Re: [Idr] Does SRGB TLV in draft-ietf-idr-bgp-prefix-sid qualify for BGP Attibute discard?

["Susan Hares" <[e-mail]>]

Eric: 

 

I think we are approaching a convergence on the need for a clear problem description and that a UPDATE that is not MALFORMED will cause problems. The restriction for attribute discard restriction on routing considered that issue. 

 

I will write-up a problem description for the SID and SRGBs and send it to the list.  This write-up will probably come on Tuesday (after the IETF deadline) due to my “day job” work.  

 

Thank you for your questions.

 

Susan Hares 

 

From: Idr [mailto:[e-mail]] On Behalf Of Eric C Rosen
Sent: Tuesday, June 26, 2018 3:39 PM
To: Susan Hares; 'Ketan Talaulikar (ketant)'; 'Robert Raszuk'
Cc: 'idr@ietf. org'
Subject: Re: [Idr] Does SRGB TLV in draft-ietf-idr-bgp-prefix-sid qualify for BGP Attibute discard?

 

On 6/26/2018 3:17 PM, Susan Hares wrote:



AFAIK, the normative references do not provide adequate information on the manageability and expectations on the timing for the BGP Peer segments.


I think the normative references  make it clear that SIDs and SRGBs are not dynamically changing values.




Most BGP attributes have some effect on routing, the question is whether discarding a malformed attribute will result in inconsistent routing (loops).





[Sue] Yes – this is the question I am trying to discuss.  I’ve provided scenarios where it might fail.


If you can provide a clear description of a scenario that results in inconsistent routing, that would be interesting.  However, I don't think you've described such a scenario yet (or I missed it somehow).  

The real danger with all this traffic engineering stuff is not that an UPDATE might be malformed, but that it might be properly formed but have incorrect values.  But that's the nature of traffic engineering.