cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
418
Views
2
Helpful
2
Replies

Segment Routing RFC8402 Prefix-SID Scope (BGP&IGP), IGP Adj-SID Scope

Hussain Sameer
Level 1
Level 1

Hi All,

I found in Segment Routing Architecture RFC 8402

 

"A BGP-Prefix segment is global (unless explicitly advertised otherwise) within the SR domain."

"An IGP-Prefix segment is global (unless explicitly advertised otherwise) within the SR domain."

"An IGP-Adjacency segment is local (unless explicitly advertised otherwise) to the node that advertises it."

 

But I can not understand how an IGP or BGP Prefix-SID, can not be in the global SR domain scope, or an IGP Adj-SID can be in the global domain scope, can anyone explain? how can this be implemented and what is the benefit of that?

 

Best Regards,

Hussain Sameer

 

2 Replies 2

Harold Ritter
Level 12
Level 12

Hi @Hussain Sameer 

According to RFC8402:

Local Segment: In SR-MPLS, this is a local label outside the SRGB.
   It may be part of the explicitly advertised SRLB.  In SRv6, this can
   be any IPv6 address, i.e., the address may be part of the SRGB, but
   used such that it has local significance.  The instruction associated
   with the segment is defined at the node level.

A perfect example of that would be an adjacency SID, which is used to steer the traffic to a specific adjacency on a specific node.

Regards, 

Harold Ritter
Sr Technical Leader
CCIE 4168 (R&S, SP)
harold@cisco.com
México móvil: +52 1 55 8312 4915
Cisco México
Paseo de la Reforma 222
Piso 19
Cuauhtémoc, Juárez
Ciudad de México, 06600
México

Hi Harold,

Thanks for your replay,

I see your idea, you are talking about the local segment which is Adj-SID, that can be part of SRGB in SRv6, but the case is whether the SID is globally significant or locally significant, in my question I'm wondering how the Prefix-SID should be in the SRGB but also can be locally significant.

Please see below my research about this case:

This paragraph from the same RFC 8402 says:

   o  While SR allows a local segment to be attached to an IGP prefix,
      where the terminology "IGP-Prefix segment" or "Prefix-SID" is
      used, the segment is assumed to be global (i.e., the SID is
      defined from the advertised SRGB).  This is consistent with all
      the described use cases that require global segments attached to
      IGP prefixes.

Also, I found in RFC 8665 this paragraph:

            L-Flag:  Local/Global Flag.  If set, then the value/index
               carried by the Prefix-SID has local significance.  If not
               set, then the value/index carried by this sub-TLV has
               global significance.

 and in this RFC 8667:

               L-Flag:  Local Flag.  If set, then the value/index
                        carried by the Prefix-SID has local
                        significance.  By default, the flag is UNSET.

 Upon the above, I see that this can be implemented, but yet cisco does not implement it, I can confirm that based on this PDF https://www.segment-routing.net/tutorials/2016-09-27-segment-routing-igp-control-plane/

pages 34 and 36 mentioned that this flag bit is always unset.

Also, weirdly, adj-sid can be global as RFC 8402 explicitly says that as below:

   IGP-Adjacency Segment: an IGP-Adjacency segment is an IGP segment
   attached to a unidirectional adjacency or a set of unidirectional
   adjacencies.  By default, an IGP-Adjacency segment is local (unless
   explicitly advertised otherwise) to the node that advertises it.
   Also referred to as "Adj-SID".

and in the same pdf above cisco does not implement it, as this flag is always set, please see pages 52 and 53

However, I can not find any info about the BGP prefix SID, and how it can be locally significant, I searched in this RFC 8669 and found that the flags are undefined as below:

HussainSameer_0-1710793103088.png

As I know now how IGP Prefix-SID or Adj-SID can be locally significant by using L bit, appreciate your idea about how BGP Prefix-SID can be determined as locally significant also.

Yet I still have not found any reason for implementing such a feature !!

Best Regards,

Hussain Sameer