> There are decades of mailing lists archives at nanog and others that have > the same thing -- 1) stressed out ops guy 2) buggy code (tac says need to > load latest code as first step) 3) L2 mess -- most of those examples of > epic failure are ipv4 related, but many are just ethernet fails. > > If your point is that IPv6 cannot be deployed at scale, i have a list of > meaningful counter examples where in fact it does work. > > And as already mention, the mailing list archive at nanog and others is > full of folks with poor design or gear.
Did you actually read the whole story? Did you read Jeff Wheeler's presentation, referenced in the comment? http://inconcepts.biz/~jsw/ipv6_nd_problems_with_l2_mcast.pdf Steinar Haug, AS 2116