cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
492
Views
0
Helpful
2
Replies

prefix

1. In the following link page 51, line 3, what does prefix mean? "You cannot have RTB generate a prefix for 160.0.0.0"

https://www.cisco.com/c/en/us/support/docs/ip/border-gateway-protocol-bgp/26634-bgp-toc.pdf?dtid=osscdc000283

2. What's the role this command " passive-interface "?

Thanks

2 Replies 2

omz
VIP Alumni
VIP Alumni

"You cannot have RTB generate a prefix for 160.0.0.0 without the generation of an entry for 160.10.0.0"

The rule for advertising an aggregate-address is that we must have at least one of the longer prefixes inside of our BGP table.

Giuseppe Larosa
Hall of Fame
Hall of Fame

Hello miracle_david@yahoo.com ,

BGP focuses on advertising prefixes. A prefix is made of a base subnet and a prefix length (short equivalent of the subnet mask) in the case in the BGP study guide the prefix is 160.0.0.0/8, but the actual subnet is 160.10.0.0/16. So the books describe methods that allow to generate the prefix 160.0.0.0/8 without also advertising the 160.10.0.0/16. The trick is to use  static route for 160.0.0.0/8 to null0 and then to advertise it either using a network command that will give origin i internal or by redistibute static that would lead to origin incomplete ?. The origin is a BGP attribute and internal is preferred over incomplete.

 

However, I'm not sure this example is totally accurate as the summary-only keyword should block the component route 160.10.0.0/16 (the most specific prefix in BGP terms).

 

2)  passive-interface

The passive-interface command is used in an IGP like OSPF, EIGRP or IS-IS or  to advertise the subnet associated to an interface without attempting to build an adjacency / neighborship over it.

In the context of eBGP it can be used to make the eBGP next-hop known to an iBGP peer for having advertised the transit subnet with the other AS in the iGP and the passive-interface allows to do this safely without the risk of creating an unwanted iGP relationship at IGP level with the router in the other AS (the eBGP peer and BGP next-hop).

In iBGP sessions the BGP next-hop attribute is not changed so to make it possibie for remote iBGP peers to accept the prefix either you use the passive-interface to advertise the transit link subnet in the iGP so that the BGP next-hop is reachable (= there is an IGP route for it) or you change the BGP next-hop to self using an additional command in defining each of the iBGP sessions neighbor x.x.x.x next-hop self.

Both approaches are possible.

More generally in the context of IGP routing the passive interface command can be used for all user facing Vlans where there is no need to build an iGP adjacency and a virtual default gateway can be provided using HSRP, or VRRP or GLBP the so called FHRP = First Hop Redudancy Protocol.

 

Hope to help

Giuseppe

 

Review Cisco Networking for a $25 gift card