From: D.H. Williams (draythw@gmail.com)
Date: Mon Jul 17 2006 - 14:21:43 ART
that link on the Cost community is interesting. Thanks all for the replies.
On 7/16/06, Scott Morris <swm@emanon.com> wrote:
>
> Assuming that you aren't going to be able to reprogram bgp in the lab, a
> couple of things pop into mind:
>
> 1. Make everything else before AS-Path selection equal. Therefore the
> first difference would be the AS Path
> 2. Use something called the BGP Cost Community
> See: http://www.cisco.com/warp/public/459/25.shtml
> or
>
> http://www.cisco.com/en/US/products/sw/iosswrel/ps1829/products_feature_guid
> e09186a008053b5b6.html
> That one seems to be supported in a variety of IOS releases, but is more
> tied to MPLS stuff. So I would doubt that.
>
> We can take AS path AWAY from the selection proces by using "bgp bestpath
> as-path ignore", but that is a hidden command (although well documented!)
> and likely won't be part of your test.
>
> I would view it more as "avoiding" the things higher in the preference
> list
> than "overriding" them.
>
> -----Original Message-----
> From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
> D.H.
> Williams
> Sent: Sunday, July 16, 2006 8:31 PM
> To: ccielab@groupstudy.com
> Subject: BGP selection process
>
> Okay, I'm curious if anybody knows (or can provide a link) that explains
> the
> different ways to override those BGP methods that are higher in the
> selection process.
>
> To leave the AS, Weight and LP.
>
> For traffic coming in, AS-Path, and MED.
>
> So, is there a way to only take into account AS-Path, for example (even
> though it is further down the list)? Effectively ignoring the other
> methods
> (weight, lp, which are taken into account first). Obviously I know of the
> always compare med, but I was curious about possibly any other ways to
> override things that are higher up in the selection process.
>
> Thanks!
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
This archive was generated by hypermail 2.1.4 : Tue Aug 01 2006 - 07:13:47 ART