Symptom
- When MED updates for ECMP BGP routes received from 2 eBGP peers and both are updated simultaneously, it will only update the metric in RIB for one of the paths. This leads to loss of ECMP in routing table.
DCSW# sh ip bgp 8.8.8.8
BGP routing table information for VRF default, address family IPv4 Unicast
BGP routing table entry for 8.8.8.8/32, version 47
Paths: (2 available, best #2)
Flags: (0x08001a) (high32 00000000) on xmit-list, is in urib, is best urib route, is in HW
Multipath: eBGP
Path type: external, path is valid, not best reason: newer EBGP path, multipath, no labeled nexthop, in rib
AS-Path: 65200 , path sourced external to AS
10.1.2.2 (metric 0) from 10.1.2.2 (10.1.2.2)
Origin incomplete, MED 105, localpref 100, weight 0 <<<<
Advertised path-id 1
Path type: external, path is valid, is best path, no labeled nexthop, in rib
AS-Path: 65100 , path sourced external to AS
10.1.1.2 (metric 0) from 10.1.1.2 (10.1.1.2)
Origin incomplete, MED 105, localpref 100, weight 0
Path-id 1 not advertised to any peer
DCSW# sh ip route 8.8.8.8
IP Route Table for VRF "default"
'*' denotes best ucast next-hop
'**' denotes best mcast next-hop
'[x/y]' denotes [preference/metric]
'%' in via output denotes VRF
8.8.8.8/32, ubest/mbest: 1/0
*via 10.1.2.2, [20/6], 00:07:59, bgp-65000, external, tag 65200 <<<<<
via 10.1.1.2, [20/105], 00:07:59, bgp-65000, external, tag 65100
Conditions
- Receiving equal cost paths from 2 eBGP peers.
- Multipath configured.
Workaround
- Clear the route "clear ip route "
- Configure a route-map to set all routes received from eBGP neighbors to same MED so as to not impact ECMP.