BGP Flapping

[Psmurali89]

Hi All

I got BGP configured between 2 DC using loopback interface. It was working fine but recently the BGP is flapping, there were no changes made to this configuration and the physical link itself is not flapping. Am not sure what the issue is, any suggestions is appreciated. 

DC1
====

!
interface Loopback10
description ** DC2 IGP **
ip address 192.168.249.1 255.255.255.255
ip ospf 50 area 1
!
neighbor 192.168.248.1 remote-as 64515
neighbor 192.168.248.1 description ** DC2 **
neighbor 192.168.248.1 password xxxxxxx
neighbor 192.168.248.1 update-source Loopback10
!

DC2
====

!
!
interface Loopback10
description ** DC1 IGP **
ip address 192.168.248.1 255.255.255.255
ip ospf 50 area 1
!
neighbor 192.168.249.2 remote-as 64515
neighbor 192.168.249.2 peer-group GROUP1
neighbor 192.168.249.2 description ** DC1 **
neighbor 192.168.249.2 password xxxxx
neighbor 192.168.249.2 update-source Loopback10

BGP Flapping - log message from DC1
=====================================

%BGP-3-NOTIFICATION: received from neighbor 192.168.248.1 4/0 (hold time expired) 0 bytes
%BGP-5-NBR_RESET: Neighbor 192.168.248.1 reset (BGP Notification received)
%BGP-5-ADJCHANGE: neighbor 192.168.248.1 Down BGP Notification received
%BGP_SESSION-5-ADJCHANGE: neighbor 192.168.248.1 VPNv4 Unicast topology base removed from session BGP Notification received
%BGP_SESSION-5-ADJCHANGE: neighbor 192.168.248.1 IPv4 Unicast topology base removed from session BGP Notification received
%BGP-5-ADJCHANGE: neighbor 192.168.248.1 Up

BGP output
==========

Connections established 1153; dropped 1152
Last reset 00:00:37, due to BGP Notification received of session 1, hold time expired

BGP Flapping - log message from DC2
=====================================

%BGP-3-NOTIFICATION: sent to neighbor 192.168.249.2 4/0 (hold time expired) 0 bytes
%BGP-5-NBR_RESET: Neighbor 192.168.249.2 reset (BGP Notification sent)
%BGP-5-ADJCHANGE: neighbor 192.168.249.2 Down BGP Notification sent
%BGP_SESSION-5-ADJCHANGE: neighbor 192.168.249.2 VPNv4 Unicast topology base removed from session BGP Notification sent
%BGP_SESSION-5-ADJCHANGE: neighbor 192.168.249.2 IPv4 Unicast topology base removed from session BGP Notification sent
%BGP-5-ADJCHANGE: neighbor 192.168.249.2 Up

BGP output
==========

Connections established 1154; dropped 1153
Last reset 00:02:33, due to BGP Notification received, hold time expired

[Flavio Miranda]

Hello

 The cause of the message is simple. Peers are not receiving keepalives. Why it is happening need to be found. 

I suggest this link to guide you. 

https://www.cisco.com/c/en/us/support/docs/ip/border-gateway-protocol-bgp/116377-troubleshoot-bgp-mtu.html 

[MHM Cisco World]

ip address 192.168.249.1 255.255.255.255 <<-

neighbor 192.168.249.2 remote-as 64515 <<- the two IP is different this can make BGP flapping 

[paul driver]

Hello
You say the there is no interface errors but as this is an IBGP peering then you have ospf for the loopback reachability so
If this stable?

Can you post the output of the following please? 

sh int x/x | in MTU|reli|In|in|CRC

Router#ping
Protocol [ip]:
Target IP address: 192.168.249.2
Repeat count [10]:
Datagram size [100]:
Timeout in seconds [2]:
Extended commands [n]: y
Ingress ping [n]:
Source address or interface: 192.168.248.1
DSCP Value [0]:
Type of service [0]:
Set DF bit in IP header? [no]: y
Validate reply data? [no]:
Data pattern [0x0000ABCD]:
Loose, Strict, Record, Timestamp, Verbose[none]:
Sweep range of sizes [n]: y
Sweep min size [36]: 1200
Sweep max size [18024]: 1500
Sweep interval [1]: 50

conf t
no logging console
logging buffered 7
debug bgp ipv4 unicast 192.168.249.2
sh logging 

[Psmurali89]

Hi

I think you are right here with regards to MTU. The OSPF is stable and no interface errors etc. 

So this loopback is used for MPLS LDP and the LDP is also stable. Its just the BGP keeps flapping. 

I just checked this after you pointed the MTU. In DC1 the link terminates in L2 switch where the MTU is 9000, the L3 is in router where the MTU is 1500.

Whereas in DC2 the L2 access port has MTU of 1500 and the L3 SVI is 1500 (L2 access port and L3 SVI in same switch). I think I just need to change the L2 port MTU in DC2 switch to 9000. I will make this change asap and let you know the outcome. 

[MHM Cisco World]

Friend you use LO as update source but you use different IP under bgp neighbor.

Just check IP.