|
Hi Rob, > In summary, unless there is some nuance to your question that I misunderstand, I don’t think that Cisco, or anyone else, is copying your ideas on composite metrics. I think that the
basic idea of using composite metrics have been known in the industry for a long time. The composite metric that eigrp uses depends on the following: 1)
Bandwidth. 2)
Load. 3)
Delay. 4)
Reliability. 5)
MTU. So, there is no factor “number of hops” taken into consideration in the metric calculation until this was added in the new EIGRP version that you call “named mode EIGRP” with wide metric that includes K6 (number
of hops) as an addition to the metric calculation. This was something new introduced by the Numbering Exchange Protocol (NEP) ID and presented on the IETF 101 remotely, and your colleague asked what are the benefits of adding this factor to the metric, so it
wasn’t something on EIGRP and it was something described on the new IGP NEP. Regards, Khaled Omar From: Rob Wilton (rwilton) <rwilton@xxxxxxxxx> Hi Khaled, I’m no expert on EIGRP or your NED draft, but note that EIGRP is documented in RFC 7868, that was published as an informational RFC on May 2016. According to datatracker, the first EIGRP draft submitted to IETF appears
to be 18th Feb 2013. Both the final RFC and initial draft appear to document composite metrics including the K6 coefficient that is part of Wide Metric support. The first revision of your NEP draft was submitted on Jun 2017, which seems to suggest that EIGRP supported composite metrics for some number of years before your first internet draft submission on NED. But it is generally patents that protect ideas rather that copyright, and a quick patent search reveals US5088032A filed by Cisco in 1988 which describes the use of composite routing metrics: <snip> The metric information includes the topological delay time for a transmission, the path bandwidth for the narrowest bandwidth segment of the path,
the channel occupancy of the path, and a count of the number of gateway circuits through which the path runs (the "hop count"). Based on this metric information, a single "composite metric" is calculated for the path. </snip> In summary, unless there is some nuance to your question that I misunderstand, I don’t think that Cisco, or anyone else, is copying your ideas on composite metrics. I think that the basic idea of using composite metrics
have been known in the industry for a long time. Hopefully this answers your question. Regards, From: ietf <ietf-bounces@xxxxxxxx>
On Behalf Of Khaled Omar Hi all, I would like to ask why ideas submitted into ietf drafts are applied without a notice to the draft authors ?!!!
Before NEP ID, there was no IGP uses a composite metric such as NEP uses, and now I can see that cisco added a third factor to its eigrp metric calculation (they call it K6) and became as NEP ID. The funny thing is that during NEP presentation at the ietf meeting someone from cisco asked “what are the benefits to add this factor to the metric”, and then you see this factor added to their IGP. Actually, I received no notice from cisco before applying such thing to their IGP. Can someone tells me what this means or what to call this? Regards, Khaled Omar |
