BGP route-poisoning ??

[pkhatri]

Hi,

A question for the Cisco insiders, I guess..

Here's an excerpt from the KnowledgeNet BGP 3.0 course:

"Only the route selected as best is propagated to the neighbors. However, a route is never sent back on the same BGP session that it was received. On the contrary, when a neighbor is selected, the best next-hop, the local router, makes sure that the neighbor is not pointing back to the local router by poisoning the route and sending a withdraw message to that neighbor.

This is to avoid a potential routing loop problem where the neighbor router selected as the best next-hop relies on the local router as the best next-hop."

I have not seen evidence of such "poisoned" routes being sent in my testing. Is this possibly something that used to happen on older IOSs ?

All help appreciated.

Regards,

Paresh.

[Harold Ritter · Accepted solution]

The BGP poisoned reverse should ocur under the following circumstances.

router A advertises route X to router B and all of a sudden router B starts advertising route X to router A and router A selects it as a best path, router A sends back a withdrawal message to router B.

If you test this exact scenario you should see the withdrawal message.

Hope this helps,

[pkhatri]

Thanks, that does help.

Paresh