Quote:
Originally Posted by Ignitionnet
I'm not sure how customer routers are going to fall offline as a result of VM enabling IPv6.
If the customer-owned routers don't support it they just won't send the DHCPv6 request?
The issue with the Superhubs is more that VM will want them migrated to IPv6 on their HFC-side interface, the major point of IPv6 for cable companies is to allow for more addresses than RFC 1918 allows for HFC-side addresses, so they need them to support it.
Zero point in VM deploying it if their owned CPE don't support it, hence the dependency.
A full migration to IPv6 would allow for clean, centralised management of all Liberty Global cable territories without workarounds to compensate as there would no longer be RFC 1918 IPv4 address overlap between them forcing it.
Sorry if I'm missing something, which I might be!
|
Obviously can't comment on the HFC side of things, but on the consumer side there have been any number of issues with IPv6. It seems that although quite a lot of routers
support IPv6, their implementation is broken in some capacity - they don't route properly, they drop packets, or they just have no proper security in mind.
Imagine this: Pure IPv4 network is running just fine. Then a router suddenly gets an IPv6 lease and starts broadcasting that to the other devices on the network. Those devices get an IPv6 address, so naturally next time they lookup google.com, they see the AAAA record, try to connect via IPv6 only to have the router drop the packet. Suddenly, Google no longer works, all from just
enabling IPv6.
I hope the situation is better these days, but certainly a couple of years ago there was a mess. I believe (can't cite a source though) that Vista's initial IPv6 support was also in some way broken, though I hope it has been fixed since then - still, customers aren't always known for keeping up to date. All they will see is one day it works fine, the next it doesn't.
Also, it needs to be done.