[mpls] MPLS-RT review of draft-bryant-mpls-sfl-framework-04

["Bocci, Matthew (Nokia - GB)" <[e-mail]>]

Hello authors and WG,

I am an MPLS-RT reviewer assigned to review draft-bryant-mpls-sfl-framework-04.

I have reviewed the draft and found that it is well written and coherent, and is a potentially useful solution.

However, I have a concern with the following text in section 4 describing ECMP and EL/ELI:

“  2.  The operator can elect to use [RFC6790] Entropy Labels which, in
       a network that fully supports this type of ECMP, results in the
       ECMP decision being independent of the value of the other labels
       in the label stack.”

I don’t think you can assume that intermediate LSR implementations only use the EL when spraying packets. Although RFC6790 does mention this as one option, I believe the more common case is to take into account other labels in the stack including the EL. The EL is effectively saying “all other things being equal, here is some additional information to distinguish flows”. It is in addition to, but not an alternative to the rest of the label stack. If the SFL value for packets on a given flow differs, but the EL is the same, then the LSR would still spray the packets because the label stack would be different.

I don’t think that it is reasonable to mandate that LSRs only consider the EL and ignore all other labels in the stack given the impact on the deployed base for just a few end-to-end applications.

I don’t think this draft should be both defining SFLs and trying to solve the general ECMP problem for applications of SFLs. Instead, it should focus on the SFL itself. My suggestion would be to change section 4 to say that ECMP considerations should addressed by an application of SFL if ECMP is a concern for that application.


Best regards

Matthew