Situation: Got a BI4k with two gig-e uplinks to provider X, and a gig-e
uplink to provider Y.
Assume a route, 10.0.0.0/8, is fed by both providers (there are three BGP
peers, two for X and one for Y, and BGP multipath is enabled). Provider
Y's route is localpref'ed, and is best:
Network Next Hop Metric LocPrf Weight Path
*>i 207.x.0.0/16 207.x.x.34 0 180 0 yyyy 6079 i
* 207.x.0.0/16 209.x.x.157 0 100 0 xxxx 6079 i
* 207.x.0.0/16 209.x.x.213 0 100 0 xxxx 6079 i
When tracerouted to the above destination FROM THE FOUNDRY CLI, it works
fine (goes via yyyy). However, whence tracing from a host *behind* the
foundry (that then goes THROUGH the foundry), it takes one of the xxxx
routes.
Any clues as to why the FIB would ignore the RIB?
-- Alex Rubenstein, AR97, K2AHR, alex@nac.net, latency, Al Reuben --
-- Net Access Corporation, 800-NET-ME-36, http://www.nac.net --
This archive was generated by hypermail 2b29 : Mon Aug 04 2003 - 04:10:05 EDT