[routing-wg] New on RIPE Labs: Power Prefixes Prioritisation for Smarter BGP Reconvergence

Dear Olivier and Job,

Thanks for your feedback, and sorry for delayed answer (I thought I was
already subscribed to the mailing list, but I was not). I was sure the
question about PIC and ADD-PATH was going to appear, and I am going to
address it in the presentation. The answer is that we mainly target
scenarios where you can't use PiC either because the operator is using next
hop self (which we understand is fairly common) or because there is no IGP
to update the pointer.

Going back to Job's questions:

1) Would it perhaps make sense to rank updates by BGP next-hop (using a
metric like traffic volume), rather than destination prefixes  themselves?
Each BGP next-hop in effect represents a group of prefixes.

We are not pushing any ranking algorithm. We used the traffic because it
was the most relevant concern at that moment, but we can also consider
other algorithms. For instance, we received a comment about considering
critical services like the DNSs at the top of the list. For sure I agree we
can consider ranking  traffic by next hop, but as we are mainly targeting
the next hop self situation this can be tricky.

3) Are you looking to publish your work in context of IETF or IRTF?

We are still to decide how and where to present this. IETF is definitely
one option. Any suggestions about how to go about resenting this in the
IETF?
do you suggest to just have an informative presentation in something like
GROW WG or do you think it would make sense to write a draft describing the
mechanism and see if any WG is willing to take this on?


BR,

Juan Brenes
-- 
JUAN BRENES BARANZANO
Universidad Carlos III de Madrid
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ripe.net/ripe/mail/archives/routing-wg/attachments/20171024/3a4f044f/attachment.html>