Hi Matt, Thanks for commenting, my initial thoughts are inline below in blue:
- The first is presenting a clear "end-to-end" test scenario. While the past two plugfests have been great for pluggables and NOS vendors, I believe our ultimate end goal is certifying from NIC to pluggable to chassis to NOS to L3 routing protocols and vice versa. The membership proposal should outline this as a future goal, but the correct place for this ultimately (in any level of detail )is in the test plan itself, not in the membership proposal. However we should make reference to the test plan in the membership proposal, and that's missing right now, I'll fix that. - Second is full holding strong transparency & privacy - such as all testing scripts, procedures, and configuration files will be made publicly available. However in the case of a failed scenario, the involved vendors would receive remediation time (say 2 months) to re-test before results went public. For example, we've seen a few NOS vendors not correctly program pluggable transceivers - usually this is due to a NOS error of some sort, doesn't seem fair for a pluggable vendor to be advertised as failing due to a NOS vendors problem. Agree on the transparency for scripts, procedures, config files etc...I will add that to the proposal, its an important point. Results (I believe) should be treated differently however. I'm wary of publicly listing failures, but an open discussion here is worthwhile. A couple points to consider - we want our public list to show what's working. If a product (or certain combination of products) isn't listed, its already a red flag that its either a) not yet tested b) failed. Either way there's work to be done. - We want people to look at this list/matrix as evidence of what works, not scouring it looking for the failures. - the possibility that failures may be publicly aired, may become a deterrent for companies to even participate. - not listing the failures eliminates the last issue you raise, i.e. a module is listed as failing because a NOS programmed it incorrectly - A clause or some wording around the use of press - generally UNH-IOL events and activities have strict rules against any participant from doing media capture (video, photos, etc). This makes sense for a traditional consortium that may have pre-release or pre-production equipment, however in the case of Open Networking - many of the products are GA and existing mass market. It seems incredibly important for us (under the OCP charter) to continue education in the industry, presenting media is crucial to this effort. Obviously members could opt out if requested, but generally I'd like to see something a bit more flexible then "request UNH-IOL staff to operate the camera, only group shots, no close-ups" norm. This could be worded under the Public Relations section. Ahhh! Change scares me! Just kidding. Sort of. I'd like to hear other opinions on this as well. I get that education is important, and we need to have a way to enable that. Speaking for myself, much of my role at UNH-IOL has been ensuring confidentiality and protecting confidential information about products at plugfests and in our test beds. So, going in the other direction is disconcerting for me, but I recognize that the Open Networking community is unique in this regards. I see the value in opening this up, but I do think their should be some limits in place, but relax those limits from where were are today (the 'no close ups' norm you mentioned). Do you have any proposed wording for this? I'll wait a few more days for further comment and feedback and then provide an updated membership proposal. With these ideas incorporated. thanks! David On Tue, Jul 14, 2015 at 1:16 PM, Matt Peterson <m...@peterson.org> wrote: > Carlos, > > I would like to see3 specific items addressed in the proposal. > > > - The first is presenting a clear "end-to-end" test scenario. While > the past two plugfests have been great for pluggables and NOS vendors, I > believe our ultimate end goal is certifying from NIC to pluggable to > chassis to NOS to L3 routing protocols and vice versa. > > > - Second is full holding strong transparency & privacy - such as all > testing scripts, procedures, and configuration files will be made publicly > available. However in the case of a failed scenario, the involved vendors > would receive remediation time (say 2 months) to re-test before results > went public. For example, we've seen a few NOS vendors not correctly > program pluggable transceivers - usually this is due to a NOS error of some > sort, doesn't seem fair for a pluggable vendor to be advertised as failing > due to a NOS vendors problem. > > > - A clause or some wording around the use of press - generally UNH-IOL > events and activities have strict rules against any participant from doing > media capture (video, photos, etc). This makes sense for a traditional > consortium that may have pre-release or pre-production equipment, however > in the case of Open Networking - many of the products are GA and existing > mass market. It seems incredibly important for us (under the OCP charter) > to continue education in the industry, presenting media is crucial to this > effort. Obviously members could opt out if requested, but generally I'd > like to see something a bit more flexible then "request UNH-IOL staff to > operate the camera, only group shots, no close-ups" norm. This could be > worded under the Public Relations section. > > --Matt > > On Mon, Jul 13, 2015 at 11:44 AM, Carlos Cardenas < > car...@cumulusnetworks.com> wrote: > >> Howdy, >> >> As a follow up to today's call on the Open Network Consortium Testing >> section, the meeting minutes have been updated to reflect the link to the >> Membership Proposal document (link here as well: >> http://files.opencompute.org/oc/public.php?service=files&t=10b85b2c499f52fe77c3d75e89115759 >> ). >> >> For those that are interested, please review and provide feedback no >> later than 31 July. >> >> Thanks. >> >> +--+ >> Carlos >> >> _______________________________________________ >> opencompute-networking mailing list >> Unsubscribe: >> http://lists.opencompute.org/mailman/options/opencompute-networking >> >> opencompute-networking@lists.opencompute.org >> http://lists.opencompute.org/mailman/listinfo/opencompute-networking >> >> > > _______________________________________________ > opencompute-networking mailing list > Unsubscribe: > http://lists.opencompute.org/mailman/options/opencompute-networking > > opencompute-networking@lists.opencompute.org > http://lists.opencompute.org/mailman/listinfo/opencompute-networking > >
_______________________________________________ opencompute-networking mailing list Unsubscribe: http://lists.opencompute.org/mailman/options/opencompute-networking opencompute-networking@lists.opencompute.org http://lists.opencompute.org/mailman/listinfo/opencompute-networking
