cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
2920
Views
0
Helpful
3
Replies

BGP - Inject Default Route

jason crockett
Level 1
Level 1

Hello,

Our network is running MPLS using BGP at several remote sites. Some of these remote sites do not have a separate internet connection so their internet traffic is backhauled to our data center on the east coast. We are opening an additional data center near the west coast and would like to backhaul internet traffic at our remote offices on the west coast to this new data center. But right now I am trying to understand how our internet traffic is currently being backhauled to the east coast data center. I understand how this traffic leaves the remote site MPLS router but what I do not understand is how it terminates at our data center MPLS router. I know one way of doing this is to add the neighbor x.x.x.x default-originate command to the data center MPLS so that it sends the default route 0.0.0.0 to it's neighbor. But I do not see this command in the BGP configuration on the data center MPLS router. How else can this be done?

Here is the BGP configuration on the data center MPLS router.

router bgp 12345

bgp log-neighbor-changes

neighbor D3 peer-group

neighbor D3 remote-as 5678

neighbor D3 soft-reconfiguration inbound

neighbor D3 route-map D3-IN in

neighbor 10.64.99.1 remote-as 12345

neighbor 10.64.99.1 next-hop-self

neighbor 10.64.99.1 soft-reconfiguration inbound

neighbor 10.64.99.16 remote-as 12345

neighbor 10.64.99.16 next-hop-self

neighbor 10.64.99.16 soft-reconfiguration inbound

neighbor 69.269.68.169 peer-group D3

3 Replies 3

jason crockett
Level 1
Level 1

Here is some more info that may be helpful.

Remote office router BGP configuration

We have dual MPLS providers at this site. Both terminate in the same 2911 router.

router bgp 12345

no synchronization

bgp log-neighbor-changes

network 10.50.1.15 mask 255.255.255.255

network 10.50.15.0 mask 255.255.255.0

network 10.50.115.0 mask 255.255.255.0

network 172.16.15.0 mask 255.255.255.0

network 172.16.115.0 mask 255.255.255.0

neighbor D3 peer-group

neighbor D3 remote-as 5678

neighbor D3 soft-reconfiguration inbound

neighbor D3 route-map D3-IN in

neighbor D3 route-map D3-OUT out

neighbor MA peer-group

neighbor MA remote-as 19855

neighbor MA soft-reconfiguration inbound

neighbor MA route-map MA-IN in

neighbor MA route-map MA-OUT out

neighbor 10.2.0.41 peer-group MA

neighbor 62.15.21.190 peer-group L3

no auto-summary

Peter Paluch
Cisco Employee
Cisco Employee

Hello,

Your BGP is currently not configured to inject a default route. Do you see it in your routing tables on any router, learned by BGP? If yes then it may be injected there by your ISP instead of your own BGP routers. You could enter the

show ip bgp 0.0.0.0

to see who advertises this network. Ideally, if you have access to some of your locations that use the backhaul to transport internet traffic to your data center, you should inspect their routing tables carefully, as the answer must already lie there - if the internet is working, the routing tables must already be set up properly.

One comment: you are using the neighbor X.X.X.X soft-reconfiguration inbound commands in your BGP configuration. This command was used years and years ago to permit your router to store the complete unfiltered database of all routes the neighbor X.X.X.X sent to you. This allowed your router to apply new inbound policies without dropping the BGP peering and retransmitting the prefixes again. It was a naive workaround around the former BGP limitation in which you could not ask your neighbor to resend you its BGP database. The obvious drawback is an increased memory footprint. Since RFC 2918 which dates back to September 2000, however, the BGP is extended with a native capability to request the neighbor to resend all routes of a particular address family (the so-called ROUTE REFRESH), and hence, this soft-reconfiguration inbound is obsolete. You should remove these configuration lines as you do not need them, and they make your router only to consume more memory.

Best regards,

Peter

Thanks for the response. I definitely do not see BGP configured to inject a route. That is why I wasn't sure how the remote sites know to send internet traffic to the data center MPLS router. Perhaps it is injected by the ISP. I will reach out to them.

Here is the show ip bgp 0.0.0.0 command run from the remote office router. It's using the 3rd path because we manipulated the weight to 3000.

Remote Office Router#show ip bgp 0.0.0.0

BGP routing table entry for 0.0.0.0/0, version 89405

Paths: (4 available, best #3, table default)

  Not advertised to any peer

  19855 19855

    10.2.0.41 from 10.2.0.41 (64.47.0.79)

      Origin IGP, localpref 100, valid, external

      Extended Community: RT:19855:88088226

  19855, (received-only)

    10.2.0.41 from 10.2.0.41 (64.47.0.79)

      Origin IGP, localpref 100, valid, external

      Extended Community: RT:19855:88088226

  5678 5678

    62.15.21.189 from 62.15.21.189 (206.165.188.153)

      Origin IGP, localpref 100, weight 3000, valid, external, best

  5678 5678, (received-only)

    62.15.21.189 from 62.15.21.189 (206.165.188.153)

      Origin IGP, localpref 100, valid, external

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: