Problem Peering

[matou]

Hello,

I have a problem announcing a prefix on the Internet with a new router
Here is the topology.
R1 has a BGP configuration with two ISPs with a full-view. the prefix 192.1.1.0/22 ​​is announced and this prefix is ​​well propagated on both ISPs. There is the ip route config 192.1.1.0 255.255.252.0 Null0
I just installed a new router to configure another bgp session with a Internet Exchange IX for peering.
R1 and R2 are connected in /31
R1(192.1.1.3/31) <--> R2 (192.1.1.4/31) and ospf is configured on both interfaces.
On R2, there is a default route: ip route 0.0.0.0 192.1.1.3
On R2 there is now BGP with IX, the BGP configuration is ok, and I announce 192.1.1.0/22 ​​but it does not propagate. The prefix 192.1.1.0/22 ​​is in the routing table with ospf though.

Thanks for your help.

[Harold Ritter]

Hi @matou ,

> On R2 there is now BGP with IX, the BGP configuration is ok, and I announce 192.1.1.0/22 ​​but it does not propagate.

You are probably referring to prefix 192.1.0.0/22, right? How do you originate this prefix in BGP on R2 (redistribution, network statement)? Could you please provide the configuration you use to originate this prefix in BGP?

Regards,

Harold

[matou]

Hi @Harold Ritter , yes right

R2 is n9k-c93180yc

ip prefix-list BGPv4_OUT seq 10 permit 192.1.1.0/22

route-map BGPv4_OUT permit 10
match ip address prefix-list BGPv4_OUT

router bgp 654321
router-id 192.1.1.4
no enforce-first-as
address-family ipv4 unicast
network 192.1.1.0/22
redistribute static route-map BGPv4_OUT
neighbor <IP ROUTE SERVER IX>
remote-as 123456
description rs1
address-family ipv4 unicast
route-map BGPv4_OUT out
soft-reconfiguration inbound always
neighbor <IP ROUTE SERVER IX>
remote-as 123456
description rs2
address-family ipv4 unicast
route-map BGPv4_OUT out
soft-reconfiguration inbound always

Thank you

[Harold Ritter]

Hi @matou ,

192.1.1.0/22 is not a valid prefix. 192.1.0.0/22 is.

Can you provide the output for the following commands from R2:

show ip route 192.1.0.0/22

show bgp ipv4 uni neighbor <IP ROUTE SERVER IX> advertised-routes

Regards,

[matou]

Hi @Harold Ritter 

192.1.0.0/22 is an exemple fictitious. The real prefix is ​​another one, It really exists.

n9k-01-XX-XX# show bgp ipv4 uni neighbor X.X.X.X routes advertised

Peer X.X.X.X routes for address family IPv4 Unicast:
BGP table version is 9568, Local Router ID is 192.1.0.3
Status: s-suppressed, x-deleted, S-stale, d-dampened, h-history, *-valid, >-best
Path type: i-internal, e-external, c-confed, l-local, a-aggregate, r-redist, I-i
njected
Origin codes: i - IGP, e - EGP, ? - incomplete, | - multipath, & - backup, 2 - b
est2

Network Next Hop Metric LocPrf Weight Path
*>l192.1.1.0/22 0.0.0.0 100 32768 i

[Harold Ritter]

Hi @matou ,

Thanks for the additional information. It looks like the prefix is being advertised to the IX. What is the issue then?

Regards,

Harold

[matou]

Hi @Harold Ritter ,

When I do tests from outside, for example from another AS member of peering, the request goes through IP transit and not through peering. And tools like "bgp he", "bgp tools" do not add new peers.
Can the command on R1 "ip route config 192.1.1.0 255.255.252.0 Null0" have an impact on propagation ?

Thank you

[Harold Ritter]

Hi @matou ,

> When I do tests from outside, for example from another AS member of peering, the request goes through IP transit and > not through peering.

You need to check in this other IX member if the route is received from the IX route server and why it prefers the path through your transit provider.

> And tools like "bgp he", "bgp tools" do not add new peers.

I am sure how and how often this information gets updated.

> Can the command on R1 "ip route config 192.1.1.0 255.255.252.0 Null0" have an impact on propagation ?

No, it doesn't

Regards, 

Harold

[matou]

Hi @Harold Ritter 

Thank you for reponse,

But, the other IX member has a looking glass tool and when I do command "show ip bgp as", there are just Transit IP routes for my ISP, but not peering routes does not exist.

[Harold Ritter · Accepted solution]

Hi @matou ,

Your best bet at this point would be to contact the support staff at the IX and work with them to see what could be the issue. 

Regards,

Harold

[matou]

Hi @Harold Ritter 

Fixed with IX support. Filtering issue on routes. Problem solved

Thank you

[Harold Ritter]

You are very welcome @matou and thanks for the feedback