BGP Confederation - far-ends announcements not being seen

RouterOS Forwarding Protocols
1

Hello,

I have a simple 3 AS BGP Confederations - 1,2 and 3. All belong to the outer AS ‘10’.

AS1 <—> AS2 <—> AS3

Announcements from AS1 to AS2 and vis-versa are fine.
Announcements from AS2 to AS3 and vis-versa are fine as well.

However, from AS3, I would like to see the routes announced to AS2 from AS1, re-announced. Meaning, I want to get to AS1 routes through the BGP path.

The “redestribute Other BGP” seems to be for other instances - as I only have one.
I also tried “ignore AS path” and that didn’t help either.

Do I need to create some prefix lists of some sort?

Thanks in advance.

2

I believe I figured it out and I have the each other’s far-end routes being announced through.
Here is what I did … not 100% sure if this is the correct way, so please comment.

On AS2s edge router, which is the glue between the two peering ASs, I created an aggregate item with a summary of the subnet of AS1 and listed it as “BGP-OUT-FROM-AS1”.

I then when to the BGP peer item for AS3, and put in as the outfilter the “BGP-OUT-FROM-AS1”.

Whol-la .. AS1’s routes are now showing up in the received routes on the AS3 edge router.

Lastly, I reversed the above for the traffic to get back the new path by creating a BGP-OUT-FROM-AS3 and applying that to the peer facing AS1.

Again, I am not 100% if this is the best way to handle this. Thoughts? Comments?

Thanks,

3 
AS 1
/routing bgp instance
set 0 confederation-peers=2,3

AS 2
/routing bgp instance
set 0 confederation-peers=1,3

etc…
4

Thanks for the reply Blake. AS1 isn’t directly connected to AS3, so that will work? I obviously have this set on the instance on AS2 as it’s the “man in the middle” - but didn’t think do to this on AS1 or AS3 since they are not direct talkers.

5

You’re right, its only needed on AS2.

Can you post your BGP peer and filter configs?

6

Sorry for the delayed reply. Here is my config on the inter-connected AS2 router.

Let me preface that again, I am not sure if this is the best way to accomplish this. I take it that there is probably several ways to solve this …


add address-families=ip as-override=no default-originate=never disabled=no hold-time=3m in-filter="" instance=default multihop=no name=\
    ibr1.gpt1 nexthop-choice=default out-filter=BGP-OUT-1 passive=no remote-address=172.29.0.2 remote-as=1 remove-private-as=no \
    route-reflect=no tcp-md5-key="" ttl=default update-source=172.29.0.1 use-bfd=no

add address-families=ip as-override=no default-originate=never disabled=no hold-time=3m in-filter="" instance=default multihop=no name=\
    ibr1.mob2 nexthop-choice=default out-filter=BGP-OUT-3 passive=no remote-address=172.29.0.6 remote-as=3 remove-private-as=no \
    route-reflect=no tcp-md5-key="" ttl=default update-source=172.29.0.5 use-bfd=no




add advertise-filter=BGP-OUT-3 attribute-filter="" disabled=no include-igp=no inherit-attributes=yes instance=default prefix=\
    subnet.from.as1.0/24 summary-only=yes suppress-filter=""
add advertise-filter=BGP-OUT-1 attribute-filter="" disabled=no include-igp=no inherit-attributes=yes instance=default prefix=\
    subnet.from.as3.0/24 summary-only=yes suppress-filter=""