[c-nsp] dumb BGP ipv6 peer group question

Nick Hilliard nick at foobar.org
Wed Dec 28 08:37:56 EST 2011

On 27/12/2011 23:25, John Brown wrote:
> I'm trying to turn up some IPv6 peers.  Most of them are already IPv4 peers.
> I have a peer group defined in the IPv4 world.
> I try to do
> cr01.lax(config-router)#address-family ipv6 unicast
> cr01.lax(config-router-af)#neig 2001:XXX:YY::ZZ remote AAAAA
> cr01.lax(config-router-af)#neig 2001:XXX:YY::ZZ peer Any2IX-LAX-PEER
> And I get the following error
> % Members of peer-group must use the same transport


In fact, peer groups were not primarily designed for simplifying
configuration, but rather were implemented to allow the IOS BGP engine to
optimise CPU usage by grouping together similar peer configurations and
consequently rib calculations.  So from the outset, what you are trying to
do here is outside the scope of peer groups.


More information about the cisco-nsp mailing list