Re: [Teas] Moving forward with draft-ietf-teas-ietf-network-slices

[Adrian Farrel <[e-mail]>]

Hi Shaofu,

> Thanks for your hard work for the slice architecture.
>
> I have a little comments,  as we know, there are slice scenarios that have
> nothing to do with VPN, for example, multiple slices in backbone can be
> used for connections between metros. So it is strange to put a usecase
> (VPN+) of slice be as a scheme for slice architecture, as section 6.3 described.

I think there are several things to say on this subject.

The first is that you're right that the use cases for network slicing go far beyond providing VPNs. But I would note that "connectivity services" are very similar to VPNs: they are expressed in terms of a connectivity matrix between a number of endpoints. Actually, you'll notice that this is comparable to a network slice.

I think the next thing to notice is that section 6.3 is not a use case. The whole of section 6 is about ways to realise (i.e., build) network slices, not a discussion of how to use slices.

In fact, ACTN and VPN+ are both TEAS working group architectural solutions (although VPN+ isn't an RFC as yet). That's why they're included in this section.

Best,
Adrian