Just a nit, but looking at the -03, version, I can see how the proposed
Geneve option is parsed, but what is the plan for the future?
Currently, the first 4 octets of the option data have a fixed format.
If the H flag-bit is present (in the 5th bit of the option data), then
the last 38 octets are the HMAC information (which seems to have
redundant type and length fields).  The remaining octets are divided
into a list of IPv4 or IPv6 addresses.  (Presumably the type is
determined by context.)

But it's not clear to me how this generalizes.  It seems like the tail
of the option data is intended to be a series of TLV items, but how does
one find the dividing point between the SF address list and the TLVs?

Dale

_______________________________________________
nvo3 mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/nvo3

Reply via email to