I think also including END.T which will select specific FIB,  and  the 
4291-style IPv6 address may be in last SID entry in Last SRH of Packet, other 
SID may not be rfc429-style IPv6@;

--------------------------------------
Cheers !


WANG Weibin

From: spring <[email protected]> On Behalf Of Ron Bonica
Sent: 2019年10月18日 5:33
To: Pablo Camarillo (pcamaril) <[email protected]>; SPRING WG List 
<[email protected]>
Subject: Re: [spring] draft-ietf-spring-srv6-network-programming :Section 4.16.2

Pablo,

I see your point. An RFC 4291-style IPv6 address can replace END with USD, but 
it can’t replaced END.X with USD.

                                                            Ron




Juniper Business Use Only
From: Pablo Camarillo (pcamaril) <[email protected]<mailto:[email protected]>>
Sent: Thursday, October 17, 2019 5:45 AM
To: Ron Bonica <[email protected]<mailto:[email protected]>>; SPRING WG 
List <[email protected]<mailto:[email protected]>>
Subject: Re: [spring] draft-ietf-spring-srv6-network-programming :Section 4.16.2

Hi Ron,

Are you sure you get the same behavior with an interface address?

Can you please explain me how do you decapsulate and forward the inner 
-encapsulated- IPv6 packet on a link with a high IGP metric when the packet 
arrives with a DA=IPv6 interface address?

Thanks,
Pablo.

From: Ron Bonica <[email protected]<mailto:[email protected]>>
Date: Wednesday, 16 October 2019 at 01:59
To: "Pablo Camarillo (pcamaril)" 
<[email protected]<mailto:[email protected]>>, SPRING WG List 
<[email protected]<mailto:[email protected]>>
Subject: RE: [spring] draft-ietf-spring-srv6-network-programming :Section 4.16.2

Pablo,

You haven’t answered my question.

Why specify a USD flavor of END, END.X and END.T when you can get the exact 
same behavior by populating SID[0] with an IPv6 address that identifies a real 
interface?

USD seems redundant.

                                        Ron




Juniper Business Use Only
From: Pablo Camarillo (pcamaril) <[email protected]<mailto:[email protected]>>
Sent: Tuesday, October 15, 2019 12:45 PM
To: Ron Bonica <[email protected]<mailto:[email protected]>>; SPRING WG 
List <[email protected]<mailto:[email protected]>>
Subject: Re: [spring] draft-ietf-spring-srv6-network-programming :Section 4.16.2

Ron,

The USD flavour is used in any T.Encaps policy applied within the SR domain.

Cheers,
Pablo.

From: spring <[email protected]<mailto:[email protected]>> on 
behalf of Ron Bonica 
<[email protected]<mailto:[email protected]>>
Date: Tuesday, 15 October 2019 at 01:03
To: SPRING WG List <[email protected]<mailto:[email protected]>>
Subject: Re: [spring] draft-ietf-spring-srv6-network-programming :Section 4.16.2

Authors,

Sorry, I meant to say USD, not USP.

                         Ron




Juniper Business Use Only
From: Ron Bonica
Sent: Monday, October 14, 2019 2:36 PM
To: SPRING WG List <[email protected]<mailto:[email protected]>>
Subject: draft-ietf-spring-srv6-network-programming :Section 4.16.2

Authors,

The USP flavor of the END, END.X and END.T functions isn’t needed.

USP is the default IPv6 behavior. So, if the source node specifies SID[0] as an 
RFC 4291 address on the SR egress node, you get the USP behavior for free.

                                                           Ron



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

Reply via email to