1- Why routes aren't seen as receiving routes from VMX2 when they are really advertised from VMX2 to VMX5 ?
Because routes that fail BGP AS Path or cluster-ID sanity checks, as well as VPN routes that do not have a matching target community, do NOT get put into the RIB-IN table, thus you will never see it.
https://community.juniper.net/discussion/route-advertized-through-bgp-not-seen-in-received-routes-on-the-neighbouring-router#bm28cc22d8-09d3-4aaf-9a44-2dc9ca2d610a
When you fixed the AS-Path issue, you could see it.
2- Why the path taken from the loop interface aren't successful and pinng sourcing from 10.12.15.1 are successful ?
What were the exact traceroute commands and traceroute results (paste them here?)?
Are you sure the path was "64515 to 64512 and end to VMX4" as opposed to going via 64516 first?
Also, looking at the config you only did the override and advertise-peer-as on one side of the network.
I don't see it on VMX1 and VMX3, who would now be trying to advertise 10.15.101.1, learnt from AS64515 (vMX 5), to AS64515 (vMX 4)
The same scenario is true for 10.15.16.1, except, this would only be the case, if the route for 10.15.16.0/24 is the one learnt from vMX5, as opposed to vMX6.
If it's the one learnt from vMX6, then that last one is a bit weird and should work.
Original Message:
Sent: 04-03-2025 22:16
From: xine32
Subject: BGP override feature and AS path loop
H,i,
I'm currently studdying for my next Juniper certification exam and as for business need around ASPATH override feature and similar function. I know my design do not respect BGP best practice at some point, my only gold was to experiment AS-OVERRIDE feature with Juniper v-lab .
So to understand AS override I used Juniper V-LAB "Multi-AS with Dual Route Reflectors" topologie esased all router and replace their configuration with the one attached. At the beginning of my lab all BGP session were established and routeur are exchange as expected.
So to experiment AS-override feature I added the following
on lo0 I added multiple secondary ip on each router VMX4, VM5 & VMX6
VMX4 10.1.101.1/24 10.1.102.1/24..... UP TO 10.1.105.1/24
VMX5 10.15.101.1/24 10.16.102.1/24..... UP TO 10.16.105.1/24
VMX6 10.16.101.1/24 10.16.102.1/24..... UP TO 10.16.105.1/24
After alterated export policy to respective ASN router were exchanged everywhere as expected.
After that I change the ASN of VMX4 from ASN 64501 to 64515 same as for VMX5, change the respect export policy to ASN64512 on VMX4 and commit the change. I also update peering configuration of VMX3 & VMX1 for peering configuration with vmx4 and also commit the change.
As expected, routes from VMX4 now appear in all router part of ASN64515 except on VMX5 since BGP path prenvention mecanism was prevent the advertisement of these route form VMX2 and VMX6 to VMX5.
On VMX2 I wasn't seen even route been advertised from router VMX2 to VMX5 which this behavior is different from other vendor like who in case vmx2 would be advertising these route to vmx5 and let vmx5 to not accepting theses routes due of BGP path loop prevention function.
After I add the command "advertise-peer-as" in the group to VMX5 and commit the configuration
I seen the route from VMX4 been advertived to VMX5 but on VMX5 I did not seen them as received but not been put in the routing table due of BGP loop prevention. (Adding "advertise-peer-as" command in the peering configuration with VMX2 did not help better, in fact it did not changed anything in the observe behavior.... I wasn't think this change something, but I tried just in case....)
on VMX6 I updated the peering configuration to VMX5 with "advertise-peer-as as-override" command.
------------------------------
Xine
------------------------------