Re: Friday's Random Comment - About: Arista and FIB/RIB's

On 29 April 2016 at 13:25, Nick Hilliard wrote: >> The more paths you receive from different sources, the more likely it >> is that this list of 120k "superfluous" prefixes will converge >> towards zero. > > Agreed that small numbers of paths are most unlikely to create the > conditions for this p

Standards for last mile performance

The CRTC hearing went well (thanks for all your help). One of the unanswered questions was how to set performance standards for the last mile to ensure people get advertised speeds (within reason). I had asked the question about contention ratio and it appears there is no proper way to set such

Re: Standards for last mile performance

For us (FTTH) we had/have enough aggressive foresight to do smaller splits.. 1x16. Some are doing 1x2's or 1x4's at the corner somewhere into 1x16's or 1x8's, so at the point where you start to hit decent saturation you can just shrink the upstream split and fuse onto a new upstream strand / optic.

Superfluous advertisement (was: Friday's Random Comment)

A use case for a longer prefix with the same nexthop: F / \ D E | | B C \ / A Suppose A is a customer of B and C. B has a large address space: 10.1.0.0/16. B allocates a subset to A: 10.1.1.0/24. B advertises the longer prefix to its backup provider C. C propagates it to E and

RE: Superfluous advertisement (was: Friday's Random Comment)

Simpler, with B and C peered: F / \ B---C \ / A If B does not send the /24 to F, then F will send all the traffic to C, even if A wanted a load balance. Maybe I could ask the community: Why do you advertise longer prefixes with the same nexthop as the shoter prefix? Is it this use cas

Re: Superfluous advertisement (was: Friday's Random Comment)

> If B does not send the /24 to F, > then F will send all the traffic to C, > even if A wanted a load balance. > > Maybe I could ask the community: > Why do you advertise longer prefixes with the > same nexthop as the shoter prefix? > Is it this use case, or something else? it is a common TE use

Reporter's questions

Folks I've long lurked here and thought to ask about the story below and related. The news here is that Hurricane Electric is opening Africa pops and likely bringing down prices significantly. I made some (semi-informed) guesses about the relation of backhaul/transit to ISP costs. Anyone with dat

Re: BGP FlowSpec

I was looking into using this mechanism for blocking DDoS on Juniper devices, but at the time, they only supported 8k flowspec entries/routes and this was not sufficient to deal with the problem. My fallback was to poison the routing table with null routes, but the problem with this was that it did

RE: Superfluous advertisement (was: Friday's Random Comment)

> A use case for a longer prefix with the same nexthop: > >F > / \ > D E > | | > B C > \ / >A > > Suppose A is a customer of B and C. This is possible, but only remotely probable. In the real world, D and E are likely peered, as are B and C. Further, it's quite possible for

Re: Superfluous advertisement (was: Friday's Random Comment)

>>F >> / \ >> D E >> | | >> B C >> \ / >>A >> >> Suppose A is a customer of B and C. > > This is possible, but only remotely probable. In the real world, D and > E are likely peered, as are B and C. "likely?" with what probability? any measurement cite please. nothing exa

Re: BGP FlowSpec

On 30 Apr 2016, at 19:56, Pierre Lamy wrote: > to null out the destination rather than the source. --- Roland Dobbins