On Mar 27, 2014, at 2:24 PM, Christopher Morrow <[email protected]> wrote:
> On Thu, Mar 27, 2014 at 11:22 AM, joel jaeggli <[email protected]> wrote: >> I think when were were discussing this in the w.g. that it has the >> potential to make things such as recursive next-hop calculation much > > yup, certainly... you probably have to think long and hard about any > standard advice from vendors/third-parties about configuration > optimizations ;( > >> harder since using link-local-scope addresses that aren't on-link for >> you would seem problematic. maybe using the loopback is an acceptable >> alterntive for some applications but you loose access to some tools this way > > see my comment about complexity... oh, included below, how nice of you :) > >>> complexity is going to cause you pain, it is going to cause you >>> problems and it is going to lengthen outages :( avoid complexity. I think this is the critical element, many things here will cause pain. Using link-local for these purposes has so many unintended side-effects it's not something anyone rational will deploy. Just because I can stick a pencil in my eye doesn't mean it's a good idea nor something I should write a how-to document for. I have a redundant eye, but at some point the consequences are likely to be quite catastrophic. - Jared _______________________________________________ OPSEC mailing list [email protected] https://www.ietf.org/mailman/listinfo/opsec
