Hey Hannes,

> if *just* loose source routing is required, then the same can get achieved
> using nested IP tunnels.

I would agree with above.

In fact one could state a question which encap would be more efficient
two IPv6 extension headers or 6 IPv4 headers :)

Also data plane processing wise one could perhaps observe that
flexible handling of multi level IPv4 headers is already supported
today in a lot of deployed hardware ....

> point is that you can't get away with loose source routing semantics for
> supporting all the SR use-cases.

Care to list those ? Perhaps for sake of documentation it may turn
helpful. Maybe for simplicity just limit those use cases to IP
forwarding paradigm only :)

Thx,
R.

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

Reply via email to