Hi Lucy.

Lucy yong <[email protected]> writes:

> IMO: this is the reason why we need to define a new service type for
> this.

I've tried following this thread and have also read the draft. I admit
to having difficulty understanding what the actual issue is or what
text is being asked for in the framework document.

It sounds to me like the crux of the issue is that for the case of
providing L3 (IP) service to Tenant Systems (TSs), there are two
sub-cases:

 - all the TSs are on the same subnet, so they all can reach each
   other without going through a router

 - TSs are on different IP subnets, with the collection of those
   subnets comprising a single NVO3 Virtual Network. 

Are you saying we need a different "service type" to distinguish those
two cases? If so, what do you mean by "service type"?

I think of NVO3 providing either an L2 or L3 service to TSes. I.e., 2
different types of service. Are you proposing that the L3 case be
split into two subcases?

If so, I agree that the two subcases are different, but I've always
thought of them as being the same service. I.e., that L3 service would
handle both cases. Why would we want to split them into different
"services"? And in any case, one could look at the first case (just
one subnet) as just being a special case of the second case, so why
handle it specially?

Do I understand correctly?

Thomas

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

Reply via email to