On 10/28/2010 21:55 EDT, Tony Li wrote: > > Hi Hannu, > > Thanks for the comments. > > >> 1. Section 3.1 Improved routing scalability >> >> You talk about "the control plane", but it is unclear what is the >> Internets control plane? RFC 1958 that was referred in section 2 doesn't >> mention anything about the control plane. Is FIB in your definition on >> control plane? So, do you think that the solution should support the >> scalability of FIBs? RIB is in my opinion on the routing control plane? >> But what about the DNS? Is flow control on the control plane that the >> document addresses? > > > Ok, thanks. I'll expand on this a bit.
Elsewhere I've used phrases like "control plane transport" for control plane infrastructure such as data paths to exchange routing information. But DNS etc. are different, since both the control and forwarding planes can function without them. I like the phrase "infrastructure services". >> I do not think that there is a consensus whether mobility is part >> of the global routing system. Why should it be? > > Actually, it should not be. However, it is a function of the routing > and addressing architecture. This is exactly the bug that we're > trying to fix. How mobility is done depends on and is shaped by how routing and addressing are done. Some routing and addressing approaches make mobility simple to solve. None of them make it truly hard. (See also the privacy issues with putting too much support for mobility in routing and addressing.) Scott _______________________________________________ rrg mailing list [email protected] http://www.irtf.org/mailman/listinfo/rrg
