Re: [Bier] WG adoption call for draft-chen-bier-frr-02

[[e-mail]]

Hi WG,






I think it's an implementation issue, so I oppose the adoption.


Obviously, the implementation scheme, i.e, FRR-BIFT per neighbor, described in this document has serious scalability problems. That means that too many tables need to be updated when topology events occur. There may be a simpler implementation that doesn't have to be done in this way.






For example, we can let BIFT entry contain both primary NBR and backup path (note that the backup path may be direct NBR, or remote NBR, or segment-list, according to IGP TI-LFA result).



There are primary FBM and backup FBM. The primary FBM contains the Bit-Positions of those BFERs that has the same primary NBR, the backup FBM contains the Bit-Positions of those BFERs that has the same backup path. In this implementation, when a BFR received a BIER packet, and if the primary NBR fails, a copy will be sent to backup path, and the bitstring contained in the copy is the result of "original bitstring of the received packet" & "primary FBM" & "backup FBM".






Another implementation may let primary FBM and backup FBM of BIFT entry to be same, to contain the Bit-Positions of those BFERs when their primary NBR are same and their backup path are also same. However, this implementation couples the setting of primary FBM and backup FBM. It's not recommended.






Regards,


PSF













原始邮件



发件人：张征00007940
收件人：[e-mail];
抄送人：[e-mail];
日 期 ：2021年03月16日 14:41
主 题 ：[Bier] WG adoption call for draft-chen-bier-frr-02


_______________________________________________
BIER mailing list
[e-mail]
https://www.ietf.org/mailman/listinfo/bier


A 2-week WG adoption call begins for the following draft:

https://datatracker.ietf.org/doc/draft-chen-bier-frr/


Please indicate your support or objection by March 30th, 2021.


Authors, please respond to the list indicating whether you are aware of any IPR that applies to this draft.


Thanks,


Sandy (As WG secretary, on behalf of Greg/Tony)