Kamikodanaka, Nakahara-ku, Kawasaki-shi, Kanagawa 211-8588, Japan E-mail： lee.chung...@jp.fujitsu.com<mailto:lee.chung...@jp.fujitsu.com> ----------------------------------------------------------
_______________________________________________ opnfv-tech-discuss mailing list opnfv-tech-discuss@lists.opnfv.org https://lists.opnfv.org/mailman/listinfo/opnfv-tech-discuss
