Hi, Iljitsch, > There is currently no writeup of how to use the flow label for ECMP. > And as far as I can tell there are no implementations of this either. > Which is a real shame.
There are a few proposals on how to set it (e.g., Blake's and mine). -- This could be a good way to start. > - we shouldn't lock down the flow label such that only one flow label > per flow is allowed because this would impede future innovation The only problem with this is that if you allow a given flow to use multiple flows, the flow-label reuse frequency might increase, and collisions might occur. -- then you violate the requirement on uniqueness fo each pair (src addr, dst addr). > - zero flow labels are still created by many systems, but these would > hamper a flow label based ECMP. Rewriting zero flow labels into a > real flow label somewhere in the network would therefore be a useful > function Some of these boxes set the flow label inconsistently. See the paper published by Malone et al (referenced in Blakes flow-label I-D, and probably in mine, too). If you want consistency, you might have to rewrite both, unfortunately. Thanks! -- Fernando Gont e-mail: ferna...@gont.com.ar || fg...@acm.org PGP Fingerprint: 7809 84F5 322E 45C7 F1C9 3945 96EE A9EF D076 FFF1 -------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------