The examples in my previous note sort of provided backing for my view
that the MAP endpoint IPv6 prefix can be limited to a maximum of a /64,
thus making the IID fully conformant both to RFC 4291 and to RFC 6052.
Obviously, the more IPv4 address bits you can stick into the EA bits,
the fewer the number of MAP rules you have to provision. The picture of
a MAP endpoint IPv6 prefix can be drawn like this:
+-----------------------------+-------------------+-----------+
| Rule IPv6 prefix | IPv4 suffix | PSID |
+-----------------------------+-------------------+-----------+
| | | |
| Common to all CEs sharing | Common to all CEs | Unique to |
| the same BMR | sharing the same | this CE |
| | IPv4 address | |
If we expect a Rule IPv6 prefix to be a /48 and a PSID to be no more
than 8 bits in length (i.e. CEs get 512 ports each, sharing ratio is
128), limiting the endpoint IPv6 prefix to 64 bits leaves room for eight
bits of suffix. That in turn means each mapping rule serves 16384 CEs.
If the Rule IPv6 prefix backs off to a /40, we can get two million CEs
per mapping rule.
I think limiting the prefix length to 64 bits is reasonable. Comments?
Tom Taylor
_______________________________________________
Softwires mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/softwires
