Been typing during breaks in a 300+ mile trip. I was more speaking to the adverse impacts of v4 redirects BUT.....
Unless you think there is some awesome reason that people are going to need redirects that have multiple routers on the shared lan WITH hosts that are not speaking some dynamic routing protocol with said routers, redirects are unnecessary and promote bad architecture. If that is your goal, then I would understand what is going on here. I've seen a lot of poorly architected networks that have nearly broken or melted down due to abusing something that is feasible in this space. I'm not sure what the true use case is for redirects here. Jared Mauch On Aug 13, 2010, at 12:07 PM, "Hemant Singh (shemant)" <[email protected]> wrote: > -----Original Message----- > From: [email protected] [mailto:[email protected]] On Behalf Of > Jared Mauch > Sent: Friday, August 13, 2010 11:55 AM > To: Randy Bush > Cc: ipv6 deployment prevention > Subject: Re: Router redirects in Node Requirements document > >> Agreed. Anyone wanting to do this is not connected with actual > operations and should be questioned as to what the operational > requirements are. Accepting and sending redirects also opens up dos > vectors >to devices when implemented poorly. > > If you have a problem with ND Redirect being a DOS vector, then you > should have raised an issue with RFC 4861. Anyway, RFC 4861 is clearly > aware of the DOS vector with Redirect, because RFC 4861 mentions > Redirect rate limiting in section 8.2. DOS vector is a very weak reason > to shoot down "Redirect functionality MUST be implemented by a router in > the Node Req document". > > [A router SHOULD send a redirect message, subject to rate limiting,] > > Hemant -------------------------------------------------------------------- IETF IPv6 working group mailing list [email protected] Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
