Send BGP full table to an peer

[sebastien3]

Hello,

I have an ASR1001 with 2 full view BGP. I would like to send the full table of ASR1001 to another peer.

The peer wants me to be their second IP transit.

What type of configuration to perform and the right method to send a full view ?

Thank !

[Georg Pauwen]

Hello,

by default, if you do not filter anything, a peer ends the entire (full) BGP routing table to its other peer. What exactly do you mean by:

-->  I have an ASR1001 with 2 full view BGP

What is '2 full view' ?

and

--> The peer wants me to be their second IP transit.

What do you want to accomplish ? Which peer needs to do what ? A schematic drawing of the topology you have in mind would be helpful...

[MHM Cisco World]

Best Path or full route what is different ???
for example R1-R2-R3 
if the R1 has neighbor to R2 only then the R2 will send the Best Path "select best path of route from R1" to R3
if the R1 has neighbor to R2 & R3 then the R3 will receive all route from R1.

[paul driver]

Hello

can you confirm - it sounds like you currently have bgp peering towards two isps and you would then like to advertise all prefixes being learned from one of these isp towards an additional bgp peer?

[Giuseppe Larosa]

Hello @sebastien3 ,

>>

I have an ASR1001 with 2 full view BGP. I would like to send the full table of ASR1001 to another peer.

The peer wants me to be their second IP transit.

Your router can send its own best path for each prefix  either coming from ISP1 or from ISP2 to this new peer.

By default only the BGP best path can be progatated to other peers for a given prefix.

There are some features like additional paths that could allow to advertise also a backup path, but given the number of routes in a full table is in the order of 880,000 it is not recommeded to try to enable this feature as it can overload your device or the peer device

As noted by others all you need is to configure the eBGP session with the new peer without any filter

Hope to help

Giuseppe

[sebastien3]

Hello all !

Here is the diagram of the BGP architecture.

BGP R1 router configuration is as follows for peer R2 :

!
interface GigabitEthernet0/3/1
 description * Connection to router BGP R2 *
 ip address A.A.A.2 255.255.255.252
 no ip redirects
 no ip unreachables
 no ip proxy-arp
 negotiation auto
!
!
router bgp XXXXX
 bgp router-id A.A.A.100
 no bgp enforce-first-as
 bgp log-neighbor-changes
 bgp graceful-restart restart-time 120
 bgp graceful-restart stalepath-time 360
 bgp graceful-restart
 neighbor A.A.A.1 remote-as 64512
 neighbor A.A.A.1 description BGP-FULL-VIEW
!
 address-family ipv4
  network A.A.A.0 mask 255.255.255.0
  network B.B.B.0 mask 255.255.255.0
  neighbor A.A.A.1 activate
  neighbor A.A.A.1 soft-reconfiguration inbound
  neighbor A.A.A.1 route-map PEERING_64512 in
  neighbor A.A.A.1 filter-list 60 in
  distance bgp 200 200 200
 exit-address-family
!
ip as-path access-list 60 permit ^64512_
ip as-path access-list 60 deny .*
!
route-map PEERING_64512 deny 10
 match ip address prefix-list AS64512_PREFIXES
!
route-map PEERING_64512 permit 20
 description *** PERMIT routes only for AS64512 ***
 match as-path 60
 set local-preference 84
!
ip prefix-list AS64512_PREFIXES description - subnet IPv4 [PEER R2] -
ip prefix-list AS64512_PREFIXES seq 5 deny Z.Z.Z.0/24
ip prefix-list AS64512_PREFIXES seq 10 permit 0.0.0.0/0
ip prefix-list AS64512_PREFIXES seq 20 permit 0.0.0.0/8 le 32
!

Here are the errors I get on R1 when I try to establish BGP session with R2 :

May 12 16:54:47.440: %BGP-5-ADJCHANGE: neighbor A.A.A.1 Up
May 12 16:55:24.257: %BGP-3-NOTIFICATION: received from neighbor A.A.A.1 3/1 (update malformed) 0 bytes
May 12 16:55:24.257: %BGP-5-NBR_RESET: Neighbor A.A.A.1 reset (BGP Notification received)
May 12 16:55:25.998: %BGP-5-ADJCHANGE: neighbor A.A.A.1 Down BGP Notification received
May 12 16:55:25.998: %BGP_SESSION-5-ADJCHANGE: neighbor A.A.A.1 IPv4 Unicast topology base removed from session  BGP Notification received
May 12 16:55:29.167: %BGP-5-ADJCHANGE: neighbor A.A.A.1 Up
May 12 16:55:43.530: %BGP-3-NOTIFICATION: received from neighbor A.A.A.1 3/1 (update malformed) 0 bytes
May 12 16:55:43.530: %BGP-5-NBR_RESET: Neighbor A.A.A.1 reset (BGP Notification received)
May 12 16:55:43.532: %BGP-5-ADJCHANGE: neighbor A.A.A.1 Down BGP Notification received
May 12 16:55:43.532: %BGP_SESSION-5-ADJCHANGE: neighbor A.A.A.1 IPv4 Unicast topology base removed from session  BGP Notification received
May 12 16:55:48.905: %BGP-5-ADJCHANGE: neighbor A.A.A.1 Up
May 12 16:56:15.684: %BGP_SESSION-5-ADJCHANGE: neighbor A.A.A.1 IPv4 Unicast topology base removed from session  NSF peer closed the session
May 12 16:56:15.684: %BGP-5-NBR_RESET: Neighbor A.A.A.1 reset (NSF peer closed the session)
May 12 16:56:15.684: %BGP-5-ADJCHANGE: neighbor A.A.A.1 Down NSF peer closed the session
May 12 17:02:46.271: %BGP-3-NOTIFICATION: sent to neighbor A.A.A.1 active 6/2 (Administrative Shutdown) 0 bytes
May 12 17:02:46.271: %BGP-5-NBR_RESET: Neighbor A.A.A.1 active reset (Admin. shutdown)
May 12 17:02:46.271: %BGP-5-ADJCHANGE: neighbor A.A.A.1 active Down Admin. shutdown

BGP configuration on R2 is very simple with no filters !

Now the R2 router no longer responds to the ping due to the above NSF errors !!!! I do not understand what is happening...

Thanks for your help !

[sebastien3]

SOLVED !!

R2 is an ASR1001 with 4Gb of RAM... By upgrading the router to 8 GB of RAM, no more NSF errors !!

[Giuseppe Larosa]

Hello @sebastien3 ,

>> R2 is an ASR1001 with 4Gb of RAM... By upgrading the router to 8 GB of RAM, no more NSF errors !!

4GB of RAM is not enough to host a BGP full view this was the root cause after loading some prefixes the R2 was going low in memory.

The message errors are misleading but you have found a working solution.

Hope to help

Giuseppe