%BGP-3-NOTIFICATION: sent/received from neighbor 10.1.1.2 2/3 (BGP identifier wrong)4 byte

[TCC_2]

Core Issue:

This situation is possible if the IP addresses are re-used on the neighboring routers resulting in duplicate BGP router IDs. 

R2#
*Mar  1 00:16:02.111: %BGP-3-NOTIFICATION: received from neighbor 10.1.1.2 2/3 (BGP identifier wrong) 4 bytes 0A010101

R1#
*Mar  1 00:20:06.235: %BGP-3-NOTIFICATION: sent to neighbor 10.1.1.2 2/3 (BGP identifier wrong) 4 bytes 0A010101
R1# FFFF FFFF FFFF FFFF FFFF FFFF FFFF FFFF 002D 0104 000A 00B4 0A01 0101 1002 0601 0400 0100 0102 0280 0002 0202 00

To resolve the issue, review the BGP router IDs on the peering routers by issuing the  show ip bgp summary command. If they are same, change the BGP router ID by configuring unique IP addresses on the interfaces being used for the router ID, or by issuing the bgp router-id command under the BGP router configuration mode. Clear the BGP process with the clear ip bgp {* |  neighbor-address |  peer-group-name} [soft [in |  out]] command. If this does not fix the problem, remove the BGP configuration from the routers and reconfigure them.

Resolution:

When Border Gateway Protocol (BGP) is configured on a router, an IP address that uniquely identifies the router is used as BGP router identifier (ID). Cisco IOS  determines this identifier the same way it selects the Open Shortest Path First (OSPF) router ID, by choosing the highest IP address on a loopback interface, or a physical interface if there is no loopback interface configured.  The BGP router ID information is exchanged between the peers in the Open message when they attempt to form neighbor relationship.

If the routers attempting to form neighbor relationship have the same BGP router ID, they send a notification message to the other peer and close the session. The message BGP-3-NOTIFICATION: received from neighbor x.x.x.x 2/3 (BGP identifier wrong) 4 bytes or BGP-3-NOTIFICATION: sent to neighbor x.x.x.x 2/3 (BGP identifier wrong) 4 bytes is displayed on the router console if the bgp log-neighbor-changes command is enabled on the router.

BGP Peering

iBGP

eBGP (includes eBGP using loopback addresses)

iBGP using route reflectors

Router#sh ip bgp neighbor <neighbor-address> | in state
  BGP state = Active

[usmansa1]

Today I recieved this problem while configuring my two routers R3 and R4 which are connected to each other through Fast Ethernet. I have configured interface loop 0 on R3 and R4 as 172.16.104.4 and also advertise the same IP address from both routers through BGP. When the BGP selected the router ID and send the it to each other request here is conflict came when they see they have the same router ID. So I suggest every body who recieved this error to please check the router ID on both sides. I am sure the error will be find there....

[Mahadev Patil]

Hi All,

Today i have found the same issue. After that change the router loopback IP address.

Mahadev Patil

[ali_sendme]

In lab test i was getting this too, however running "clear ip bgp *" resolved this issue.

[brett.stone]

I was getting this same message even though I had different BGP identifiers on each router.  In order to figure out my issue I had to do a show ip bgp summary command and from there I was able to see that I had the same GRE tunnel created on each side and the BGP was barfing because of it.  I simply deleted the dupe GRE tunnel and my issue was resolved.

[Ashis Patra]

Thanks a lot.. I do had same problem, and its resolved.

[antonkolev]

I had the same problem today when I set up a lab and accidently place same ip address on two bgp routers - when I change ip address on one of the router - still the same error and output from bgp summary saying the bgp state is idle - 

Fix it with Clear ip bgp * 

[adfsdafs adsfafds]

What adverse effect does duplicated RID on non-peers have? IOS does not complaint when it has 2 iBGP peers with the same RID.

[sac.hin7864556]

I have faced this error too in lab environment and it was occurred when i tried to establish multiple neighborship at one time in my case its both (ebgp and ibgp) so in operations when dealing with multiple neighbors (with single loopback address) then create and establish BGP neighborship one by one with peer routers