I have a customer hammering my links with Microsoft backups/syncs/etc and I wanted to bias them out a mostly unused 500M connection we have. I built this to bump the local preference from 500 which three of my peers share. My target AS# is 8075 and one of the prefixes is 220.127.116.11. When I do sho ip bgp 18.104.22.168 my local pref is still 500 (pasted last). What did I mess? Thanks for your time.
router bgp MY_AS#
neighbor 8.28.xxx.xxx remote-as 16657
neighbor 8.28.xxx.xxx description Lightower
neighbor 8.28.xxx.xxx version 4
neighbor 8.28.xxx.xxx allowas-in
neighbor 8.28.xxx.xxx soft-reconfiguration inbound
neighbor 8.28.xxx.xxx route-map Lightower-Ext in
neighbor 8.28.xxx.xxx route-map localonly out
route-map Lightower-Ext permit 10
match as-path 50
set local-preference 500
route-map Lightower-Ext permit 12
match as-path 55
set local-preference 600
ip as-path access-list 50 deny _8075$
ip as-path access-list 50 permit .*
ip as-path access-list 55 permit _8075$
sho ip bgp 22.214.171.124
MY_AS# 3356 8075
8.28.xxx.xxx from 8.28.xxx.xxx (my.ex.ternal.ip)
Origin IGP, localpref 500, valid, external, best
Solved! Go to Solution.
I too thought the config was correct. AS8075 and 126.96.36.199 both belong to Microsoft and are out of my control. To keep the embarrassment to a minimum I've hidden our AS as MY_AS# in the config.
Still, when I do show ip bgp from any of my bgp routers the route 8075 does not default to my Lightower peering at 16657 and the path through 16657 shows a local preference of 500. Even on the router with this config! :(
I'm at a loss what to do next.
Jon Marshall has the correct answer. (I'll mark it so when the page does not fail)
Thank You. I was waiting for overnight to run clear ip bgp but hadn't thought about running it conditionally. d'oh.
I was just about to say that you could use an access list matching the prefix as the match criteria instead of the as path...
Either way, the reset did it. Glad you got it resolved.
On a side note, check if your BGP neighbors support route refreshing:
sh ip bgp neighbors x.x.x.x | s Neighbor capabilities
Route refresh: advertised and received(old & new)
If they do, get rid of the soft configuration inbound, since it doubles the size the BGP routing table. With route refresh, to request an update, you only need to use:
clear ip bgp x.x.x.x in