Why the BGP next-hop is inaccessible?

[iceking2nd]

There are some configuration of my Cisco 6506-E, the IOS version is (s72033_rp-ADVENTERPRISEK9_WAN-M), Version 12.2(18)SXF17b

C6506E#show run | b r b

router bgp 64517

template peer-policy Route-Opt

route-map Route-Opt in

route-map BROADCAST_LOCAL out

exit-peer-policy

!

bgp router-id 124.202.129.196

no bgp default ipv4-unicast

bgp log-neighbor-changes

neighbor 124.202.11.21 remote-as 17964

neighbor 124.202.11.25 remote-as 17964

neighbor 172.18.0.13 remote-as 64513

neighbor 172.18.0.17 remote-as 64513

maximum-paths 4

!

address-family ipv4

redistribute static

neighbor 124.202.11.21 activate

neighbor 124.202.11.21 inherit peer-policy Route-Opt

neighbor 124.202.11.25 activate

neighbor 124.202.11.25 inherit peer-policy Route-Opt

neighbor 172.18.0.13 activate

neighbor 172.18.0.13 route-map BROADCAST_LOCAL out

neighbor 172.18.0.17 activate

neighbor 172.18.0.17 route-map BROADCAST_LOCAL out

maximum-paths 4

no auto-summary

no synchronization

network 118.145.16.0 mask 255.255.254.0

network 124.202.128.56 mask 255.255.255.252

network 124.202.129.196 mask 255.255.255.255

exit-address-family

C6506E#show route-map Route-Opt

route-map Route-Opt, permit, sequence 10

Match clauses:

as-path (as-path filter): 101

Set clauses:

ip next-hop 202.99.1.30

Policy routing matches: 0 packets, 0 bytes

route-map Route-Opt, permit, sequence 65535

Match clauses:

Set clauses:

local-preference 200

Policy routing matches: 0 packets, 0 bytes

C6506E#show ip route 202.99.1.30

Routing entry for 202.99.1.30/32

Known via "static", distance 1, metric 0

Redistributing via bgp 64517

Advertised by bgp 64517

Routing Descriptor Blocks:

* 172.18.0.13

Route metric is 0, traffic share count is 1

172.18.0.17

Route metric is 0, traffic share count is 1

C6506E#ping 202.99.1.30

Type escape sequence to abort.

Sending 5, 100-byte ICMP Echos to 202.99.1.30, timeout is 2 seconds:

!!!!!

Success rate is 100 percent (5/5), round-trip min/avg/max = 1/1/1 ms

C6506E#show ip bgp regexp _4808$

BGP table version is 14031, local router ID is 124.202.129.196

Status codes: s suppressed, d damped, h history, * valid, > best, i - internal,

S Stale

Origin codes: i - IGP, e - EGP, ? - incomplete

Network          Next Hop            Metric LocPrf Weight Path

*  27.131.220.0/22  202.99.1.30                            0 17964 65008 17816 4837 4808 i

BJ-CY-JXQ-C-M35.1-SC-C6506E#show ip bgp 27.131.220.0

BGP routing table entry for 27.131.220.0/22, version 0

Paths: (2 available, no best path)

Multipath: eBGP

Not advertised to any peer

17964 65008 17816 4837 4808

202.99.1.30 (inaccessible) from 124.202.11.21 (202.99.1.27)

Origin IGP, localpref 100, valid, external

Why the next-hop 202.99.1.30 is inaccessible?

[Edison Ortiz · Accepted solution]

BGP checks the interface the next-hop was learned from. If 202.99.1.30 was learned via a different interface than the one use to peer with 124.202.11.21, then the route will be marked as inaccessible.

[iceking2nd]

How can I fix it?

Take a host route from 172.18.0.13 in BGP?

Or disable any feature in BGP?

[iceking2nd]

Establish BGP Peer with Loopbak interface,and set the ebgp-multihop.

Thanks.

[Edison Ortiz]

Using the loopback for BGP peering was a very sound approach.

By using the loopback, you need to advertise it via an IGP and the routers peering with BGP will learn the loopback subnet via the same interface the next-hop was learned from.

Thanks for the rating and welcome to the forums!