On 05/19/2016 06:42 AM, Koenraad Lelong wrote:
Op 10-05-16 om 13:29 schreef Carlos E. R.:
Please allow my ignorance, I'll place a question for my learning :-)
I thought the general purpose is to get a large range (/64) IPv6 addresses from the ISP. A single one is assigned to the external interface of the entry router, another single one to the internal interface, and then the router hands over (2^64 -2) addresses to the millions of possible internal machines, each one with an IPv6 real world address, so that any internal machine is reachable (if the router firewall permits it) from the world.
A little late, but I can't resist.
AFAIK, *all* addresses of the given prefix are intended for the *internal* network. Just like in ipv4 the external address and internal address or a router should be in a different subnet (=prefix in ipv6 language). Otherwise, how is the router going to decide which way a packet has to go ? Internal : 192.168.0.1/24, external 192.168.0.2/24. Where does it send a packet for 192.168.0.3/24 ? Likewise for ipv6. It is possible to devide the prefix further in smaller prefixes, but, AFAIK, that's not preferred.
Koenraad.
As with IPv4, the LAN side of the router must be on the same subnet as the local network, so you still use 1 address there. With IPv4, the other side of the router has an address on a different subnet or may use just the interface ID for point to point links. With IPv6, the link local address is typically used, instead of a separate subnet to reach the next router. Also, IPv4 loses 1 address to the broadcast address. As there is no such thing as a broadcast address in IPv6, this no longer applies.
P.S. I'm no guru, I'm trying to understand ipv6. So please correct me if I'm wrong. Also links to study-material are welcome !
This is an excellent book: http://shop.oreilly.com/product/0636920023432.do -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org