I'm glad that we are pulling this discussion back to technical track. Some additional comments about QTIP inline.
On Thu, Jan 5, 2017 at 7:09 AM Cooper, Trevor <trevor.coo...@intel.com> wrote: I am glad we will discuss this in the Test Working Group meeting tomorrow … the testing projects have evolved quite a bit since the first plugfest. We should evolve wherever possible to normalize and avoid duplication BUT also give projects autonomy they deserve to decide their interests and drive their objectives. This seems to be what is emerging? Functional testing VIM + NFVI -> Functest MANO -> ? Compliance -> Dovetail Performance testing NFVI (+ VNF) characterization -> Yardstick Compute, memory, storage sub-system (platform) performance -> Qtip This is the major application of QTIP. A synthesis evaluation of platform performance by benchmarking test. As a companion delivery, QTIP is also developing a platform for performance benchmarking, including common utility modules, base templates and benchmark result sharing (possibly on opnfv testresult site) Data-plane performance (switching technology, virtual and physical networks) -> VSPERF Storage performance -> Storperf Controller performance -> CPerf Performance trend analysis -> Bottleneck /Trevor *From:* Yujun Zhang [mailto:zhangyujun+...@gmail.com] *Sent:* Wednesday, December 21, 2016 6:36 AM *To:* Gaoliang (kubi) <jean.gaoli...@huawei.com>; test-wg < test...@lists.opnfv.org> *Cc:* TECH-DISCUSS OPNFV <opnfv-tech-discuss@lists.opnfv.org>; Morgan Richomme (morgan.richo...@orange.com) <morgan.richo...@orange.com>; Jose Lausuch <jose.laus...@ericsson.com>; Cooper, Trevor <trevor.coo...@intel.com>; mark.bei...@emc.com; Yuyang (Gabriel) <gabriel.yuy...@huawei.com> *Subject:* Re: [test-wg] [all] TestPerf EcoSystem diagram now editable Thanks for the detail explanation. As many test-wg members are on holiday, I think we may not able to arrive at a consensus on the process soon. The modification is actually not affecting any other projects but yardstick and QTIP. It reflects my understanding of QTIP projects, for the past and also for the future. And I didn't see any objection on the modification content (not the process) in the past week. Reverting it to previous version will mislead the community. So I would like to keep the modified version unless we have a scope change requested from the working group and approved by TSC. Do you think it will work? We may discuss about process in testperf meeting. But to my understanding, community wiki is open for editing unless locked. This is a way to show open spirit and improve efficiency. Of course that people who abuse this right shall be banned. My two cents. On Wed, Dec 21, 2016 at 4:50 PM Gaoliang (kubi) <jean.gaoli...@huawei.com> wrote: Hi Yujun, To my knowledge, the Draft Testperf Ecosystem Diagram was discussed and published at the first plugfest. Its orientation is to avoid overlap works between our testing projects and also to reflect the relations. Then after a lot of discussion within the community which leads to common consensus, the Diagram was updated during the OPNFV Berlin Summit (please refer to the presentation "Conversation with the Testing community"[1]). Now, the Updated Diagram has been published in WIKI page for 6 months and perceived stable community-wide. I believe if some modifications of the Diagram are expected, we should discuss first, then we do the changes. After all, we are working as a community. In addition, since any change to the Diagram may result in different relations between our testing projects, direct changes to the original Diagram will make things more complicated. It's better to have our discussion based on the summit version of our Testperf Ecosystem Diagram for changes. I'd like to work together with you to improve the modification process and keeping things in order. For this purpose, I would like help add a topic in the Testperf Weekly Meeting. I believe we can make further progress together! [1] https://gerrit.opnfv.org/gerrit/gitweb?p=functest.git;a=blob;f=docs/com/img/OPNFV_testing_group.png;h=4b8c2c053e0143b1a9abc7e54fd9e7671ccfcee8;hb=HEAD Thanks, Kubi *发件人**:* Yujun Zhang [mailto:zhangyujun+...@gmail.com] *发送时间**:* 2016年12月16日 16:43 *收件人**:* Gaoliang (kubi); test-wg *抄送**:* TECH-DISCUSS OPNFV; Morgan Richomme (morgan.richo...@orange.com); Jose Lausuch; trevor.coo...@intel.com; mark.bei...@emc.com; Yuyang (Gabriel) *主**题**:* Re: [test-wg] [all] TestPerf EcoSystem diagram now editable See my replies inline. On Fri, Dec 16, 2016 at 3:43 PM Gaoliang (kubi) <jean.gaoli...@huawei.com> wrote: Hi Yujun, Thanks for your timely reply! First, it is a very praiseworthy effort of you on making the diagram editable! What I concerned is the process of the modification of the Testperf Ecosystem Diagram. As you know, Testperf Ecosystem Diagram is not only to show what the one project will do , but also the relationship between each projects. It is on the Testing Work Group Home Page and shows the consensus to the whole community. Another important function of this diagram is that It could help the test projects to avoid overlap. *We shouldn’t modify the Ecosystem Diagram before consensus. *Otherwise it may delivers misleading information to the whole community. I suggest to discuss the modification process of our Testperf Ecosystem Diagram in our testperf weekly meeting. Before the discussion taking place, *we’d better go back to the structure[1] of ecosystem on which we have consensus *and keep it editable. Currently, what is in my mind about the process is that maybe a proposal first, then we could discuss together and decide how to make the modification. [1] https://wiki.opnfv.org/download/attachments/2926690/OPNFV_testing_group.png?version=1&modificationDate=1467636653000&api=v2 Regards, Kubi This is exactly my motivation of the modification. I hope we can find out the original agreement so we can know why it is delivering a misleading message and we can work together to improve the process to keep things in order.
_______________________________________________ opnfv-tech-discuss mailing list opnfv-tech-discuss@lists.opnfv.org https://lists.opnfv.org/mailman/listinfo/opnfv-tech-discuss
