Assuming you're wanting to use the full T3 all of the time .. unless it
becomes unavailable, a simple configuration would be to only pickup full
routes from the primary T3 connection, then BGP default from the backup
provider. Or you could local_pref "up" destinations via the primary path so
that AS path (etc..) aren't factored. The stuff suggested in rfc1998 would
work nicely for introducing a bit of symmetry for ingress traffic .. assuming
that at least the backup provider supports it.
> We have two T3 connections to two different ISP's using 7507 routers.
> We run BGP between each ISP and IBGP between our two 7507's.
> One T3 is a full T3 flat rate and the other is a falt fee at 15MB
> but burstable to 100MB (supposedly) with added cost over 15MB.
15MB, 100MB... Hrmm.. Perhaps the 100Mb thing comes from their considering
aggregate of both directions (which is still only ~88Mb)..
> Anyway our traffic currently is split about 50/50 load wise between
> those two circuits - at roughly 20MB per circuit outbound.
> We want to influence traffic so it goes over the full T3 to more
> fully utilize that paid for bandwidth.
> Suggestions? I'm unclear how to do it since routes such to
> places like www.cisco.com only show up in our route tables
> on our BBN connection - implying it's preferred or cisco has
> their internet connection off of BBN.
You should still see the BGP path from the "other" provider on the router
where you peer locally, although the path likely won't be selected as best.
If you opt to use it, local_pref will override AS path if you're learning the
destination from both providers.
-danny
This archive was generated by hypermail 2b29 : Sun Aug 04 2002 - 04:12:06 EDT